Foreign direct investment task force report, March …...6 ECB Foreign direct investment – Task...

FORE IGN D IRECT INVESTMENTTASK FORCE REPORT

MARCH 2004

FORE

IGN

DIR

ECT

INVE

STM

ENT

– TA

SK F

ORC

E RE

PORT

MAR

CH 2

004

In 2004 all ECB publications will feature

a motif taken from the

€100 banknote.

FORE IGN D I RECT I NVE STMENTTASK FORCE REPORT

MARCH 2004

© European Central Bank, 2004

AddressKaiserstrasse 2960311 Frankfurt am Main, Germany

Postal addressPostfach 16 03 1960066 Frankfurt am Main, Germany

Telephone+49 69 1344 0

Websitehttp://www.ecb.int

Fax+49 69 1344 6000

Telex411 144 ecb d

All rights reserved. Reproduction foreducational and non-commercial purposesis permitted provided that the source isacknowledged.

As at November 2003.

ISBN 92-9181-483-0 (print)ISBN 92-9181-484-9 (online)

3c ECB

Fore ign d i rec t inves tment – Task force repor tMarch 2004

CONTENTSC ONTENT SEXECUTIVE SUMMARY 5

INTRODUCTION 16

Summary of the mandate 16

Structure of the report 17

1 CONCEPTUAL ISSUES RELATEDTO THE FULLY CONSOLIDATED SYSTEMAND THE COVERAGE OF INDIRECT FDIRELATIONSHIPS 18

Introduction 18

Companies with indirect links ofownership. 20

The case of “fellow” / “sister”companies 25

Conclusions 28

2 PRACTICAL ASPECTS RELATEDTO THE COVERAGE OF INDIRECT FDIRELATIONSHIPS 31

Introduction 31

Statistical consolidation versusaccounting consolidation 31

Current practices: results of thequestionnaire 38

Different consolidation approachesand the geographical allocation oftransactions and positions:impact on the compilation of theEuropean aggregates 40

Conclusions and recommendations 46

European database on ownershipstructures 47

3 VALUATION OF FDI EQUITY STOCKS 51

Introduction 51

Related decisions adopted by theStatistics Committee and the WorkingGroup on Balance of Payments andExternal Reserves 51

Results of the questionnaire onvaluation of FDI equity stocks 53

National feasibility studies on howto compile FDI in listed companies’shares on the basis of both marketvalues and book values 56

4 REINVESTED EARNINGS 65

Introduction 65

International standards concerningreinvested earnings and clarificationsadopted by the TF-FDI 65

Current collection and compilationmethods for reinvested earnings in theEuropean Union 69

International Accounting Standards(IAS) 73


5 ISSUES RELATED TO OTHER CAPITALIN FDI STATISTICS 77

Introduction 77

Current practices in the treatmentof Other Capital in the EU MemberStates 77

Conclusions 80

The application of the directionalprinciple: practical issues andempirical evidence 83

6 IDENTIFICATION AND TREATMENT OFSPECIAL PURPOSE ENTITIES (SPES) 87

Introduction 87

Reasoning for the identification ofSPEs and problems in dealing withSPEs 87

Definitions of SPEs in internationalguidelines 89

Importance of SPEs and offshorecountries in European statistics 91

Overview of the impact of offshorecentres on European aggregates 92

Results of the empirical exerciseon the importance of SPEs in EUMember States 95

Summary 96

Conclusions 97

4ECBFore ign d i rec t inves tment – Task force repor tMarch 2004

7 ALLOCATION OF FDI INWARD STOCKSBY COUNTRY OF “ULTIMATE BENEFICIALOWNER” (UBO) 98

Conceptual framework 98

Practical application in some EUcountries 99

EU aggregates and UBO allocation 101


8 GENERAL CONCLUSIONS ANDRECOMMENDATIONS 104

Introduction 104

Conclusions and recommendationsfor individual items 104

Prioritisation and timing forimplementation of the TF-FDIrecommendations 113

Issues for follow-up work 114

LIST OF ANNEXES 115

Annex 1: Compilation of FDI stocksat t+9 months 116

Annex 2: Definitional issues relatedto FDI/Other capital 118

Annex 3: Summary of the answersto the questionnaire onOther Capital 125

Annex 4: Application of thedirectional principle indifferent countries 128

Annex 5: Questionnaire on importanceof the main sub-itemsof “Other Capital” in FDIstatistics 131

Annex 6: National descriptionsof UBO-based FDIstatistics 133

Annex 7: Reinvested earnings innational accounts andimpact on GDP 137

Annex 8: List of members of theTask Force on ForeignDirect Investment 140

SUPPLEMENTARY DOCUMENT:SPECIAL PURPOSE ENTITIES 141

5cECB


EXECUTIVE SUMMARY

E X E CU T I V E S UMMARY

I N TRODUCT I ONSUMMARY OF THE MANDATE

1. The Task Force on Foreign DirectInvestment (TF-FDI) was set up by theWorking Group on Balance of Paymentsand External Reserves Statistics (WG-BP&ER), jointly with the Eurostat Balanceof Payments Working Group (BoP WG),to investigate on the matters included inthe mandate, focusing particularly on thepractical and consistent implementation ofthe various related principles/definitions.

2. The main objectives of the TF-FDI were toidentify “best practices” with a view tominimising inconsistent treatments withineuro area/European Union (EU) memberstates. In this respect, par-ticular attentionwas paid to the accuracy of thegeographical allocation of the relevanttransac-tions and positions.

3. The mandate of the TF-FDI was articulatedaround the five following items:

(i) Valuation of stocks: the TF-FDI wasmandated to investigate the practicality ofthe conceptual agree-ments reached by theSTC in this area, i.e. the problemsassociated with the provision of additionalbreakdowns required for listed and non-listed companies for the production ofmemoran-dum items.

(ii) Other capital: the TF-FDI was mandatedto investigate and put forward practicalsolutions for the problems linked to:

a) the application of the “directionalprinciple”;

b) the identification and impact of tradecredits, financial leasing, debtsecurities subscribed by associatedcompanies;

c) the treatment and classification oftransactions/positions on “othercapital” when Monetary and FinancialInstitutions (MFI) are involved.

(iii) Reinvested earnings: the TF-FDI wasmandated to review the practical aspects ofthe compilation meth-ods for reinvestedearnings (declaration by respondents,calculation by the compilers, interimestima-tions etc.).

(iv) Identification and treatment of SpecialPurpose Entities (SPEs): the TF-FDIshould investi-gate practical problemsassociated with the identification andtreatment of transactions/positions ofSPEs, mainly “other financialintermediaries” or “financial auxiliaries”,in the context of direct in-vestmentrelationships.

(v) Treatment of indirect FDI relationshipsand allocation of FDI inward stocks bycountry of “ultimate beneficial owner”(UBO): the TF-FDI should investigate thepractical application of the “fullyconsolidated system” in member states.Possible solutions to the problem ofobtaining information on group structuresshould be examined with reference also tothe costs that they would entail. Thepossibility and implications of classifyinginward FDI stocks by the country of UBOwould be analysed.

4. At their respective meetings in March2003, the ECB WG-BP&ER and theEurostat BoPWG gave further guidance tothe TF-FDI. The WG-BP&ER stressed that“departure from international statisticalstandards should only be proposed by theTF in case a substantial critical mass ofmember states were in favour and forsound practical reasons, mainly addressedto avoid distortions in the euro area/EUaggregates”.

5. Both working groups stated that theCurrent Operating Performance Concept(COPC) should be the reference conceptfor the compilation of reinvested earnings.It was also stressed that for the practicalapplication of this concept, some


simplifications could be needed so as tofacilitate its practical application.Additionally, the WG-BP&ER encouragedthe TF-FDI to further examine thepracticalities in compiling aggregates (inparticular as regards ‘extraordinaryprofits’ and any possible distinctionbetween financial and non-financialcorporations).

FEATURES OF PRESENT COLLECTION SYSTEMSAND MOST COMMON PROBLEMS6. The TF-FDI started its work with a reviewof member states’ current practices in order toidentify possible best practices as well asproblems encountered by the nationalcompilers. The analysis of current practicesshowed large differences in the applica-tion /non-application of international standards aswell as regarding data collection andcompilation methods.

7. The collection/compilation methods are atthe moment in a transitional period, where somemember states have started to move fromsettlement-based systems to systems based onsurveys and direct report-ing. In many cases thefeatures of the collection system used is crucialfor the application or non-ap-plication ofinternational standards.

POTENTIAL BENEFITS OF HARMONISINGCOLLECTION SYSTEMS IN THE FIELD OF DIRECTINVESTMENT8. The coexistence of different collectionsystems is to some extent at odds with the needto produce consistent European aggregates andguarantee a certain degree of homogeneity. It isobvious that sharing solutions to commonproblems as well as a shift towards furtherstandardisation should be deemed positive stepswhich may certainly provide substantialbenefits.

INDIRECT FDI RELATIONSHIPS9. Across the analysis carried out by the TF-FDI, it became obvious that the fifth item of themandate (mostly in relation to the coverage ofindirect FDI relationships) required a higher

prioritisation since it influenced how tointerpret the conclusions of most other items.For that reason, item (v) was split into threeparts, the first two referring to indirect FDIrelationships (from the conceptual and thepractical viewpoints, respectively) and the thirdone referring to the UBO.

I . CONCLUSIONS AND RECOMMENDATIONS

10. The most significant issues encountered inthe analysis of the items of the TF-FDI mandateare briefly presented below. The TF-FDI putspecial emphasis on the provision of practicalrecommendations and proper justification forthese recommendations. Additionally,appropriate prioritisation of the measuressuggested is provided, highlighting the mosturgent problems to be solved.

CONCEPTUAL ISSUES RELATED TO THE FULLYCONSOLIDATED SYSTEM AND THE COVERAGE OFINDIRECT FDI RELATIONS11. Firstly, the TF-FDI agreed that indirectrelationships should undoubtedly be covered byFDI statistics.1 As a basis for evaluating andrecommending “best practices” regarding thetreatment of indirect FDI relationships, the TF-FDI found it necessary to first agree on aunique interpretation of the internationalrecommendations, on a conceptual level.

12. In particular, two different types of indirectrelationships were identified: (i) parentcompany – affiliate and (ii) sister/fellowcompanies2, i.e. pertaining to the same group,but without either direct or indirect links ofownership. The TF-FDI agreed that, for thefirst category, flows and stocks should beclassified by the parent company as outward

1 This recommendation is very relevant for all FDI items, namelyequity capital, reinvested earnings and other capital. For a moredetailed analysis of the conceptual treatment suggested for thedifferent FDI components, please refer to chapter 1.

2 Both terms, i.e. “fellow” companies and “sister” companies areindistinctly used throughout the report to refer to the same kind ofcompanies, namely those pertaining to the same multinationalgroup but with neither direct nor indirect control over oneanother.

7cECB


EXECUTIVE SUMMARY

FDI and by the (indirectly owned) directinvestment en-terprise as inward FDI. In thecase of sister companies, flows and stocksshould be classified as outward FDI by thecountry which provides the investment orgrants the loan and as inward FDI for thecountry receiving the investment/loan.

13. The TF-FDI recommends that all memberstates agree to put into practice these mainrecommendations as well as the more detailedmethodology described in chapter I of the reportso as to avoid asymmetries and inconsistenttreatments.

PRACTICAL ASPECTS RELATED TO THE COVERAGEOF INDIRECT RELATIONSHIPS

SIMPLIFICATION PROPOSALS TOWARDS THECOVERAGE OF INDIRECT FDI RELATIONSHIPS14. The TF-FDI considers that a full applicationof the fully consolidated system (FCS) by allcoun-tries is virtually unfeasible on practicalgrounds. On the other hand, restricting FDIstatistics to only cover direct relationshipswould not be compliant`with internationalstandards and the outcome would offer a loweranalytical value. The most important difficultywas how to find practical ways for collectingthe necessary information, since the longer thechain of links between companies, the moredifficult it is to get access to the balance sheet offoreign subsidiaries with no direct link to thedomestic mother company.

15. In order to find an alternative solution to thefull application of the FCS, which at the sametime could be deemed consistent withinternational standards and easier to apply inpractice, the TF-FDI explored differentalternatives. In turn, the TF-FDI suggests asim-plification of the FCS rules as the minimumwith which all countries should be compliant.Such a minimum approach would narrow downthe risk of asymmetries and would reduce theimpact on the European aggregates of thedifferent methodologies applied in memberstates:

16. The two admissible simplifications thatshould constitute the bottom line for allpractises at the EU level would be:

(i) The coverage of indirect links of ownershipabove 50% (direct links of ownership above10% would still need to be covered)

(ii) The coverage of (direct and indirect) links ofownership above 10%, calculated as the simpleproduct of the subsequent links of ownershipalong a chain.

GEOGRAPHICAL DISTRIBUTION OF FDI FLOWS/STOCKS RELATED TO INDIRECT FDI LINKS17. The use of non-fully harmonised criteria forthe geographical allocation of countrycontributions to the European aggregates (inrelation to the existence of indirect FDIownership links) implies a high risk of doublecounting and/or missing information.

18. With a view to avoiding such a risk, the TF-FDI recommends that, for both reinvestedearnings and FDI equity stocks, all (indirect)FDI transactions/positions should begeographically allocated to the company withwhich the investor/direct investment enter-prisemaintains a direct link of ownership (immediateaffiliate or immediate parent company).

19. It is acknowledged that this criterion mayresult in less valuable statistics from ananalytical viewpoint. For this reason, the TF-FDI would encourage countries to collect andpublish additional information on thegeographical allocation of FDI flows and stocksbased on the residence of the ulti-matebeneficiary owner.

EUROPEAN DATABASE ON OWNERSHIPSTRUCTURES20. The TF-FDI analysed the issue on aEuropean database on ownership structuresfrom two differ-ent points of view: (i) aspotential data providers, and (ii) as users of theinformation. The TF-FDI acknowledged thatthe existence of a centralised database withinformation about the structure of multinational


groups would be seen as a very useful tool forthe compilation of FDI statistics

21. From the point of view of potential dataproviders, the main findings of the TF-FDIpointed out that the provision of the necessaryinformation would imply a number ofsignificant problems related to resources,confidentiality issues, technical problems, etc.

22. From the point of view of potential users,the TF-FDI is of the opinion that a harmonisedand multilat-eral solution should be highlywelcome. In this regard, the TF-FDI suggeststhat a solution could be explored through theongoing project on the construction of aEuropean Business Register currently underdevelopment by Eurostat in collaboration withthe ECB. It is also suggested that other bodies,for instance the ECB’s WG-BP&ER and theEurostat’s BoPWG, elaborate the list of userrequirements which would permit that the finalproduct could be used for the compilation ofFDI statistics.

VALUATION OF FDI EQUITY STOCKS

23. The TF-FDI considered practical problemsfor the implementation of the STC decisionsconcern-ing how to value FDI equity stocks. Noalternative market valuation methods wereconsidered in this analysis. The decisions of theSTC could be summarised as follows:

(i) FDI in listed companies’ shares shall bevalued on the basis of stock exchangeprices in the euro area i.i.p.

(ii) FDI in non-listed companies’ shares shallbe valued on the basis of book values inthe euro area i.i.p.

(iii) Book values consist of the application ofownership percentages to the sum ofselected accounts extracted from theliabilities side of the target FDI company’sbalance sheet (according to the commondefinition of OFBV).

(iv) Two (four) memorandum items will becompiled on a centralised way: inward andoutward euro area FDI based on market valuesand book values for all types of companies,respectively (with no geographical or sectordetails).

(v) To this aim, inward and outward FDI equitystocks should be reported to the ECB with asplit between listed and non-listed FDIcompanies, and FDI stocks in listed companies’shares should be reported on the basis of bothmarket values and book values.

24. In reviewing all possible practical problemsthat the implementation of all these proposalscould entail, the TF-FDI considered the absenceof FDI surveys for the compilation of stockstatistics as a major difficulty. Such a problemhas implications on the ability of certaincountries to implement the decisions adopted bythe STC as regards valuation of FDI equitystocks.

The TF-FDI is of the opinion that thecompilation of FDI stocks should be based oninformation collected via FDI surveys. Theprovision of annual FDI stocks based onaccumulation of b.o.p. flows should bediscontinued as soon as possible. In relation tothis subject, the TF-FDI ranks this issue as thefirst priority for any follow-up work subsequentto the delivery of this report.

25. Concerning practical solutions to collect thenecessary information to comply with the STCagreements, the TF-FDI makes the followingrecommendations:

DISTINCTION BETWEEN LISTED AND NON-LISTEDCOMPANIES26. The TF-FDI considered the following aspossible and acceptable information sources:

(i) registers of (resident) listed companiesmaintained by stock exchange authorities;

(ii) information provided by respondents;

9cECB


EXECUTIVE SUMMARY

(iii) manual distinction based on internaldatabases and/or publicly availablesources (e.g. financial press, stockexchange web sites, etc.)

VALUATION OF STOCKS IN LISTED COMPANIES27. On the basis of the results of individualnational feasibility studies carried out by sixcountries, The TF-FDI has come to theconclusion that:

The collection of FDI equity stocks for listedcompanies on the basis of two differentvaluation methods (market values and bookvalues on the basis of the common definition ofOFBV) can be deemed feasible and not toocostly for countries running FDI surveys

Good/acceptable practices– the most feasible way to collect market

values and book values is through theinformation pro-vided by respondents viathe addition of supplementary questions tothe FDI surveys.

– Additionally, individual valuation methodsbased on stock exchange prices combinedwith inter-nal databases and publiclyavailable information have also proved to bea viable way to get information on marketvalues, especially in the case of inward FDI.

Non-acceptable practices:– Leaving the choice to respondents on the

valuation criterion (market values or bookvalues) they wish to use to report FDIstocks. This can neither ensure theprovision of the nec-essary information tothe ECB nor guarantee the compilation ofconsistent FDI equity stocks.

– Application of perpetual inventory methods/accumulation of b.o.p. flows. 3 This relieson the reasons previously explained.

COMPILATION OF FDI STOCKS AT T+9 MONTHS28. The TF-FDI concluded that, in the currentsituation, only four member states are already ina position to provide pure stocks data based on

surveys within the required timeliness. Theothers can only accumulate flows to the lastavailable stock (perpetual inventory method),usually adjusted for exchange rate changes, andin a few instances for price changes. Theprovision of data with the re-quiredgeographical breakdown (shortly on step-3basis) does not seem to pose significantproblems for most countries.

REINVESTED EARNINGS

29. A review of current practices revealed thatsome member states have not yet established asys-tem to calculate/estimate reinvestedearnings.

The TF-FDI deems the non-inclusion ofreinvested earnings as the most crucialproblem in this area. This difficulty seems to beclosely connected with the lack of FDI surveys,which should be resolved promptly, in line withthe proposal made for FDI equity stocks.

30. All other TF-FDI recommendations arebasically determined by how reinvestedearnings (RIE) are calculated. RIE arecalculated as the difference between twovariables: total profits from current operationsand dividends payable. The first component isnormally available later than the second oneand, hence, RIE (or rather total profits) areoften temporarily estimated from the projectionof total profits as presented in the last availableFDI survey.

TOTAL PROFITS31. International standards prescribe theapplication of the Current OperatingPerformance Concept for the measurement oftotal profits, excluding e.g. extra-ordinarygains and losses. Only five EU countries arecompliant with this so far. In considering theneed to adapt systems to the application of theCOPC, the TF-FDI suggests two types of

3 Exception made of the delivery of provisional estimates by end-September (where applicable) and of real-state investments.


information sources, both connected with theaccounting statements of the respondents: (i)companies’ public accounts and (ii) restrictedinformation internally available to thecompanies.

32. Although the split between ordinary andextraordinary gains/losses in accountingstatements is not necessarily consistent withstatistical definitions, it was considered by theTF-FDI to be an imme-diately available proxyfor the time being. The first information source(public annual accounts) on its own cannot beconsidered as an acceptable proxy for the COPCwithout additional information internallyavailable to respondents, notably, thegeographical breakdown of the information.Therefore, a combi-nation of both informationsources (i.e. public accounts and internalinformation) would be necessary in any case.

33. The development of the new IAS will implya more specific definition of the componentswhich may serve as a firm basis for theharmonisation of member states’ application ofthe COPC. However, the devel-opment of thenew IAS may pose an additional difficulty forcompilers to properly apply the COPC, to theextent that only very exceptional results will beexcluded from the ordinary profits and losses.

34. The TF-FDI concluded that inconsistenttreatments caused by different practices implyserious distortions for the euro area/EU currentaccount. Therefore, since the data necessary fora COPC valuation of profit is available from therespondents’ accounting:

(i) The same concept for the compilation oftotal profits, namely the COPC, should beused by all member states and exceptionalresults should be appropriately excludedfrom the current account.

(ii) As current practices within the EU indicatehow difficult this may be on practicalgrounds, the TF-FDI concluded thatacceptable solutions for the application ofthe COPC should aim at covering at least

the reduced number of companies whichcontribute the most to extraordinaryresults.

35. Concerning the second recommendationabove, the experiences of some member states isthat a reduced number of companies involved inFDI relations contribute to most of theextraordinary gains/losses. For othercompanies, the all-inclusive approach may beapplied, since it often provides similar results tothe COPC.

An acceptable practice would therefore be toapply the COPC, as a minimum, only to suchcompanies (namely the biggest ones plusholding companies) in each Member State, andto collect the rest on an all-inclusive basis.

DIVIDENDS36. Time of recording: international standardsprescribe the recording of dividends whenpayable rather than when they are paid. Theforeseeable increase in the use of directreporting through surveys may bring thepractices closer to international standards, asthey are likely to reflect accruals-basedaccounting data. This changeover will,however, not take place in the short term.Nevertheless, asymmetries will only occur inshort time spans, since the difference betweenpayable and paid is usually only a matter of timeallocation during a fairly limited period.

37. The treatment of dividends stemming fromexceptional capital gains may be a problem in sofar as it affects the calculation of reinvestedearnings. While exceptional results are notincluded in total profits (according to the COPCdefinition), it is questionable whether or not,once payable, they should be recorded in thecurrent account as income on direct investment.

38. As regards the provision of funds toaffiliates to cover losses, some countries recordit as nega-tive dividends, while others record itin the financial account.

11cECB


EXECUTIVE SUMMARY

CONCLUSIONS– Concerning the time of recording, the TF-

FDI suggests that member states keep onwith their current practices for practicalreasons and due to the limited impact on theresulting statistics in longer time spans.However, member states are requested toswitch from dividends paid to dividendspayable when moving towards directreporting systems.

– Concerning payment of dividends stemmingfrom exceptional capital gains, inaccordance with international standards, theTF-FDI recommends their recording in thefinancial account as FDI disinvestments,thus not entering in the calculation of RIE.

– As to contributions to cover losses in directinvestment enterprises, in line withinternational recommendations, the TF-FDIproposes that these transactions should berecorded in the Financial Account, asadditional investment flows and not asdirect investment income.

39. Finally, as a reference to therecommendations related to the treatment ofindirect FDI relation-ships, it should also benoted that the coverage of reinvested earningsgenerated by indirectly related direct investmententerprises should at a minimum meet one of thefollowing simplified rules:

(i) The coverage of indirect links of ownershipabove 50% (direct links of ownership above10% would still need to be covered)

(ii) The coverage of (direct and indirect) links ofownership above 10%, calculated as the simpleproduct of the subsequent links of ownershipalong a chain.

OTHER CAPITAL40. The TF-FDI tried to seek clearer guidanceon the inclusion/exclusion of some borderlinecases within FDI other capital. In particular, theTF-FDI addresses the followingrecommendations:

– Preferred shares should be excluded fromother capital and recorded as DirectInvestment/Equity capital unless they takethe form of non-participating shares

– Permanent debt (e.g. subordinated loans,perpetual bonds, etc.) should be included inDirect Invest-ment/Other capital, regardlesswhether or not it takes the form ofsecurities.

– Trade credits, financial leasing, and anyother type of inter-company loans should beincluded in Direct Investment/Other capital,while financial derivatives (in accordancewith final agreements between the ECB andthe IMF) should be excluded from FDIstatistics.

– When both parties involved in lendingactivities are MFIs, financial intermediariesor financial auxiliaries, only permanent debtshould be included in Direct Investment/Other Capital. This rec-ommendation couldraise some confidentiality concerns in somemember states, as the granting of permanentdebt to affiliate companies in the bankingsector is usually rather limited. In suchcases, the contributions to the Europeanaggregates could be flagged as confidential.

41. Additionally, the TF-FDI particularly triedto find practical solutions to collect thenecessary information from reporting agents. Inthis framework, the TF-FDI is of the opinionthat the two main problems concerning FDIother capital are:

(i) the incomplete coverage of both transactionsand stocks between affiliated companies,such as securities and trade credits, lendingactivities between fellow companies (i.e.companies with the same ultimate parentcompany but not belonging to the sameownership chain), etc. and

(ii) the partial application of the directionalprinciple by some member states.


42. The general collection methods for FDIflows and stocks can be split into two maincategories: survey-based and settlement-basedsystems. The practical solutions acknowledgedby the TF-FDI for a consistent application ofthe directional principle could most probablynot be deemed very innovative, but no otheralternatives have been found.

43. For survey-based systems, the TF-FDI is ofthe opinion that the most effective way tocollect the necessary information would be theaddition of questions to the survey form,requesting separately each element of othercapital and taking into account the directionalaspect of the investment. One alternative to thedirect request of separate information fromreporters could be to instruct the reporters onhow to reclassify (from inward to outward FDIor vice versa) the funds provided by affiliates totheir parent companies.

44. For settlement-based systems, the codesused to collect information from reportersshould be expanded (where necessary) toinclude the elements of other capital required.They should also in-clude information on thedirection of the investment to satisfy therequirements of the directional prin-ciple. TheTF-FDI recommends that instructions toreporters should also be expanded to specify therequirements. A database on FDI relationshipsof resident companies, is a useful tool to ensurethat any other capital transaction involvingaffiliated companies is effectively recordedunder direct investment, although itsmaintenance normally requires a significantamount of resources.

IDENTIFICATION AND TREATMENT OF SPECIALPURPOSE ENTITIES (SPES)45. Due to the increasing role of SpecialPurpose Entities in the provision of intra-groupfinancing and other services, the TF-FDIexamined (i) the appropriateness of collectingseparate statistics for this type of companies;and (ii) whether an alternative treatment fortransactions and positions in which SPEs areinvolved should be applied. Concerning the

second point, the TF-FDI considered thepossibility of “passing through” this kind ofenterprises (i.e. do not record either assets orliabilities) in those cases in which SPEs do notcarry out any real economic activity in theterritory in which they are located.

The TF-FDI disregarded both options (i.e. adifferent treatment and the collection ofseparate statistics) on the grounds thatinternational standards recommend treatingSPEs as any other FDI enterprise (exceptionmade for some special cases4) and do notrequire any separate statistics for this kind ofinstitutions. Additionally, the non-existence ofa single harmonised definition of SPE wouldhamper their identification as well as theapplication of different rules for the recordingof transactions and positions in which theseentities are involved.

At present, most countries do not distinguishtransactions/positions with non-resident SPEsfrom those with any other foreign counterpart.Furthermore, most member states do notseparately identify domestic SPEs in theirregular statistics.5 A change in the methodologyapplied to these companies would be confrontedwith the difficulty to calculate consistenthistorical series.

– Against this background, the TF-FDIrecommends the inclusion of transactions/positions of/with SPEs or SPE-likecompanies in b.o.p./i.i.p. reportingconcerning the contributions to the euroarea/EU aggregates.

– Notwithstanding all the practical andconceptual difficulties previously stated, theTF-FDI recommends that the possibility tocollect separate statistics for SPEs continue

4 For instance, in the case of holding companies (for which areclassification in the economic sector of activity isrecommended) or SPEs with a financial nature (for which it isrecommended excluding from FDI statistics intra-group lendingand borrowing vis-à-vis other related corporations with afinancial nature).

5 One country excludes SPEs from national statistics, since, if thatwere not the case, national statistics would be blurred by thevolume of financial transactions between non-resident entitieschannelled through domestic SPE’s.

13cECB


EXECUTIVE SUMMARY

being assessed by both working groups andin the framework of ad-hoc workshops inthe future.6

– Following the latest decisions of the IMFand the ECB, SPEs principally engaged infinancial intermediation for a group ofrelated enterprises should be included inthe category of affiliated finan-cialintermediaries and, therefore, inter-company loans with any other institutionincluded in the category of MFIs/affiliatedfinancial intermediaries should be excludedfrom direct investment and should berecorded in other investment.

ALLOCATION OF FDI INWARD STOCKS BYCOUNTRY OF “ULTIMATE BENEFICIAL OWNER”(UBO)

46. Following its mandate, the TF-FDI analysedthe possibility and implications of classifyingin-ward FDI stocks by the country of the UBO.The compilation of FDI statistics based on theUBO prin-ciple implies allocating FDI stocksaccording to residence of the entity that exercisecontrol on the capital stock considered.

47. The TF-FDI assessed the impact of applyingthe UBO principle on the intra/extra-EUallocation and concluded that such impact wassignificant in most of the cases analysed(namely on two out of the three countries forwhich this information was available)

DEFINITION OF UBO48. Ultimate beneficial owners are the firstpersons proceeding up along the chain that arenot con-trolled by any other company. Thisdefinition should be applied at least to equitycapital.

PRACTICAL METHODS TO COLLECT INFORMATIONBASED ON THE UBO PRINCIPLE49. Two approaches were identified by the TF-FDI to apply the UBO through the FDI sur-veys: (i) direct collection of UBO-based FDIstocks from respondents; and (ii) calculation bythe compiler on the basis of more basic

6 To this aim, co-ordination should be ensured with the relatedwork currently being developed in the OECD.

information (e.g. on all intermediate ownersplus percentages of ownership) collected fromrespondents.

CALCULATION OF EU AGGREGATES BASED ON THEUBO ALLOCATION.50. In the case of the EU aggregates, theapplication of the UBO criterion may implysome double recording related to the inwardFDI stocks held by non-euro area countries. Forthis reason, the TF-FDI recommends that theUBO should only be applied in those cases inwhich EU direct in-vestment companies aredirectly owned by an investor located in anextra-EU country. However, the value of theFDI equity stocks controlled by extra-EUcountries should also reflect the con-solidatedvalue of the group, including other affiliates(inside or outside the EU/euro area), in linewith the recommendations related to thetreatment of indirect FDI relationships.

51. The TF-FDI did not hold a conclusivediscussion on the practical ways in which theseproposals could be implemented in practice.Therefore, it is proposed that some further workin this area should be part of the follow-up tothe TF-FDI.

PRIORITISATION AND TIMING FORIMPLEMENTATION OF THE TF-FDIRECOMMENDATIONS

52. The TF-FDI was requested by the StatisticsCommittee to provide an appropriateprioritisation of its recommendations, with aclear emphasis on those actions which wereconsidered more urgent for the quality of FDIstatistics. Following this request, the TF-FDIhas divided its recommenda-tions into threecategories according to the potential distortionsthat departing from its recommen-dations couldentail for the European aggregates: high,medium and low importance, respectively.


53. Additionally, a second dimension refers to theeffort that each individual action would requirefrom member states and the time lag with whichthe application of its recommendations could bereasonably expected. On that basis, the proposedactions could also be classified in three addi-tionalcategories: short, medium and long-term. The TF-FDI did not intend to define (in terms of more

Importance

Timeframe High Medium Low

Short-term • All countries should start compiling • Contributions toFDI equity stocks and reinvested cover losses ofearnings on the basis of the results direct investmentof FDI surveys, at least annually. 1) enterprises should

• FDI equity stocks should be collected be recorded in theseparately for listed (both book 2) and financial account.market values) and non-listed companies.

• All indirect FDI relationships 3) shouldbe conceptually treated in accordancewith the interpretation of standardsoutlined in chapter 1.

• All (indirect) FDI transactions/positionsshould be geographically allocatedto the immediate affiliate or parentcompany.4)

Medium-term • The COPC should be used by all MS. 5) • Contribute to the development of a• The components of other capital should European database with information

be identified on the basis of the about the structure of multinationalrecommendations provided in chapter 6. groups.

• Payment of dividends from exceptionalcapital gains should be recorded in thefinancial account (thus not entering inthe calculation of RIE).

Long-term • Indirect FDI relationships 6) should • Dividends shouldcover in practice (as a minimum) be recorded wheneither (i) indirect links of ownership payable ratherabove 50%; or (ii) direct and indirect than when paid.links of ownership above 10%,calculated as the product of thesubsequent links of ownership alonga chain.

• The directional principle shouldbe (fully) applied by all member statesfor FDI flows and stocks.

1) Exception made of provisional results to be provided at T+9 and real-state investments. The following non-acceptable practices shouldbe abandoned: (i) to leave the choice to the respondents on the valuation criterion (market values or book values); and (ii) the applicationof a perpetual inventory method/accumulation of b.o.p. flows to compile stocks.2) Based on the common definition of own funds at book value.3) To the extent that they can be identif ied, considering the practical diff iculties existing at present, as addressed in chapter 2 of thisreport.4) For both reinvested earnings and FDI equity stocks.5) MS may focus on a reduced number of companies (the biggest ones and/or holding companies) to perform the distinction betweenordinary and extraordinary gains and losses.6) For all elements of FDI statistics (namely equity capital, reinvested earnings and other capital).

Table 23 Matrix of conclusions: priorit isation and timing for implementation of theTF-FDI recommendations

specific timing) the deadlines corresponding toeach slot, since this was considered out of itsmandate.

54. By combining both dimensions (i.e.importance and timeframe), the most significantrecommenda-tions of the TF-FDI have beenintegrated in a matrix for illustrative purposes.

15cECB


EXECUTIVE SUMMARY

1. ISSUES FOR FOLLOW-UP WORK

55. Due to its limited time horizon, the TF-FDIcould not expand the topics defined in itsmandate with other subjects identified in thecourse of its investigations. Additionally, theTF-FDI did not hold discussions on morestrategic issues. Therefore, it is proposed thatsome work could follow the delivery of thisreport in the following areas:

– Elaborate an implementation calendar withspecific deadlines to put therecommendations of the TF-FDI in practice.

– Monitor on a regular basis theimplementation status of the TF-FDIrecommendations as well as other mattersrelated to FDI (e.g. exchange of experiencesand information on FDI) through, forinstance, the regular meetings of theworking groups and/or ad-hoc workshops.Among the different issues to be consideredin the future, the TF-FDI recommends thatthe possibility to collect separate statisticsfor SPEs continue being assessed in thefuture.7

– Develop a twofold monitoring task, whichshould focus on: (i) the definition of the newinternational accounting standards; and (ii)the update of the IMF Balance of PaymentsManual. This monitoring task should aim atpromoting further convergence betweenstatistical and accounting standards, whilekeeping in mind that FDI statistics shouldalways be able to serve analytical needsfrom the macroeconomic viewpoint.8

– Explore practical ways to put the proposalsto compile statistics based on the UBO inpractice.

– Elaborate the list of user requirements forthe European Business Register projectcurrently being developed by the Eurostat’sBusiness Statistics Directorate. Such acontribution should ensure that the finalproduct will have the necessary features forthe compilation of FDI statistics.


8 In particular, the TF-FDI discussed two alternatives to try toapproximate statistical rules to accounting standards: (i) changethe 10% rule defining all FDI relationships to a 20% criterion;(ii) consider only indirect FDI relationships over 50% (i.e.restrict the coverage of indirect relationships to cases ofmajority control). The TF-FDI tentatively expressed apreference for the second option, which is already a practicalsimplification addressed in this report.


SUMMARY OF THE MANDATE

1. The Task Force on Foreign DirectInvestment (TF-FDI) was set up by the ECBWorking Group on Balance of Payments andExternal Reserves Statistics (WG-BP&ER),jointly with the Eurostat Balance of PaymentsWorking Group (BoP WG), to investigate onthe matters included in the mandate, focusingparticularly on the practical and consistentimplementation of the various principles/definitions involved.

2. The main objectives of the TF-FDI were toidentify “best practices” with a view tominimising divergence in the treatments appliedby euro area/European Union (EU) MemberStates. In effect, any such inconsistencies may,not only have a direct impact on the calculationof the euro area/EU aggregates, but alsoendanger the homogeneity of the final aggregateresults. In this respect, particular attention wasto be paid to the accuracy of the geographicalallocation of the relevant transactions andpositions.

3. The mandate of the TF-FDI was articulatedaround the five following items:

(i) Valuation of stocks: the TF-FDI was toinvestigate the practicality of theconceptual agreements reached within theWG-BP&ER and the STC in this area1,i.e. the problems associated with theprovision of additional breakdownsrequired for listed and non-listedcompanies for the production ofmemorandum items. More specifically,the TF-FDI should investigate (a) thefeasibility of the separate provision ofstock data on listed and non-listed FDIcompanies and (b) whether stocks onlisted (resident and non-resident) FDIcompanies may be valued on the basis ofboth market values and book values (thelatter using the common definition of“own funds at book value” approved bythe WG-BP&ER and the STC)2.

(ii) Other capital: the TF-FDI was mandatedto investigate and put forward practicalsolutions for the problems linked to:

a) the application of the “directionalprinciple”;

b) the identification and impact of tradecredits, financial leasing, debtsecurities subscribed by associatedcompanies;

c) the treatment and classification oftransactions/positions on “othercapital” when Monetary and FinancialInstitutions (MFI) are involved.

(iii) Reinvested earnings: the TF-FDI was toreview the practical aspects of thecompilation methods for reinvestedearnings (declaration by respondents,calculation by the compilers, interimestimations etc.), with the participation ofEurostat Unit B1 (National Accounts).

(iv) Identification and treatment of SpecialPurpose Entities (SPEs): the TF-FDIwould investigate practical problemsassociated with the identification andtreatment of transactions/positions of SPEs,mainly “other financial intermediaries” or“financial auxiliaries”, in the context ofdirect investment relationships.

(v) Treatment of indirect FDI relationshipsand allocation of FDI inward stocks bycountry of “ultimate beneficial owner”(UBO): the TF-FDI would investigate onthe practical application of the “fullyconsolidated system” in Member States.Possible solutions to the problem ofobtaining information on group structuresshould be examined with reference also tothe costs that they would entail. The

1 Reference documents “ST/WG/BP/FDIIMPLE.DOC”, dated29 October 2001 (WG-BP&ER) and “ST/STC/BP/FDIREPORT.DOC”, dated 20 November 2001 (STC).

2 According to the above-mentioned reference document“ST/WG/BP/FDIIMPLE.DOC”

I N T RODUCT I ON

17c ECB


INTRODUCTION

possibility and implications of classifyinginward FDI stocks by the country of UBOwould be analysed. The results and workongoing in other CMFB (Steering Groupon Multinationals) or Eurostat groups(FATS, Business register) would be takeninto account. Eurostat Unit D1 (Businessregister) would be invited to participate.

4. At their respective meetings in March 2003,the ECB WG-BP&ER and the Eurostat BoPWGgave further guidance to the TF-FDI. The WG-BP&ER stressed that “departure frominternational statistical standards should onlybe proposed by the TF in case a substantialcritical mass of Member States were in favourand for sound practical reasons, mainlyaddressed to avoid asymmetries in the euroarea/EU aggregates”. Both WGs stated that theCurrent Operating Performance Concept(COPC) should be the reference concept for thecompilation of reinvested earnings. It was alsostressed that for the practical application of thisconcept, some simplification could be needed inorder to avoid excessive costs. Additionally,the WG-BP&ER encouraged the TF-FDI tofurther examine the practicalities in compilingaggregates (in particular as regards“extraordinary profits” and any possibledistinction between financial and non-financialcorporations).

5. The focus of the work of the TF-FDI, as perits mandate, was on the identification of bestpractices with a view to minimisingasymmetries by the implementation of practicaland compatible solutions over a short- tomedium-term horizon.

STRUCTURE OF THE REPORT

6. In February, the TF-FDI presented a firstreport on the first three items of the mandate.The structure of the report was based on theorder in which the various items under studywere listed in the mandate. The items coveredby the report were valuation of equity stocks,reinvested earnings, and other capital.

7. Starting during the first phase, one issueappeared repeatedly in the discussions, namelythe potential use of (non-)consolidated accountsfor the compilation of FDI statistics. Since theissue on the treatment of indirect FDIrelationships has proved crucial for the finalconclusions of the TF-FDI, a broader analysishas been made of these aspects. Also thestructure of the report has been adapted to thissituation and the first two chapters deal with thetheoretical and practical aspects of the FullyConsolidated System and indirect FDIrelationships.

8. The following three chapters deal with thefirst items of the mandate, namely valuation ofequity stocks, reinvested earnings and othercapital. The chapter on reinvested earnings hasbeen developed in line with the clarificationsmade by the two Working Groups. The chapteron other capital has been supplemented with theresults of a study on the importance of the sub-components following the request of theBoPWG. Chapters VI and VII deal with the twofinal items of the mandate (exception made ofthe coverage of indirect FDI relationships,which is tackled in the first two chapters, asmentioned in the previous paragraph), namelythe identification and treatment of SpecialPurpose entities and the allocation of stocks bycountry of ultimate beneficiary owner. A finalchapter provides a summary of the mainconclusions and recommendations of the TF-FDI.


INTRODUCTION

9. This chapter is an attempt to clarifysomewhat the conceptual backgroundestablished by international standards. Morespecifically, this chapter fosters the adoption ofa unique methodology applicable to somespecific cases for which international standardsmay leave some room for interpretation.3

10. This chapter has an introductory naturestemming from the fact that it tackles, purely onconceptual grounds, general aspects which arerelevant to the interpretation of other parts ofthe report. In particular, the recommendationsaddressed by this chapter should be consideredas to how the conclusions of, for instance,chapters 2 (Practical solutions for the coverageof indirect FDI relationships), 3 (Valuation ofFDI equity stocks) and 4 (Reinvested earnings)should be applied.

11. To be more specific, internationalstandards prescribe that direct investmentstatistics should cover all directly and indirectlyowned subsidiaries, associates and branches.The incorporation of indirectly related FDIaffiliates to the value of the total directinvestment should be done through theappropriate process of consolidation.

12. This chapter aims at clarifying further howto interpret standards with regard to thecoverage of indirect FDI relationships. It isimportant to stress that it is restricted to theconceptual analysis of some aspects concerningthe methodology applicable to the compilationof FDI statistics. The following chapter (2) willstudy in detail any practical problems for theapplication of such a methodology, currentpractices as well as the difficulties linked to thecompilation of the European aggregates and thepossible use of consolidated accounts for thecompilation of FDI statistics.

13. The identification of FDI relations hasbeen traditionally based on the methodologycontained in the OECD Benchmark Definitionof Foreign Direct Investment (the Benchmark)

1 CONC E P TUA L I S S U E S R E L AT ED TO TH E F U L LYCON SO L I DAT ED S Y S T EM AND TH ECOV ER AG E O F I ND I R E C T F D I R E L AT I ON SH I P S

and in the IMF Balance of Payments Manual(BPM5). As part of such methodology, the so-called Fully Consolidated System (FCS) ismeant to identify those enterprises in which thedirect investor has directly or indirectly a directinvestment interest. Thus, FDI statistics shouldcover transactions and positions between directinvestors and all FDI enterprises which are partof the FCS.

14. The traditional presentation of the FCS isusually illustrated by the following chart:

3 It should be borne in mind that the ongoing process of updatingthe BPM5 could trigger significant revisions in internationalstandards in the forthcoming years. Such revisions could helpovercome some of the most significant practical difficultiescurrently faced by compilers and identified in this report.

15. The FCS basically illustrates whichenterprises below company N in the chainshould be considered as subsidiaries, associatesor branches and whether or not they should becovered by FDI statistics. According to thediagram and the FCS rules, companies A, B, C,D, E, F, K and L should be covered by FDIstatistics.

16. While, as a general reference, the FCShelps to define which companies in the exampleshould be considered in FDI statistics (leavingaside how difficult collecting such a detailedpicture of the multinational groups’ structure

N

60% 10% 30% 9% 70%

Company A Company D Company F Company H Company K

55% 60% 25% 100% 100%

Company B Company E Company G Company J Company L

12%

Company C

Chart 1 Ful ly consol idate system

19c ECB


1 Conceptual issuesrelated to the fullyconsolidated system

and the coverageof indirect FDIrelationships

might be), there are some more specificquestions that may not be so clearly answeredby international standards and the FCS in itstraditional presentation. For instance, let usconsider the following example:

Enterprise 1 (France)

Assets Liabilities

€100 €100(shares of Ent. 2) (Equity capital)

Enterprise 2 (Ireland)

Assets Liabilities


Enterprise 3 (United States)

Assets Liabilities

€100 €100(Equipment) (Equity capital)

17. The (unconsolidated) balance sheet ofthese enterprises could initially be as follows:

18. The methodology addressed byinternational standards (BPM5 and theBenchmark) may not suffice to determine whichtransactions/positions between 1 and 3 shouldbe recorded under FDI and how. Some typicalexamples of transactions that may generatedoubts are (i) equity transactions below 10%between companies without direct links ofownership (1 and 3); (ii) whether reinvestedearnings generated by 3 should be attributed to1; (iii) whether the valuation of the equitycapital stocks based on the “own funds at bookvalue” of 2 should include retained earnings /reserves generated by 3; etc.

19. In the next sections, these and otherexamples will be analysed case by case. Sectionone deals with stocks and section two withtransactions between indirectly relatedcompanies in cases such as the one presented inExample 1. Section three considers a different

Example 1 Indirect relationships grandmother – granddoughter: equity capital (stocks)


case, i.e. that of “fellow”/“sister” companies4.Finally, section four concludes by puttingforward some general conclusions andrecommendations.

COMPANIES WITH INDIRECT LINKS OFOWNERSHIP

FDI STOCKS: EQUITY CAPITAL AND OTHERCAPITAL

EQUITY CAPITAL STOCKS20. To start with, let us focus on Example 1 aspreviously described: the first question couldbe whether or not (and how) enterprise 1 shouldincorporate to the value of its equity capitalstocks of outward FDI part or all of the value ofenterprise 3. To simplify even further the casesanalysed we always focus on relationshipsresident/non-resident and implying 100% ofownership.

21. As regards the valuation of FDI equitystocks based on the common definition of OwnFunds at Book Value (OFBV)5, the problemcould be more clearly identified by consideringseparately: (i) nominal capital; and (ii)undistributed reserves (i.e. reinvestedearnings), including current year’s profits/losses carried forward.6

22. Let us begin with the first component, i.e.nominal (paid-up) equity capital. If the capitalof enterprise 3 was added to the nominal capitalof enterprise 2, the value of the outward FDIstocks of enterprise 1 would resultoverestimated.

23. In our example, a company located in FR(Ent.1) invests EUR 100 in a US company (Ent.3), through a holding company (2) located in IE.To simplify, all ownership relations entail100% ownership.

24. If the outward FDI equity stock of FRincluded the share capital of all the subsequentlinks in the chain, it would amount to: 100 (IE)+ 100 (US) = 200. However, the outward

investment of FR would just be worth 100 (andwould only be valued that much by themarkets), which is the money that Ent. 1 hasactually put in circulation.

25. With a view to further illustrating this case,we can also consider Example 2, which is basedon a real group of companies. The names of thecompanies involved have been hidden so asto overcome confidentiality problems (seeExample 2).

26. For the second calculation, the nominalcapital in the balance sheet (liabilities) of thecompanies below XXX Enterprises has not beenconsolidated with the value of thoseinvestments in the balance sheet (assets) of XXXEnterprises, i.e. the funds that XXX Enterprisestransfers to its subsidiaries. Obviously thedifference between both approaches is rathersubstantial. The conclusion would be that onlythe nominal capital of the directly-owned FDIcompany should be taken into account.

27. Let us consider now the second componentof equity capital, namely non-distributedreserves and profits (losses) in the current year,following the common definition of own fundsat book value (OFBV) approved by the STC.

28. Coming back to the original “simplistic”example (as previously shown in Example 1),let us consider now that Enterprise 3 makessome profits, which are not distributed to itsshareholders (see Chart 3).

4 Both terms, i.e. “fellow” companies and “sister” companies areindistinctly used throughout the report to refer to the same kind ofcompanies, namely those pertaining to the same multinationalgroup but with neither direct nor indirect control over oneanother.

5 The STC decided that the valuation criteria for the official euroarea series should be market (stock-exchange) prices for listedcompanies and book values (based on the common definition ofOFBV) for non-listed companies. Nevertheless, equity stocksfollowing the book valuation based on OFBV will be requestedfor all types of companies.

6 The specific treatment applicable to more detailed componentslike premiums, non-disbursed capital, capital grants, etc. wasdeveloped by the WG-BP&ER.

21c ECB




Example 2 Real case of indirect relations grandmother – granddoughter

XXX ENTERPRISES SA (2,837.5)

XXX SA SWITZERLAND

SWITZERLAND FRANCE

99.99% 99.97%

99.90%

95.10%

93%

55.50%

99.99%

99.99%

100%

100%

100%

Y10(18.1)

Y9

Y8(433.8)

Y7(11.6)

Y6

Y5(900.0)

Y4(-1.2)

Y3(69.5)

Y2(4.9)

Y1(1.9)

From the point of view of France, inward FDIstocks would amount to:

Inward FDI (only “first shot”):

Switzerland

99.99% × 2, 837.5 =

EUR 2, 837.5 million

Inward FDI stock(including FDI indirect relationships only whereaccounting data available):

Switzerland

99.99% × 2,837.5 + 99.99% × 100 × 1.9

+ 99.99% × 100 × 4.9 + 99.99% × 100 × 69.5

+ 99.99% × 99.99% × (-1.2)

+ 99.99% × 99.99% × 900

+ 99.99% × 99.90% × 11.6

+ 99.99% × 95.1% × 433.8

+ 99.99% × 55.50% × 18.1

= EUR 4, 246.3 million


29. Now the (unconsolidated) balance sheet ofthe three enterprises would look as follows:

Enterprise 1 (France)

Assets Liabilities


Enterprise 2 (Ireland)

Assets Liabilities


Enterprise 3 (United States)

Assets Liabilities

€100 (Equipment) €100 (Equity capital)€20 (Cash) €20 (Reserves)

30. Following the guidance provided byinternational standards, reinvested earningsgenerated by indirectly owned enterprisesshould also be incorporated to the totalreinvested earnings corresponding to theoutward FDI investments of enterprise 1.

31. Therefore, in order to be compliant withthese guidelines, these undistributed profitsshould also be considered within the total valueof the outward FDI equity capital stocks of FR,which should amount to EUR 100 (equitycapital of ent. 2) + EUR 20 (undistributedreserves generated by ent. 3) = EUR 120. Fromthe point of view of the enterprise located in IE,all reinvested earnings recorded as outward FDIshould also be recorded as inward FDI, with anil effect, thus, on a net basis.

Chart 3 Reinvested earnings (stocks)

23c ECB




32. Summing up the main conclusions of thissection:

The TF-FDI is of the opinion that, onconceptual grounds, for the valuation basedon book values of FDI equity stocks, thefollowing should be considered:

Concerning the nominal capital:

(i) only that of the directly owned FDIenterprises should be included in thevaluation of FDI equity stocks.

Concerning the (non-distributed) reserves, itis important to distinguish between directinvestment in the reporting economy anddirect investment abroad:

(ii) FDI abroad (outward): in addition to thereserves of the directly owned foreign FDIcompanies, reserves generated by theaffiliates of the foreign FDI companies7

should be incorporated to the total valueof the FDI equity capital in proportion tothe % of ownership across the subsequentlevels of the ownership chain.

(iii) FDI in the reporting economy (inward):the FDI company should attribute to theforeign investor (i.e. the direct owner), inaddition to its own reserves, all reservesgenerated by its directly or indirectlyowned direct investment enterprises8 inproportion to the % of ownership.

33. Therefore, as a general principle, book-value-based FDI equity stocks should cover the OFBVof the directly owned direct investment enterpriseplus the (non-distributed) reserves generated bythe (domestic and foreign) affiliates of thedirectly owned direct investment enterpriseaccording to the rules of the FCS. Obviously,concerning valuation principles for equity capitalstocks, all references to the application of theOFBV definition and whether or not reinvestedearnings generated by indirectly owned FDIenterprises should be incorporated to the stocksare only applicable to the valuation based on bookvalues. Market values based on stock exchangeprices should already incorporate all relevantinformation and, thus, do not require any furtheradjustment.

7 Both resident and non-resident.8 Both resident and non-resident.

Chart 4 Other capital (stocks)


OTHER CAPITAL STOCKS9

34. Coming back to our basic example, let usconsider a loan granted by enterprise 3 to its(indirectly related) mother company, i.e. toenterprise 1 (see Chart 4).

35. Leaving aside some practical problems suchas how to identify indirect FDI relationsbetween lenders and borrowers, the inclusion ofsuch a loan under FDI other capital seemsuncontroversial on purely conceptual grounds.However, there might be a doubt on whethersuch a loan should be recorded by FR underFDI in the reporting economy following thedirection of the cash flows or rather under FDIabroad as a disinvestment, i.e. following thedirection of the FDI relationship.

The TF-FDI recommends that such a loanshould be recorded under FDI abroad/Othercapital as a disinvestment by the country ofenterprise 1, i.e. the directional principleshould prevail, even if such an FDI relationshipis merely indirect.10

FDI FLOWS: EQUITY CAPITAL, REINVESTEDEARNINGS AND OTHER CAPITAL

EQUITY CAPITAL TRANSACTIONS36. The main question concerning this item iswhether or not transactions between indirectlyrelated companies below 10% of ownershipshould be recorded under FDI. In the examplewe have been analysing so far, let us considerthat enterprise 3 acquires 5% of the equitycapital of its (indirectly-linked) mothercompany located in FR (see Chart 2).

9 The loans referred to (or any transactions other than permanentdebt) exclude those involving financial intermediaries.

10 Enterprise 3 should record the loan under FDI in the recordingeconomy/Other capital as a disinvestment. Obviously, enterprise2 should not record anything. Any loan in the opposite direction,i.e. granted by enterprise 1 to enterprise 3, must be recorded byFR under FDI abroad/Other capital (and by USA under FDI in thereporting economy/Other capital).

Chart 2 Equity capital below 10% (f lows)

25c ECB




37. Should this transaction be recorded underFDI/Equity? If that is the case, should it beconsidered as inward or outward FDI?

The TF-FDI favours the recording of such atransaction by FR under FDI abroad/Equity/Liabilities to affiliated enterprises as adisinvestment, for the same reasons previouslyexplained, i.e. the directional principle shouldprevail. 11

REINVESTED EARNINGS38. A case which is equally applicable here hasbeen already analysed under equity capitalstocks in paragraphs 28 to 31. The conclusionsconcerning b.o.p. flows (in this case, recordedunder reinvested earnings with a counter entryin the income statement) would be equivalent tothe conclusions reached concerning equitystocks.

Consequently, the TF-FDI is of the opinionthat:

(i) concerning direct investment abroad,reinvested earnings generated by bothdirectly and indirectly owned enterprises12

should be considered in proportion to the% of ownership down the chain;

(ii) concerning direct investment in thereporting economy, the attribution ofreinvested earnings to the foreign mothercompany (i.e. the direct owner) inproportion to its ownership share shouldencompass the sum of all reinvestedearnings generated by the (directly owned)domestic FDI enterprise plus allreinvested earnings generated by the(directly or indirectly owned) affiliates ofthe domestic FDI enterprise (also inproportion to the % of ownership)13.

OTHER CAPITAL TRANSACTIONS39. The example for other capital transactionscould be the same as that already analysed forOther capital stocks. In short, the conclusionsare basically consistent with those already

provided for the consideration of stocks,namely:

It is recommended that all loans betweenindirectly related companies should berecorded under FDI other capital. For theconsideration of those transactions as eitherinward or outward FDI, the directionalprinciple should prevail, i.e. the (indirect)investor should record all transactions in FDIabroad/other capital, while the directinvestment enterprise should record alltransactions under FDI in the reportingeconomy/other capital.

THE CASE OF “FELLOW”/“SISTER” COMPANIES

40. Some of the recommendations put forwardso far could slightly vary in the case ofcompanies whose role in the group’s structureis not so clearly defined. For instance, whenneither company is the mother of the group norare they clearly at the end of the chain, it mightbe difficult to determine how the directionalprinciple should be applied.

41. “Sister companies” in the context of FDIstatistics could be defined as affiliatespertaining to the same multinational groupwhich do not have a participation/interest of10% or more in each other. Let us considerExample 2 as the basis for discussion.Enterprises 2 and 3 would be what wecall “sister” companies in this section (seeExample 3).

11 For the sake of consistency, the country of enterprise 3 (in theexample, US), should follow the same recording rules, i.e. thetransaction should be recorded under Direct investment in thereporting economy/equity/claims to direct investors as adisinvestment. A similar transaction in the opposite direction, i.e.an investment of 1 in the equity capital of 3 below 10% should berecorded by FR under FDI abroad/Equity/Claims on affiliatedenterprises.

12 In the case of companies with indirect links of ownership, bothforeign and domestic direct investment enterprises should becomprised (provided that the direct link of ownership ismaintained with a foreign direct investment enterprise).

13 Irrespective of whether they are resident or non-resident.


42. Then different transactions could take placebetween 2 and 3. Let us consider the followingthree independent cases, corresponding to thethree FDI items. All of them are supposedto happen taking as starting point the situationin T.

Example 3 Fel low/sister companies: equity capital (stocks)

27c ECB




Chart 5 Transactions between fel low/sister companies


43. According to the IMF Text Book, paragraph529, “When a direct investment enterpriseinvests in an enterprise related to its directinvestor, this investment is recorded, by theeconomy providing the investment, as residentdirect investment-abroad and by the enterprisereceiving the investment, as direct investment-reporting economy”.

44. Therefore, concerning the case of fellow/sister companies, the followingrecommendations are proposed by the TF-FDIon conceptual grounds:

• Equity capital transactions/positionsbetween sister companies not exceeding10% of ownership should be recordedunder FDI (by 2 and 3; not by 1), either asinward or outward FDI depending on thedirection of the investment.

• Reinvested earnings generated by sistercompanies with no direct equity links shouldnot be recorded under FDI by thosecompanies, i.e. 2 should not record anyreinvested earnings generated by 3.14 If the

starting point was the situation after theacquisition of 5% of 3 by 2, 2 should record5% of the reinvested earnings generated by3 (and 3 should attribute 5% of itsreinvested earnings to 2).

• Loans granted to/borrowed from sistercompanies should be recorded under FDIother capital (by 2 and 3; not by 1), eitheras inward or outward FDI depending on thedirection of the loan.

CONCLUSIONS45. This section summarises the agreementsreached by the TF-FDI towards a commonconceptual understanding of internationalguidelines. In short, this chapter is articulatedaround the distinction between two types ofcompanies with indirect links: (i) those forwhich one of the concerned companies exerts(indirect) control over the other; and (ii) thosewith neither direct nor indirect control over oneanother (i.e. the so-called “fellow”/“sister”companies).

14 Obviously, the mother company (1) should record the reinvestedearnings generated by both companies 2 and 3.

Chart 5 cont’d

29c ECB




46. For companies with an indirect link ofownership when one of them exerts indirectcontrol over the other, the conceptualagreements of the TF-FDI could be summarisedas follows:

Stocks

(i) Equity capital: concerning the nominalcapital, only that of the directly ownedFDI enterprises should be included in thevaluation of FDI equity stocks based onbook values. Concerning non-distributed)reserves, the value of equity stocksshould include, in addition to thosecorresponding to the directly owned FDIcompanies, (i) for FDI abroad, reservesgenerated by indirectly owned15 directinvestment enterprises in proportion tothe % of ownership; (ii) for FDI in thereporting economy, reserves generated bythe domestic FDI company’s affiliates16

in proportion to the % of ownership.Equity capital stocks should be classifiedas inward or outward according to thedirectional principle (see below)

(ii) Other capital: loans between this type ofcompanies should be recorded underinward or outward FDI according to thedirectional principle (see below).

Flows

(i) Equity capital: transactions below 10%should be recorded under inward oroutward FDI equity capital according tothe directional principle (see below).

(ii) Reinvested earnings: In line with therecommendations provided for equitycapital stocks, the total reinvestedearnings should include, in addition to thereinvested earnings corresponding to thedirectly owned FDI companies, (i) forFDI abroad, reinvested earningsgenerated by indirectly owned foreigndirect investment enterprises inproportion to the % of ownership; (ii) for

FDI in the reporting economy, reinvestedearnings generated by the domestic FDIcompany’s affiliates17 in proportion to the% of ownership.

(iii) Other capital: loans between indirectlyrelated companies should be recordedunder FDI other capital. For theconsideration of those transactions aseither inward or outward FDI, thedirectional principle should prevail (seebelow).

Applicability of the directional principle

All stocks and flows between these typesof companies should be classified by theinvestor (i.e. the indirect owner) as FDIabroad and by the (indirectly owned)direct investment enterprise as FDI in thereporting economy, under the relevantFDI items.

47. For the second group of FDI companieswith indirect links, namely „sister“ companies,the recommendations of the TF-FDI would beas follows:18

Stocks

(i) Equity capital: equity stocks held by sistercompanies not exceeding 10% ofownership should be recorded under FDI.The direction of the investment shoulddetermine whether stocks should beclassified under inward or outward FDI.

15 In the case of companies with indirect links of ownership, bothforeign and domestic direct investment enterprises should becomprised, provided that the direct link of ownership ismaintained with a foreign direct investment enterprise.

16 Either directly or indirectly owned, irrespective of whether theyare resident or non-resident.

17 Either directly or indirectly owned, irrespective of whether theyare resident or non-resident.

18 We mostly refer hereafter to the recording by the sistercompanies involved. The entries that, where relevant, should berecorded by the parent company (e.g. for reinvested earnings)are not considered here.


(ii) Other capital: Loans granted to/borrowed from sister companies shouldbe recorded under FDI/other capital bythe lender/borrower, either as inward oroutward FDI depending on the directionof loan.

Flows

(i) Equity capital: transactions between sistercompanies not exceeding 10% ofownership should be recorded under FDI.The character of inward/outward FDIshould be determined by the direction ofthe investment (i.e. outward FDI for theshareholder and inward FDI for theissuer).

(ii) Reinvested earnings: Reinvested earningsgenerated by these companies should notbe recorded under FDI by the other(sister) company, as long as neithercompany indirectly exerts control over theother, nor direct ownership links existbetween both companies.

(iii) Other capital: Loans granted to/borrowedfrom sister companies should be recordedunder FDI other capital by both lender andborrower, either as inward or outwardFDI depending on the direction of the loan

Applicability of the directional principle

In the case of sister companies, flows andstocks should be classified as outward FDI bythe country which provides the investment orgrants the loan and as inward FDI for thecountry receiving the investment/loan.

31ECB


2 Practical aspectsrelated to the

coverage ofindirect FDI

relationships

INTRODUCTION

48. The previous chapter of this report hasestablished common rules as to howinternational standards should be interpretedconcerning the coverage of indirect FDIrelationships in the compilation of FDIstatistics. Once a common conceptualunderstanding has been already defined, thepresent chapter aims at investigating the mostimportant difficulties currently existing for theapplication of such rules in practice.

49. Besides this introduction, this chapter isstructured in five sections. The first section hasan introductory nature and analyses thedifferences between the accounting rulesapplied to elaborate consolidated balance sheets(accounting consolidation hereafter) andstatistical rules as contained in the so-calledfully consolidated system (FCS) plus relatedmethodology for the compilation of FDIstatistics (henceforth referred to as statisticalconsolidation). The second section describescurrent practices in member states for thecoverage of indirect relationships as well assome other connected features revealed by aquestionnaire circulated within the TF-FDI.

50. Section 3 explores the consequences for thecompilation of the European aggregates of thetreatment applied to indirect FDI relationshipsthrough an illustrative example. Section 4extracts some conclusions from the analysiscarried out throughout the chapter and proposessome practical simplifications to the rules forthe coverage of indirect FDI relationshipsprescribed by international standards. Thissection also includes some criteria that shouldbe applied for the geographical allocation ofFDI transactions and positions in whichindirect FDI relationships play a role. Finally,in connection with the coverage of indirect FDIrelationships and following the instructions ofthe mandate section 5 considers possiblesolutions to the problem of obtaininginformation on group structures and, inparticular, the possibility to develop a Europeandatabase/business register.

2 PRA C T I C A L A S P E C T S R E L AT ED TO TH ECOV ER AG E O F I ND I R E C T F D I R E L AT I ON SH I P S

STATISTICAL CONSOLIDATION VERSUSACCOUNTING CONSOLIDATION

INTRODUCTION51. The inclusion of indirect links of ownershipin FDI statistics is a major challenge due to thenumerous practical difficulties that compilersusually encounter in accessing to the relevantinformation.

52. Such practical difficulties are largelyacknowledged in the manuals containinginternational statistical standards, such as theBPM5 and the Benchmark. Therefore, thesemanuals normally adopt a rather practicalapproach for the application of the FCS and themethodology brought forward for thecompilation of FDI statistics and admit the useof consolidated accounts of the companiesinvolved in FDI deals as an acceptable way tocapture indirect FDI relationships.

53. The sole use of consolidated accounts in thecompilation of FDI statistics without any othersupplementary information does not allowgetting the results that the FCS prescribes. Forinstance, accounting statements normally lackvital information for the compilation of externalstatistics such as the geographical dimension orinformation on the economic sector of activityof the affiliates. Additionally, the rules todefine the consolidation perimeter in accountingnormally differ from those linked to the so-called fully consolidated system (FCS), whichare basically defined by the 10% rule and thenotion of “majority control” (so as to identifywhich companies should be covered in FDIstatistics).

54. In particular, the connection between bothapproaches (the statistical rules applicable tothe compilation of FDI statistics and theaccounting rules applied to the elaboration ofconsolidated accounts) may be illustrated bysplitting statistics into two dimensions: (i)input, or how the basic information is collectedby the b.o.p./i.i.p. compiler, and (ii) output, orthe statistic which is finally produced. The firstdimension relies to a great extent on the


information sources from which the informationis collected. Since FDI stocks are normallycompiled via direct contact with reportersthrough FDI surveys, it might be logical toassume that reporters use a single set of(accounting) rules to elaborate both theirbalance sheet and to fill in statistical reports.Such accounting rules are aimed at assessingthe micro situation of each individual company.On the other hand, the output dimension isnormally ruled by the needs of the users, whichnormally entail a macro economic approach.Such a framework requires the application of adifferent methodology concerning aspects suchas valuation criteria, time of recording, etc.

55. In conclusion, some confusion may comeout when the notion of “consolidation” is usedwith no further specification, i.e. it might referto either the output or the input dimension. Onthe input side, the concept of consolidation isnormally associated to the notion of accountingconsolidation, i.e. the rules applied by thecompanies to elaborate consolidated accounts.On the output side, the concept of consolidationnormally refers to the FCS (as defined in theBPM5 and the Benchmark), which is meant toestablish the rules governing the coverage of allrelationships to be considered as directinvestment. Both notions are not totallycoincident even if they attach some similarities.

56. This section explores the differencesbetween the notions of statistical consolidationand accounting consolidation. Morespecifically, it tries to identify the maindifferences between the statistical guidelinescontained in the FCS plus related FDImethodology and the rules governing theelaboration of the companies’ consolidatedaccounts.

57. As it has been recognised before, the directuse of consolidated accounts for thecompilation of FDI statistics is not possiblewithout some additional information.Therefore, the comparisons in this chapter aremainly intended for illustrative purposes. Theidentification of the differences between

statistical and accounting rules concerning thescope for consolidation is meant to highlight thedifficulties that compilers may encounter at thetime of collecting information and providinginstructions to respondents. Additionally, it isalso intended to identify what kind ofsupplementary information would be necessary.

ACCOUNTING CONSOLIDATION58. The request for consolidated accounts ismeant to evaluate the true situation of acompany with a participating interest in someother affiliates (subsidiaries and branches). Theprocess of consolidation basically consists ofattributing to the consolidated enterprise allassets and liabilities of its subsidiaries andbranches, thus cancelling out all reciprocalassets and liabilities. In the profit and lossaccount, the consolidated enterprise is alsoattributed all credits and debits of theconsolidated subsidiaries and branches’ profitand loss statements.

59. In the balance sheet the most significantresults are:

– all reciprocal participations in equity capitalbetween consolidated enterprises (recordedin the assets side of their balance sheet)disappear after the consolidation process;

– the equity capital + reserves19 of theconsolidated enterprises (recorded in theliabilities side of their balance sheet) alsodisappear in the consolidation process;

– all loans and deposits between allenterprises involved in the process ofconsolidation also cancel out in theconsolidated balance sheet;

– finally, a minority ownership item is createdin the liabilities side of the consolidatedbalance sheet (accounting for the non-consolidated shareholders’ interest).Additionally, minority (non-consolidated)

19 The profits/losses generated after the consolidated enterprisesbecome members of the group do no longer disappear in theconsolidation process.

33ECB




relationships

interests in the assets side of the companiesremain in the assets side of the consolidatedbalance sheet.

60. All other assets and liabilities of theconsolidated enterprises are fully attributed tothe consolidated company, even if the links ofownership between the companies do not reach100%. In addition, the consolidation perimeteris limited to subsidiaries controlled or owned at50% as a minimum (all branches areconsolidated since, by definition, they are 100%owned by their mother companies).

61. One other aspect that may be worthmentioning is that the consolidation processmay be applied at different levels of the chain ofsubsidiaries. Therefore, the same assets/liabilities may be accounted for by enterprisespertaining to the same group in their respectiveconsolidated accounts and, from a macro-economic point of view, gross figures mayresult magnified.

IMPACT OF THE NEW INTERNATIONALACCOUNTING STANDARDS (IAS) IN THEACCOUNTING CONSOLIDATION RULES62. The introduction new IAS may trigger somechanges to the present practices:

– A major impact is that the proportionalconsolidation method20 will be restricted tothe cases of joint ventures.

– For subsidiaries, the 100%-consolidationmethod21 will be the overall rule.

– For associates, no consolidation rules areproposed. The consolidated balance sheetjust registers the value of theparticipation(s) (“equity method”) andfinancial assets/liabilities directly transactedwith those associates (e.g. loans), as withany other counterpart. The results are equalto those on non-consolidated balance sheets.

– The ownership threshold to defineassociated enterprises is 20%, i.e. differing

from the 10% rule followed in FDIstatistics.

– A set of supplementary tables will be addedto the balance sheet containing informationrelated to the subsidiaries and associates(limited information).

63. In the consolidated data there is still nodistinction required between domestic andforeign consolidated enterprises.

STATISTICAL CONSOLIDATION: THE FULLYCONSOLIDATED SYSTEM (FCS)64. The notion of consolidation in FDI statisticsis meant to produce statistics with full coverageof all direct and indirect FDI relationshipsthrough a detailed specification of the affiliatesthat should be considered as subsidiaries,associates and branches, respectively. Inparticular, the FCS states that FDI statisticsshould cover all enterprises in which the directinvestor has directly or indirectly a directinvestment interest.

65. However, as we have seen in the firstsection of this chapter, the practical applicationof the rules established by the FCS is not aneasy task. In particular, it requires a perfectknowledge of the ownership structure of thegroup, including the % of participation and thelocation (residence) of each entity as well as theavailability of detailed information on assetsand liabilities of each company pertaining to thegroup. In particular, concerning the latter point,intra-group assets and liabilities should not beconsolidated so as to allow the production ofgross figures, which are required in FDIstatistics for analytical purposes.

20 Proportional consolidation means that assets and liabilities of theconsolidated subsidiaries are incorporated to the consolidatedbalance sheet of the parent company only in proportion to its %of ownership.

21 The 100%-consolidation method means that 100% of theconsolidated subsidiaries’ assets and liabilities are incorporatedto the consolidated balance sheet of the parent company,irrespective of the % of ownership.


66. The application of the FCS implies thatreinvested earnings generated by directly orindirectly controlled affiliated enterprises areattributed to the parent company in proportionto its percentage of ownership, calculatedthroughout the ownership chain following therules of the FCS.

DIFFERENCES BETWEEN ACCOUNTING ANDSTATISTICAL RULES FOR CONSOLIDATION67. Following what has been described so far,some major differences between the twoconcepts can be established. This comparison isfurther illustrated in Table 1. The comparison ofthese two sets of rules is just meant forillustrative purposes and refers to the strictbasic concepts used in both domains. It does notcover those cases in which compilers requestsupplementary information from respondents tocomplement information extracted from theiraccounting statements since, in such cases, alldifferences could potentially be overcome.

68. On the basis of the above analysis and thedifferences between both approaches, itbecomes evident that the direct use ofconsolidated accounts (balance sheets) tocompile FDI stocks without any othersupplementary information is not possible. Dueto the different criteria used to determine theconsolidated perimeter and the differentconsolidation approaches (100% orproportional consolidation, respectively), itcould produce either an overestimation or anunderestimation of the resulting figures.

Table 1 Dif ferences between accounting and stat ist ical rules for consol idation 1)

Accounting rules Statistical rules

Consolidation perimeter � 50 % ownership � 10 % ownershipAttribution of consolidated elements 100% 2) In proportion to % of ownershipBreakdown by item no yesGeographical breakdown no yesBreakdown by counterpart no yesMeasurement basis net gross

1) This exercise is based on the accounting rules in place in a majority of countries. Such accounting rules may be different in some othercountries such as, for instance, the UK and IE.2) Exception made of minority interests.

69. The geographical dimension implicit in theconcept of FDI and the FCS constitutes anadditional problem. The elaboration ofconsolidated accounts does not distinguishbetween domestic and foreign affiliates.Furthermore, the need to compile aggregatestatistics for a group of countries integrated in amonetary union implies an additional difficulty,since international standards are mostlydesigned to provide methodological consistencyfrom a purely national viewpoint (this point isfurther developed in the next section).

70. Therefore, although the use of consolidatedaccounts (or balance sheets) is accepted byinternational standards and offers someadvantages as it offers information on morethan one level of ownership, this procedurealone does not provide the necessaryinformation to comply with statisticalrequirements. The use of consolidated accountsto produce FDI statistics always requires somesupplementary information (e.g. on grossrelationships, geographical breakdown of thecounterparts, economic activity, etc), which isnecessary to perform some adjustments.

71. Nevertheless, despite the evidentdeficiencies of directly using consolidatedaccounts in compiling FDI statistics, they mightbe a good proxy for compiling FDI equitystocks and reinvested earnings when thedomestic respondent is in charge of compilingconsolidated accounts for the group. In thosecases, links of ownership above 50% are wellcovered. For compiling other FDI items, the use

35ECB




relationships

of consolidated accounts as such is moreproblematic. In any case, the obligation toelaborate consolidated accounts demands a lotof information on intra-group assets andliabilities, which may be readily available torespondents and may become suitable forstatistical purposes. Special care should betaken to avoid double recording.

FULL STATISTICAL CONSOLIDATION (FCS) VERSUSTHE COVERAGE OF ONLY DIRECT LINKS OFOWNERSHIP72. The complexity of a full compliance with therules established by the FCS has beenextensively mentioned throughout the report.Many countries cannot go beyond the first levelof the ownership chain in their FDI statistics(see results of the questionnaire on currentpractices in the next section). By restrictingthemselves to only cover direct (“first-shot”)links of FDI, those countries also try tominimise the reporting burden for respondents.

73. Such an approach implies, concerningequity stocks (above 10%) and reinvestedearnings, the recording of: (i) as regards directinvestment in the reporting economy, onlystocks/flows vis-à-vis the foreign investor(s)that directly owns shares of the domestic DIenterprise; and (ii) concerning direct investmentabroad, only stocks/flows vis-à-vis the foreignDI enterprise(s) directly owned by the domesticinvestor. Other capital stocks/flows and equitycapital flows/stocks below 10% are attributed tothe direct counterpart.

74. While this procedure is efficient, lesscomplicated and less costly than a completeapplication of the FCS, it does not fully meetinternational standards.

COMPARISON BETWEEN (I) COVERAGE OF ONLYDIRECT LINKS OF OWNERSHIP; (II) THE“STATISTICAL CONSOLIDATION” (FCS) AND(III) THE “ACCOUNTING CONSOLIDATION”APPROACHES75. As we have seen in the previous paragraph,the comparison between the coverage of onlydirect ownership links and the FCS approach

reveals that both have advantages anddisadvantages. This section ends with acomparative analysis of the three approachesconsidered so far. This comparison is purelymeant for illustrative purposes, since, aspreviously said, while the concepts underlyingthe direct-ownership and the FCS/statisticalconsolidation approaches are meant to producestatistics, the rules governing the elaboration ofconsolidated accounts and its final product, i.e.the consolidated balance sheet, cannot be usedto that purpose without making use ofsupplementary information.

76. The most important difference betweenthe three approaches is their respectivecoverage, which is illustrated in Chart 6.A more exhaustive description of the pro’sand con’s of the three approaches is shown inTable 2. Once more, it might be convenientto underline that this comparison is madefor illustrative purposes and that it does notcontemplate mixed systems such as, forinstance, the use of accounting datasupplemented by additional informationprovided by respondents.


Cha r t 6 D i a g r am show i n g t h e d i f f e re n c e i n t h e s c ope o f t h e t h re e app ro a ch e s 1 )

B2

100% 30%

COUNTRY A COUNTRY C

C2 C4

100% C3

B3

30%

30%

40%

C4

100%C1 C3

FULLYCONSOLIDATEDSYSTEM

CONSOLIDATEDACCOUNTS

ONLY DIRECTLINKS OF OWNERSHIP

40%C1

B330%

B1

100%

100%

C2100% B1

B2100%

100%

B2

C1 100% C3

C2 30% C4

B3

40%

30%

100% B1

1) The comparison is based on the accounting rules in place in a mjority of countries. Such accounting rules may be signif icantlydifferent in some other countries such as, for instance, the UK and IE.

A

A

A

COUNTRY B

100%

100%

37ECB




relationships1.Direct-ownership approach 2.Accounting/enterprise 3.Statistical consolidation (FCS)

consolidation

Coverage • First level of the chain of • Enterprise consolidation • Data collected for all levelsaffiliates (immediate level (usually only affiliates owned of the investment chainof ownership) at more than 50 per cent) following the rules of the FCS

Pro’s

1. Concept • Simplicity (> 10%) • Simplicity (> 50%) • Fully compliant with• Acceptable approximation to international standards – FCS

international standards2. Availability of data • Data available and more • Data to some extent available

accessible than for the other (> 50 per cent ownership)two options

3. Reporting burden • Lower reporting burden forrespondents

• Lower costs for compiler4. Breakdowns • Geographical breakdown • Geographical breakdown

• Activity breakdown • Activity breakdown5. Data quality • Easier to avoid asymmetries • Good estimate of stocks and • If perfectly applied, no

• Reliability of the available profits asymmetries 1)

data • Good analytical value • Best estimate of stocks, profitsand income

• Offers the highest analyticalvalue

6. Feasibility • The most feasible to implementin the short – medium term

Con’s

1. Concept • Deviation from international • Not fully consistent with FCS • Complex for reportersstandards (50% consolidation perimeter, • Risk of inconsistency between

100%-consolidation approach, flows and stocks, since flowsetc.) are consistent with approach 1

• More complex for reportersthan 1

• Risk of inconsistency betweenflows and stocks, since flows areconsistent with approach 1

2. Availability of the data • Availability (problematic due • The longer the chain the moreto EU Regulation) 2) difficult to access to the data

• In some cases, can even not beavailable

3. Reporting burden • Reporting burden for • Highest reporting burdenrespondents higher than alt. 1 • Highest costs for compiler

• More costly for compiler thanalt. 1

4. Breakdowns • Difficulties to distinguishbetween domestic and foreignsubsidiaries

• No geographical break down• No activity breakdown

5. Data quality • Likely underestimation of • Cannot ensure a symmetricequity stocks and profits treatment (the same assets/

• Low analytical value liabilities can be accounted forby several companies along thechain)

• The 50% consolidation perimeter • Higher risk of asymmetriesmay underestimate results, (as the rules are morewhile the application of the complex)100%-consolidation method mayover estimate them

6. Feasibility • The least feasible toimplement in the short –medium term

1) Though, as it is more complex, the risk of asymmetries is much higher than in the case of the direct-ownership approach.2) Not all countries may have access to the consolidated accounts of their reporters, due to the specific accounting guidelines and national legislationin place in each country. In particular, some countries do not require consolidated accounts from resident enterprises if their annual accounts areconsolidated into the accounts of an enterprise governed by the law of a European Economic Area (with certain exceptions).

Table 2 Comparison between: ( i) direct-ownership approach; ( i i ) accounting consol idation;and ( i i i ) stat ist ica l consol idat ion (FCS)


CURRENT PRACTICES: RESULTS OF THEQUESTIONNAIRE

INTRODUCTION77. During the constant review of currentdevelopments carried out by the TF-FDI, itbecame evident that there was a significantdistance between theory and practice for thecoverage of indirect FDI relationships. For thisreason, the TF-FDI designed a questionnairewith a view to investigating current practicesand the most significant problems that countriesencounter to comply with internationalstandards.

78. This section is a schematic overview of theanswers provided by the TF-FDI members tothe questionnaire and is structured in threeparts: (i) current practices; (ii) most significantproblems to cover indirect FDI relationships;and (iii) feasibility and costs of switching to analternative system. Some other results of thisquestionnaire are shown in the last section ofthis chapter (in connection with the possibledevelopment of a European database onownership structures as a potential informationsource to cover indirect FDI relationships) andin chapter 7 (statistics based on the UBOprinciple).

CURRENT PRACTICES79. Countries were asked about which principlethey follow to compile FDI statistics as regardsthe coverage of indirect FDI relationships. Fourpossible replies were suggested:

(i) Just cover direct links of ownership.

(ii) Use consolidated accounts as anapproximation to the coverage of indirectFDI relations, without any othersupplementary information. The rulesapplied to elaborate consolidated accountsare normally based on the nationalaccounting regulation, which varies fromcountry to country (specially concerningthe scope for consolidation, i.e. normally10, 20 or 50%). Normally data can not beused directly, since the geographical

dimension (and breakdowns by sector ofactivity) is necessary. Some extrainformation is normally required and, thus,most (if not all) of the replies classifyingtheir methodology under this categoryshould rather be moved to the fourth(residual) block.

(iii) Application of the FCS as prescribed byinternational standards, by means of, forinstance, direct information requestedfrom respondents, calculations made bythe compiler based on feedback fromreporters, ITRS and other publicinformation sources (such as annualreports, financial press and websites,Dunn&Bradstreet, etc.) From the feedbackobtained from the respondents to thequestionnaire, it turned out that mostcountries included in this category alwayshave to admit exceptions to the fullapplication of the FCS due to practicalproblems, and should thus be moreproperly considered under the residualcategory.22

(iv) Other methods (basically a mixture of theprevious options).

80. Bearing in mind the above-mentionedreservations concerning the replies to thequestionnaire, the total results of thequestionnaire have been summarised in Table 3,which distinguishes between inward andoutward FDI, between stocks and flows andwith a split by FDI components.

22 For instance, Belgium includes associates of associates andGermany only includes indirectly related data when the directlink is above 50%.

39ECB




relationships

Table 3 Coverage of indirect FDI relat ionships in FDI statist ics (number of countriesincluded in each option1)

1) All EU member states except LU, i.e. 14 responses.2) Just 13 countries as RE are not available in the Spanish b.o.p.

Direct relations Accounting FCS Other methodsconsolidation

Inward FDI Flows Equity capital 8 3 2 1Reinvested earnings2) 6 4 2 1Other capital 4 3 2 5

Stocks Equity capital 6 4 3 1Other capital 4 3 2 5

Outward FDI Flows Equity capital 8 2 2 2Reinvested earnings2) 6 2 2 3Other capital 4 2 2 6

Stocks Equity capital 6 2 3 3Other capital 5 2 2 5

81. Another issue that was investigated was theextent to which countries could distinguishbetween direct and indirect FDI relationships intheir FDI data. The intention was to figure outhow easy it could be to countries to exclude(include) indirect FDI relationships from FDIfigures if a common approach was decided atthe euro area/EU level. The outcome was that,for most FDI items and countries, separatefigures for indirect relations are not available.

82. Additional information: only 3 countrieshave more information available than what isfinally published: AT, BE and DE compile somedata on indirect links of ownership which is notadded to their publications or only at thenational level (DE and AT).

– Though from the replies to the questionnaireit appeared that there is no visible differencebetween inward and outward FDI statistics asto whether countries do or do not incorporateindirect FDI relationships to their FDIstatistics, some countries revealed later thatthey have more difficulties in the case ofoutward FDI.

– Although a majority of countries incorporatesome data on indirect FDI relations to theirFDI statistics, most of them cannot

distinguish indirect from direct links ofownership, since these data are often derivedfrom enterprises’ consolidated accounts, inwhich there is no such a distinction (and it’snot a current output requirement).

– Therefore, the achievement of a unique andhomogeneous methodology across the EUcountries concerning whether or not (andhow) indirect FDI relationships should beincorporated to FDI statistics seems adifficult task in the current circumstances.Those countries which currently onlyconsider direct relations have practicaldifficulties to extend their coverage so as tocover indirect FDI links. On the other hand,countries currently including such indirectlinks in their statistics would have seriousdifficulties to exclude them, since they are notseparately distinguished (and would not bewilling to do so, as it would be considered asa “backward” step in their methodology).

MOST SIGNIFICANT PROBLEMS TO COVERINDIRECT FDI RELATIONSHIPS83. The most important problems identified bythe questionnaire are as follows:

– Access to the relevant information: thesystems to collect data on indirect relationsare normally very costly and an appropriatecoverage is difficult to guarantee.


– Identification of the target population,specially in the case of indirect relationsbelow 50%.

– Timing problems, e.g. changes in ownershipstructures are normally not available in time.

– Difficulties to check the data collected.– Difficulties to get more detailed data, since

domestic respondents may not have access tothe accounts of such indirectly related foreignaffiliates.

FEASIBILITY AND COSTS OF SWITCHING TO THEALTERNATIVE APPROACH84. Mostly for the sake of ensuring a consistentand homogeneous way of compiling theEuropean aggregates, it became clear thatcountries would need to agree on a commonapproach towards the coverage of indirect FDIrelations. The information included in Table 3showed that European countries are basicallysplit into two groups: those that stick to thecoverage of only direct links of ownership andthose that incorporate indirect FDI relations totheir FDI statistics.

85. Therefore, the respondents to thequestionnaire were asked about the feasibilityof changing their current system to thealternative solution. More than half of thecountries declared that costs associated to sucha change would be difficult to assume.

86. Another alternative that was explored wasthe possibility to keep on with current practicesin the compilation of national statistics and justchange the methodology and coverage for thecontribution to the European aggregates (so thatall countries applied a common methodology atthe extra euro area/EU level.) That solutionwould imply, de facto, the existence of twoparallel methodologies in some countries. Mostcountries rejected such a possibility as it wasdeemed not cost-effective, implied an increasein the burden on respondents and required acomplete change in the legal framework.

Nevertheless, the replies to the questionnaireexpressed some consensus on two points:

• Changes could only be acceptable to theextent that (i) all countries accepted anychange in parallel; and (ii) the final outcomeimplied a closer alignment to internationalstandards.

• Any such change should be implementedwithin a long-time perspective, since theadptation of systems and legislation wouldrequire sufficient time lag.

DIFFERENT CONSOLIDATION APPROACHES ANDTHE GEOGRAPHICAL ALLOCATION OFTRANSACTIONS AND POSITIONS: IMPACT ON THECOMPILATION OF THE EUROPEAN AGGREGATES

87. One key issue in the compilation process forthe European aggregates is the need forconsistency in the methodologies applied by allmember states. Therefore, the existence ofdissimilar consolidation approaches for thecompilation of FDI statistics is a potential riskwhose distortions should be carefully analysed.In addition, and in connection with thetreatment of indirect FDI relationships, thegeographical attribution of related flows/stockscould also trigger serious distortions in thecompilation of the euro area aggregates. Inparticular, the existence of transactions andpositions that could be recognised in thestatements of several enterprises (located indifferent countries) pertaining to the samegroup could imply some risk of omissions ordouble recording in the European aggregates.

88. To analyse in more detail the consequencesthat any decision concerning the treatment ofindirect FDI relationships could implyconcerning the European aggregates, a moredetailed example is presented in the next sub-section. Some conclusions concerning the needfor a homogeneous approach and somerecommendations concerning the geographicalallocation of transactions and positions arepresented immediately after.

41ECB




relationships

S1

S2

S3 S4

GERMANY FINLAND

UNITED STATES

SPAIN UK

EQ=60 RE=10

EQ=100 RE=15 EQ=40 RE=10 EQ=20 RE=10

P100

60

40

AR

EA

UN

ION

EU

RO

EU

RO

PE

AN

20

ILLUSTRATIVE EXAMPLE: DISTORTIONS THATDIFFERENT APPROACHES TO THE COVERAGE OFINDIRECT FDI RELATIONSHIPS MAY EXERT ONTHE EUROPEAN AGGREGATES89. This example relies on the assumption that,for the sake of simplicity, two basic variantsexist for the compilation of FDI statistics: (i)first shot approach (i.e. only direct FDIrelationships are considered); or (ii) applicationof the FCS. Concerning the second alternative,for the production of FDI equity stocks andreinvested earnings the following applies:

– in the assets side (direct investmentabroad), the domestic direct investor haveto consider, in addition to the equity capitalof the directly owned (non-resident) FDIcompanies, all reinvested earningsgenerated by such directly owned foreigncompanies as well as those generated byindirectly owned enterprises;

– in the liabilities side, in addition to theequity capital of the domestic FDIcompany, the (non-resident) directinvestor is also attributed the reinvestedearnings of both the domestic FDIcompany and its directly and indirectlyowned resident and non-resident affiliates.

90. Along these lines, let us consider amultinational group, whose mother company islocated in DE. The subsequent investments ofthe group are placed inside the euro area (FI andES), inside the EU (UK) and outside the EU(USA) respectively, according to the diagramentitled Example 1. The figures in Example 1reflect the situation at the end of 2001 in termsof equity capital and reinvested earnings(reserves) of the companies. The arrowsrepresent funds flowing from the parentcompanies to their respective affiliates. With aview to simplifying the example, all directinvestment relationships imply 100% ofownership (see Chart 7).

Chart 7 Example: multinational group with direct and indirect FDI l inks


• For the compilation of both nationalstatistics and European aggregates, theremight be three possible scenarios:

(ii) Scenario 1. FI applies the FCS for thecompilation of FDI statistics, while DE,ES and the UK compile inward andoutward FDI statistics according to the“first-shot” approach.

(iii) Scenario 2. All European countries, bothinside and outside the EU, apply the FCSfor the compilation of FDI statistics.

(iv) Scenario 3. All European countriescompile FDI statistics according to the“first-shot” approach.

• The first scenario may resemble the currentsituation in the EU, while the other twoscenarios may represent the two alternativesolutions that could be proposed to avoiddissimilar practices in the EU, i.e. eitherpromoting the application of the FCSacross all EU countries or imposing that allcountries compile FDI stocks on the basisof the first-shot approach, alternatively.

(i) According to Scenario 1 (in which only FIapplies the rules of the FCS as previouslydescribed), the following entries would berecorded in national and Europeanstatistics (the details in brackets reflectwhether the figures come from eitherequity capital or reinvested earnings andthe country of location of the companyoriginating each entry):

National statistics

– GermanyOutward FDI equity stocks vis-à-vis FI:100 (EQ FI) + 15 (RE FI) = 115Total RE vis-à-vis FI: 15Inward FDI equity stocks = 0

– FinlandOutward FDI equity stocks vis-à-vis USA:60 (EQ USA) + 10 (RE USA) + 10 (RE ES) +10 (RE UK) = 90Total RE vis-à-vis USA: 30Inward FDI equity stocks vis-à-vis DE: 100 (EQ FI) + 15 (RE FI) + 10 (RE USA) +10 (RE ES) + 10 (RE UK) = 145Total RE vis-à-vis DE: 45

– SpainOutward FDI equity stocks vis-à-vis UK:20 (EQ UK) + 10 (RE UK) = 30Total RE vis-à-vis UK: 10Inward FDI equity stocks vis-à-vis USA:40 (EQ ES) + 10 (RE ES) = 50Total RE vis-à-vis USA: 10

– United KingdomOutward FDI equity stocks: 0Inward FDI equity stocks vis-à-vis ES:20 (EQ UK) + 10 (RE UK) = 30Total RE vis-à-vis ES: 10

Euro area aggregates (only ES and FI wouldreport transactions and positions vis-à-visnon-euro area countries):Outward FDI equity stocks vis-à-vis USA +UK: 60 (EQ USA) + 10 (RE USA) + 10 (REES) + 10 (RE UK) + 20 (EQ UK) + 10 (REUK) = 120Total RE vis-à-vis USA + UK: 40Inward FDI equity stocks vis-à-vis USA:40 (EQ ES) + 10 (RE ES) = 50Total RE vis-à-vis USA: 10

European Union aggregates:Outward FDI equity stocks vis-à-vis USA:60 (EQ USA) + 10 (RE USA) + 10 (RE ES) +10 (RE UK) = 90Total RE vis-à-vis USA: 30Inward FDI equity stocks vis-à-vis USA:40 (EQ ES) + 10 (RE ES) = 50Total RE vis-à-vis USA: 10

43ECB




relationships

(ii) According to Scenario 2 (in which allcountries compile FDI statistics incompliance with the FCS), the followingentries would be recorded in national andEuropean statistics:

National statistics

– GermanyOutward FDI equity stocks vis-à-vis FI:100 (EQ FI) + 15 (RE FI) + 10 (RE USA) +10 (RE ES) + 10 (RE UK) = 145Total RE vis-à-vis FI: 45Inward FDI equity stocks = 0

– FinlandOutward FDI equity stocks vis-à-vis USA: 60(EQ USA) + 10 (RE USA) + 10 (RE ES) + 10(RE UK) = 90Total RE vis-à-vis USA: 30Inward FDI equity stocks vis-à-vis DE:100 (EQ FI) + 15 (RE FI) + 10 (RE USA) +10 (RE ES) + 10 (RE UK)=145Total RE vis-à-vis DE: 45

– SpainOutward FDI equity stocks vis-à-vis UK:20 (EQ UK) + 10 (RE UK) = 30Total RE vis-à-vis UK: 10Inward FDI equity stocks vis-à-vis USA = 40(EQ ES) + 10 (RE ES) + 10 (RE UK) = 60Total RE vis-à-vis USA: 20


Euro area aggregates:Outward FDI equity stocks vis-à-vis USA andUK: 60 (EQ USA) + 10 (RE USA) + 10(RE ES) + 10 (RE UK) + 20 (EQ UK) + 10(RE UK) = 120Total RE vis-à-vis USA + UK: 40Inward FDI equity stocks vis-à-vis USA:40 (EQ ES) + 10 (RE ES) + 10 (RE UK) = 60Total RE vis-à-vis USA: 20

European Union aggregates:Outward FDI equity stocks vis-à-vis USA: 60(EQ USA) + 10 (RE USA) + 10 (RE ES) + 10(RE UK) = 90Total RE vis-à-vis USA: 30Inward FDI equity stocks vis-à-vis USA = 40(EQ ES) + 10 (RE ES) + 10 (RE UK) = 60Total RE vis-à-vis USA: 20

(iii) According to Scenario 3 (in which allcountries compile FDI statistics accordingto the first-shot principle), the followingentries would be recorded in national andEuropean statistics:

National statistics

– GermanyOutward FDI equity stocks vis-à-vis FI:100 (EQ FI) + 15 (RE FI) = 115Total RE vis-à-vis FI: 15Inward FDI equity stocks = 0

– FinlandOutward FDI equity stocks vis-à-vis USA:60 (EQ USA) + 10 (RE USA) = 70Total RE vis-à-vis USA: 10Inward FDI equity stocks vis-à-vis DE = 100(EQ FI) + 15 (RE FI) = 115Total RE vis-à-vis DE: 15

– SpainOutward FDI equity stocks vis-à-vis UK:20 (EQ UK) + 10 (RE UK) = 30Total RE vis-à-vis UK: 10Inward FDI equity stocks vis-à-vis USA = 40(EQ ES) + 10 (RE ES) =50Total RE vis-à-vis USA: 10



Euro area aggregates:Outward FDI equity stocks vis-à-vis USA +UK: 60 (EQ USA) + 10 (RE USA) + 20 (EQUK) + 10 (RE UK) = 100Total RE vis-à-vis USA + UK: 20Inward FDI equity stocks vis-à-vis USA:40 (EQ ES) + 10 (RE ES) = 50Total RE vis-à-vis USA: 10

European Union aggregates:Outward FDI equity stocks vis-à-vis USA:60 (EQ USA) + 10 (RE USA) = 70Total RE vis-à-vis USA: 10Inward FDI equity stocks vis-à-vis USA:40 (EQ ES) + 10 (RE ES) = 50Total RE vis-à-vis USA: 10

CONCLUSIONS FROM THE EXAMPLE91. It is evident that the results obtainedfollowing the three approaches are remarkablydifferent, from the point of view of bothnational statistics and European aggregates. Theexample is not necessarily overly realistic,specially concerning the proportion of earningsgenerated by each affiliate, which has beenexaggerated for illustrative purposes.

92. Bearing this in mind, the results obtained arepresented in Table 4. The net FDI position(outward – inward FDI equity stocks) of theeuro area would be 70, 60 and 50 according tothe three scenarios. For the EU, the results ofthe three scenarios would be 40, 30 and 20. Asregards the net flows (i.e. credits minus debits)of annual reinvested earnings in the euro areab.o.p. the three scenarios would register 30, 20

and 10, respectively, while for the EU, theywould be 20, 10 and 0 (see Table 4).

93. The main conclusions that can be extractedfrom these results are summarised as follows:

(i) Scenario 1 (dissimilar approaches acrosscountries)

National statistics

– Global results cannot be deemed consistentacross countries due to heterogeneouspractices.

– Net results (outward – inward FDI) are notcomparable across countries.

– There is no symmetry between counterpartcountries recording the same FDI stocks/flows.

– The risk of double attribution of reinvestedearnings generated by indirectly owned FDIcompanies in assets (outward FDI) by morethan one country without a counter entry inliabilities (inward FDI) exists.

European aggregates

– Reinvested earnings of indirectly ownedenterprises could be recorded several times ormissing in the European aggregates withoutappropriate counter entries in liabilities(inward FDI), depending on whether thecontributing countries apply or not the FCSfor the compilation of FDI statistics.

Euro area EU

Scenario 1 Scenario 2 Scenario 3 Scenario 1 Scenario 2 Scenario 3

Equity stocks Outward 120 120 100 90 90 70Inward 50 60 50 50 60 50Net 70 60 50 40 30 20

Reinvested earnings Outward 40 40 20 30 30 10Inward 10 20 10 10 20 10Net 30 20 10 20 10 0

Table 4 Summary of the results

45ECB




relationships

– Net results (outward – inward FDI) will,thus, be distorted23

94. Obviously, the first scenario (which couldbe rather close to reality) implies manyworrying consequences on both national and,specially, on European statistics. Additionally,the lack of a common methodology does notenable a transparent approach vis-à-vis users,as practices differ among countries. A logicalconclusion would be that a common solutionshould be promoted and applied by all EUcountries. Such a common solution could becoincident with either scenario 2 or scenario 3.Let us analyse the consequences of eachscenario, without entering into how feasibleeach one could be on practical grounds.

(ii) Scenario 2 (all countries compile FDIstatistics according to the FCS)

National statistics

– Global results are comparable due tohomogeneous practices

– Net results (outward – inward FDI) are alsocomparable across countries.

– Counterpart countries record the same FDItransactions/positions in a symmetric way.

– Reinvested earnings generated by indirectlyowned FDI enterprises are recorded by morethan one country, thus implying larger grossfigures (i.e. inward and outward FDI).

– Since companies in the middle of a chain willrecord the same amounts in assets andliabilities, there will be no impact on the net.

European aggregates

– Reinvested earnings of indirectly ownedenterprises would be recorded several timesin the gross FDI figures of the Europeanaggregates. Therefore, inward and outwardFDI will register larger figures than inscenario 3.

– Net results (outward – inward FDI) are notdistorted, since reinvested earnings ofindirectly owned companies are recordedmore than once in the assets side, but are alsorecorded in liabilities by companies in themiddle of a chain, so that the net FDI justregisters reinvested earnings once.

(iii) Scenario 3 (all countries compile FDIequity stocks on the basis of non-consolidated accounts)

National statistics

– Global results can be deemed consistentacross countries, since all of them followhomogeneous practices

– Counterpart countries record the same FDItransactions/positions in a symmetric way.

– Net results (outward - inward FDI) arecomparable across countries.

– However, the net results are not correctaccording to international standards and aredifferent than in scenario 2 due to the non-recording of reinvested earnings generated byindirectly owned FDI enterprises. Whetherthe impact on the net results is positive ornegative (compared with scenario 2) cannotbe ascertained a priori.

European aggregates

– Reinvested earnings of indirectly ownedenterprises are not recorded. Therefore, gross(inward and outward) FDI figures are lowerthan in scenario 2.

– Net results (outward – inward FDI) are notconceptually correct and different than inscenario 2. Whether the impact on the netresults is positive or negative (compared withscenario 2) cannot be ascertained a priori.

23 In the example, the reinvested earnings generated by the UKcompany are recorded twice in outward FDI (accounted for byboth ES and FI), while they are omitted in inward FDI, since ESdoes not follow the FCS.


THE GEOGRAPHICAL ALLOCATION OF RELATEDFLOWS/STOCKS95. In addition to the above-mentionedproblems, the way in which these transactions/positions are broken down geographically mayeven imply further distortions for the Europeanaggregates.

96. For instance, according to scenario 3, let usassume that all countries attribute all entries tothe ultimate counterpart. In that case, the b.o.p./i.i.p. of both Finland and Germany wouldrecord extra-euro area entries under FDI abroad(assets) accounting for the reinvested earningsgenerated by the company located in the USA.Since the b.o.p./i.i.p. in Finland would recordthose reinvested earnings in liabilities againstGermany (i.e. as an intra euro area flow/stock),only the (extra euro area) asset entries would berecorded in the euro area aggregates, thusimplying a distortion in both gross and net euroarea figures. Similar distortions would occur inthe case of the reinvested earnings generated inUK or ES.

97. These distortions could be avoided if allcountries attributed all FDI flows and stocks tothe “first shot” counterpart, i.e. according to thelocation of the directly related affiliates, in thecase of outward FDI, and according to thelocation of the non-resident investor, in the caseof inward FDI.

CONCLUSIONS AND RECOMMENDATIONS

SIMPLIFICATION PROPOSALS TOWARDS THECOVERAGE OF INDIRECT FDI RELATIONSHIPS98. Chapter 1 established that compliance withinternational standards implies that indirect FDIrelationships should be effectively incorporatedto FDI statistics. The review of currentpractices has revealed how difficult this is onpractical grounds. EU member states do notcurrently follow a homogeneous approach,since they are basically split into the group ofcountries that only cover direct FDIrelationships and those other that also coverindirect relationships (to different extents).

99. The illustrative example analysed in thischapter proved that the current situation(similar to scenario 1 considered in theexample) is very harmful for the quality of theEuropean aggregates. Therefore, a commonsolution should be agreed concerning whetheror not (and to which extent) indirect FDIrelationships should be part of FDI statistics.

100. Given the numerous practical difficultiesrevealed, the TF-FDI considers that a fullapplication of the FCS by all countries isunfeasible on practical grounds. Therefore, theonly two possible alternatives for a commonapproach at the EU level seem to be: (i) that allcountries cover just direct (first-shot) FDIrelationships; or (ii) to fix a bottom lineconcerning the minimum indirect FDIrelationships that all countries should be in aposition to cover in the medium term.

101. The first alternative (only cover directFDI relationships) has the advantage ofsimplicity and a lower burden on respondents(which are obliged to report less information)and, in same cases, also on compilers. Inaddition, and mainly from the point of view ofthe compilation of supranational aggregates,this approach would suffer from fewerproblems concerning the geographicalallocation of flows and stocks. Conversely, themain problem of this approach is that the resultsobtained would not be fully compliant withinternational standards and the outcome wouldoffer a somewhat lower analytical value,specially considering the increasing role ofspecial financial vehicles, clearing centres, etc.in the investment strategy of multinationalgroups.

102. The second alternative would consist ofestablishing a bottom line for the coverage ofindirect FDI relationships that all countriesshould be in a position to surpass. Such aminimum common approach would narrowdown the risk of asymmetries and would reducethe impact on the European aggregates of thedifferent methodologies applied in memberstates. The most important difficulty would be

47ECB




relationships

the exploration of practical ways for collectingthe necessary information, since the longer thechain of links between companies, the moredifficult it is to get access to the balance sheet offoreign subsidiaries with no direct link to thedomestic mother company. For this reason, asimplification of the rules described in the FCScould reduce the obstacles existing to apply the“statistical consolidation” approach.

103. Some testimonies in the TF-FDI pointedtowards the significant proportion representedby FDI relationships above 50% over the totalFDI figures. Additionally, this information ismore easily available to domestic respondentsin those cases in which there is an obligation tocompile consolidated accounts. Therefore, itwas concluded that efforts should aim atappropriately covering at least this kind oflinks.

104. Against this background, the TF-FDIconsidered that two simplification approachesshould be deemed acceptable minimum commonstandards for the coverage of indirect FDIrelationships and could, thus, constitute thebottom line that all countries should reach inthe medium term:

(i) The coverage of indirect links ofownership above 50%.24

(ii) The coverage of direct and indirect links ofownership above 10%, calculated as theproduct of the subsequent links ofownership along a chain.

GEOGRAPHICAL DISTRIBUTION OF FDI FLOWS/STOCKS RELATED TO INDIRECT FDI LINKS105. The illustrative example analysed in thischapter revealed some problems that non-fullyharmonised criteria for the geographicalallocation of transactions and positions relatedto indirect FDI links of ownership could entailfor the quality of the European aggregates. Inorder to avoid such possible distortions, thefollowing recommendations should apply:

– Reinvested earnings should begeographically allocated to the immediateaffiliate (direct investment abroad) orimmediate mother company (directinvestment in the reporting economy), i.e.the one with which the investor/directinvestment enterprise maintains a direct linkof ownership. This criterion should applyirrespective of whether the retained profitsare actually generated by a differentcounterpart along the chain of ownership.

– Likewise, FDI equity stocks should beattributed to the immediate affiliate (DIabroad) direct investor (DI in the reportingeconomy) even if, in some cases, asubstantial part of the total value may begenerated by indirectly linked enterprisesfurther down in the ownership chain.

106. It is acknowledged that these criteriamay result in less valuable statistics from theanalytical viewpoint. For this reason, the TF-FDI would encourage countries to collect andpublish additional information on thegeographical allocation of FDI flows and stocksbased on the residence of the ultimate beneficialowner, whenever such information were not toodifficult to obtain (see chapter 7).

EUROPEAN DATABASE ON OWNERSHIPSTRUCTURES

INTRODUCTION107. In connection with the coverage ofindirect FDI relationships, the TF-FDI mandatestated that possible solutions to the problem ofobtaining information on group structuresshould be examined with reference also to thecosts that they would entail.

108. Some relevant information is oftenpublicly available in annual reports, media,commercial data providers, etc. Besides theseinformation sources, the idea of sharinginformation through a common platform, such

24 All direct links of ownership above 10% would still need to becovered.


as a centralised database to be used by Europeancompilers, has been suggested in severaloccasions. To some extent this initiative couldresemble for direct investment the envisagedrole of the Centralised Securities Database inthe area of portfolio investment.

109. The TF-FDI analysed the issue from twodifferent points of view: (1) from the point ofview of potential data providers; and (2) fromthe point of view of users of the information.The first aspect is further developed in the firstsection, in connection with a number ofnational studies carried out by the TF-FDI onthe basis of the information currently availableto NCBs. The second aspect has beenconsidered in the framework of the informationreceived by the TF-FDI right before its lastmeeting concerning an ongoing project todevelop a European Business register, which, atthat time, was under consideration by Eurostatand the ECB.

( I ) NCBS AS POTENTIAL DATA PROVIDERS

RESULTS OF THE SURVEY ON INDIRECTRELATIONSHIPS110. The questionnaire on indirect FDIrelationships mentioned at the beginning of thischapter sought the TF-FDI members’ viewswith regard to the willingness of MS toparticipate in a European register ofmultinational ownership structures as potentialdata providers, from a fairly generalperspective. All countries but 3 would agree toprovide the data. Most of them would only bewilling to update the data provided to such aEuropean database on an annual basis.

111. Finally, some problems were identified,basically linked to confidentiality constrains,need to adapt national legislation, requirementfor additional resources, and need to develop acommon platform with internationalidentification codes as a necessary prerequisite.

112. To sum up, most countries could deliverthe data on an annual basis on certainconditions, namely:

– if its usefulness is studied before any productis launched;

– if confidentiality is ensured;

– if an appropriate legal framework is foreseen.

RESULTS OF THE SUB-GROUP ASSESSING THEFEASIBILITY OF FEEDING A EUROPEAN DATABASEON OWNERSHIP STRUCTURES113. With a view to further exploring thefeasibility of such a centralised databasecontaining information on ownership structureswithin multinational groups, four countries(FR, ES, DE and IT) studied whether or not andhow they could contribute to the project. Inparticular, the four countries studied at whichfrequency and with which information theycould contribute to the feeding of the database,how costly it could be and which main obstacleswould require a way out.

114. The main findings of the countriesparticipating in the study can be summarised inthe following blocks:

Type of information available

– DE: 2 databases for trade statistics. 130.000resident/60.000 non-resident enterprisescovered respectively

– ES: external loans register. Cover all non-banks receiving/granting a loan from/toabroad.

– FR: FDI register of FR companies includingtheir links with non-resident entities.

– IT: no such register is currently available.The UIC is currently running a researchproject to study the feasibility of maintaininga business register for the compilation of FDIstatistics in IT. Some of the sources underinvestigation are Istat, Dun & Bradstreet,CERVED, CONSOB and Foreign TradeInstitute/R&P.

49ECB




relationships

Specific problems to each country

– DE: the two main problems are (i) the lack ofownership information in the data currentlyavailable in the trade databases; and (ii)confidentiality rules for the treatment of theinformation.

– ES: the main problem would be the additionalcosts that ensuring quality would entail, sincethe use of ownership information is fairlylimited at present and, therefore, is notexhaustively checked.

– FR: the two main concerns would be: (i) howto ensure confidentiality; and (ii) theadditional costs that it could entail.

– IT: difficult to answer at the current stage,since no such information is currentlyavailable.

General problems of the project

– Problems previously mentioned by some ofthe countries (non-availability of the data,how to ensure confidentiality, need foradditional resources, costs, etc.)

– Technical problems to adapt the informationthat should be available in the database to thestructure of multiple and different nationalcollection systems.

– Most public sources only cover listedenterprises. How to obtain information fornon-listed enterprises would require a muchharder and deeper investigation.

– History dimension would be required, i.e. thespecific situation of each individual group atdifferent time periods should be maintained inthe database so as to allow the compilation ofstatistics across time.

– The existence of individual identificationcodes (e.g. ISIN) for each enterprise is anabsolute must to permit the development ofsuch a database.

Overall conclusions

– The existence of a centralised database withinformation about the structure ofmultinational groups would be seen as avery useful tool for the compilation of FDIstatistics.

– However, a number of significant problemshave been identified. Some of theseproblems are deemed difficult to overcome.

– A very preliminary estimation of the costshas revealed that they could be high.

(2) NCBS AS POTENTIAL USERS OF THEINFORMATION

115. As stated in the introduction, at its lastmeeting the TF-FDI received information on aproject of the Eurostat’s Business StatisticsDirectorate which, at that time, was underconsideration. The project basically consistedof the development of a pan-European businessregister including information on themultinational groups’ ownership structure forstatistical purposes. The project would be basedon an update of Regulation (EC) No 2186/93 of22 July 1993 on Community co-ordination indrawing up (national) business registers forstatistical purposes.

116. Due to the late notice at which the TF-FDIreceived information on this project, it was notpossible to consider in more detail its potentialusefulness for the purposes of the compilationof FDI statistics. However, as stated in theprevious subsection, the TF-FDI recognised thenumerous practical difficulties that thecollection of this type of information may entailfrom the individual country perspective.

117. For this reason, The TF-FDI is of theopinion that a harmonised and multilateralsolution should be highly welcome from thepoint of view of users of the possible productsthat such a database could put at the disposal ofFDI compilers. Given the above-mentionedtime constrains, it was not possible to further


consider how the TF-FDI could contribute tosuch a project.

For this reason, as a follow-up work to the TF-FDI, it is suggested that other bodies, forinstance, the ECB’s WG-BP&ER and theEurostat’s Balance of Payments WG, elaboratethe list of user requirements which wouldpermit that the final product could be used forthe compilation of FDI statistics.

51

3 Valuationof FDI equity

stocks

ECBFore ign d i rec t inves tment – Task force repor t

March 2004

INTRODUCTION

118. In the recent years, different issuesrelated to the valuation of foreign directinvestment (FDI) equity stocks have beenconsidered in several fora. Following thoroughinvestigation, in 2001 the ESCB StatisticsCommittee (STC) and the Working Group onBalance of Payments and External ReservesStatistics (WG-BP&ER) reached someconceptual agreements related to the generalrules that should guide the valuation of thesestocks in the euro area international investmentposition (i.i.p.). Some practical difficulties toimplement these decisions were recognised,inter alia, the time schedule for putting theseagreements into practice.

Following its mandate, the TF-FDI exclusivelyconsidered the valuation criteria approved bythe STC and, thus, solely focused on how toapply the agreements reached by the STC onpractical grounds. Other valuation methods,such as macroeconomic revaluation indexes orcurrent-cost methods as presented by the USAin the November 2002 IMF BOP Committee, arenot considered in this chapter.

119. Within the above-described framework,the TF-FDI carried out an analysis of thecurrent state of play. Based on this analysis, atthe end of this chapter the TF-FDI addressessome conclusions and recommendations,putting special emphasis on their applicability.All issues related to the use of consolidatedaccounts for the valuation of FDI equity stocks(in particular, whether the common definition ofOFBV should be applied on consolidated or onnon-consolidated accounts of the directinvestment enterprises) have been alreadytackled in the previous two chapters. Inparticular, the TF-FDI addressedrecommendations on how to incorporateindirect links of ownership to the total book-value-based FDI equity stocks.

120. This chapter is in three sections. The firstone contains a brief summary of the mainrelated decisions adopted by the STC and the

3 V A LUAT I ON O F F D I E QU I T Y S TO CK S 25

WG-BP&ER. The second section summarisesthe answers of the countries to thequestionnaire on current practices and futureprospects related to the valuation of FDI equitystocks (Table 5 further illustrates the answersreceived from all countries). Finally, sectionthree presents an overview of the results of thenational feasibility studies carried out withinthe TF-FDI concerning the viability ofproducing separate figures for listed and non-listed companies and of collecting two differentvaluations (book values and market values) forFDI in listed companies. The compilation ofFDI stocks at T+9 months is separately coveredin Annex 1.

RELATED DECISIONS ADOPTED BY THE STATISTICSCOMMITTEE AND THE WORKING GROUP ONBALANCE OF PAYMENTS AND EXTERNALRESERVES

121. In the course of 2000, the STC consideredthe distortions exerted by the wide range ofvaluation criteria applied by European UnionMember States for the compilation of FDIequity stocks. The lack of a single set ofvaluation rules was acknowledged as animportant source for inconsistencies in theconstruction of the euro area aggregate. For thisreason, the STC considered the provision ofclearer guidance as a high priority, with a viewto identifying common rules for the valuation ofFDI equity stocks to which all Member Statesshould converge in their contributions to theeuro area aggregate.

122. At the time of deciding on the mostappropriate valuation rules, the analysis ofinternational standards was not fullyconclusive. While both the IMF Balance ofPayments Manual (5th edition) and the OECDBenchmark Definition of Foreign Direct

25 The analysis in this chapter did not cover the valuation of FDIequity stocks arising from real-state investments. Due to theimpossibility to send FDI surveys to the non-resident owners ofreal state in the country (nor to domestic households acquiringproperties abroad), it was concluded that the accumulation offlows could be a reasonable solution for this specific case.


Investment generally recommend the use ofmarket (i.e. stock exchange) prices, in theabsence of such market prices (i.e. for non-listed DI companies), other alternatives areadmitted. Even the use of book values for thevaluation of FDI in listed companies is notconclusively ruled out in either manual.26

123. Considering that most difficulties arelinked to the valuation of FDI companies whentheir shares are not quoted on the stockexchange, the STC decided that the followingcriteria would be the basis for the valuation ofthe euro area inward and outward FDI equitystocks in the future:

– FDI in listed companies’ shares shall bevalued on the basis of stock exchange prices;

– FDI in non-listed companies’ shares shall bevalued on the basis of book values, assumingthe lack of any appropriate market referencefor these companies.

124. In defining these criteria, the STC feltthat, whereas the concept of “stock exchangeprice” was straightforward, a commondefinition of “book values” was needed, notablyto avoid asymmetries between assets andliabilities. Indeed, book values for outward DIcould often be interpreted as accounting valuesin the investors’ books (in many casescoinciding with “historical prices”), while forinward DI, stocks are usually valued on thebasis of the domestic FDI company’s ownfunds.

125. Therefore, the STC decided that thecommon definition would exclusively be basedon the value of the FDI company’s own funds.It was considered that the price recorded in thebalance sheet of the direct investor (i.e. theacquisition/historical price) hardly reflects theevolution of the price of the company throughtime due to the strict valuation rules usually inplace in accounting.

126. The subsequent work consisted in findingout which accounts on the liabilities side of the

direct investment enterprise’s balance sheetshould be considered when assessing the totalvalue of the company based on its volume ofown funds, i.e. its own funds at book value(OFBV). Then, the calculation of FDI equitystocks would consist of applying the percentageof ownership of each direct investor to thecompany’s worth calculated this way.Following this approach, the valuation of DIstocks should show some consistency with theevolution of the true value of the company.

Book values should be understood as the % ofownership of the direct investor times the valueof the DI company based on its volume of ownfunds, which should be calculated according tothe following definition of OFBV:

– Paid-up capital (net of own shares).

– All types of reserves (including sharespremium accounts and investment grants).

– Net value of non-distributed profits andlosses (including results for the currentyear).

127. Moreover, in order to further improveeuro area FDI statistics, the STC envisagedproducing two memorandum items for the total(i.e. without sector or geographicalbreakdowns) inward and outward FDI equitystocks:

(i) FDI equity stocks on the basis of bookvalues (for all types of FDI companies),mostly to ensure continuity in the timeseries; and

(ii) FDI equity stocks marked-to-market (for alltypes of FDI companies), mostly to provideusers with some complementary information

26 Paragraph 377 of the BPM5 reads: “Although this Manual, inconcordance with the SNA, affirms the principle of using marketprice as the basis for valuation, it is recognized that, in practice,book values from the balance sheets of direct investmententerprises (or investors) often are used to determine the valueof the stock of direct investment.” This paragraph seems toimplicitly admit this valuation and does not make any distinctionbetween listed and non-listed companies.

53


stocks


March 2004

for analytical purposes and as a proxy forthe reconciliation with financial accountsstatistics (shares and other equities item).

128. For the practical implementation of allthese proposals, euro area Member Statesshould take the following steps: (i) split thereporting of the equities item within FDI in theeuro area and FDI abroad between listed andnon-listed companies; and (ii) report to the ECBFDI in equities of listed companies on the basisof both market and book values (following theagreed common definition of OFBV).27 Thevaluation of FDI in listed companies on thebasis of book values should be twofold: asdirect input for the first memorandum item andfor the calculation of ratios market valuedivided by book value, which could form thebasis for the production of the secondmemorandum item.

129. After in-depth discussion, Member Statesidentified several practical difficulties incarrying out these principles. Indeed, it wasrecognised that practical problems may alreadybe affecting the compilation of FDI stocks atpresent. Some countries may have difficulties toapply the common definition of OFBVespecially in the case of FDI abroad due to thedifficult access to the details required aboutnon-resident FDI companies. The next sectionprovides some indications on how othercountries have managed to (or plan to)overcome such practical problems to implementthese agreements.

RESULTS OF THE QUESTIONNAIRE ON VALUATIONOF FDI EQUITY STOCKS

130. Following the fact-finding exercise on thecollection of direct investment stocks(September 2000)28, the sub-group designed anew questionnaire to investigate currentpractices of Member States and possible plansconcerning the applicability of the STCdecisions on valuation of FDI equity stocks.Twelve out of the thirteen participatingcountries sent the completed questionnaire.

27 These requests only apply to step-2 aggregates. Step-3breakdowns should only be provided using market values forlisted companies and book values for non-listed companies. Norules have been specified for step-1 figures (i.e. national data).

28 See document ST/WG/BP/DISQUEST.DOC “Collection of directinvestment stocks: outcome of the questionnaire”, 30 October2000.

Ireland and Luxembourg, which did notparticipate in the work of the TF-FDI from itsinception, were not questioned.

131. The main answers to the questionnaire aresummarised in Table 5 (see next page).

DISTINCTION BETWEEN LISTED AND NON-LISTEDCOMPANIES132. Nine countries (DK, ES, FR, IT, AT,PT, FI and GR) are able to (directly) distinguishbetween listed and non-listed companies forinward stocks and five countries (DK, FR, IT,PT, FI) for outward stocks. One country (DE)plans to make this distinction for both inwardand outward stocks beginning with figures atend-2002. Three countries (NL, SE, UK) do nothave any plans. One country (BE) uses a ratiobased on the market capitalisation of listedcompanies compared to the total capitalisationof both listed and non-listed companies toprovide inward stocks broken down betweenlisted and non-listed companies.

Practical solutions

(i) Five countries, namely DK, IT, FI as wellas FR and PT (for outward stocks) rely oninformation provided by respondents tomake this distinction.

(ii) In the case of inward stocks, fourcountries (AT, PT, FI and GR) useregisters of resident listed companiesmaintained by stock exchange authorities,at least for cross-checking purposes (FI).

(iii) Only two countries (IT, FR for inwardstocks) use internal security databases toknow about companies’ status.

(iv) When the information is not provided byrespondents and no register exists or is


Table 5 Questionnaire on the valuation of direct investment stocks – summary table

Distinction Application Possibilty Valuation Application Valuation Source ofbetween of some to provide of DI stocks of the of DI stocks information

listed and form of stocks on in non-listed WG-BP&ER in listed to compilenon-listed consolidation a non- at book agreed companies at marked-tocompanies consolidated value definition market value market

basis of OFBV stocks

Inward Outwardstock stock

Belgium yes no yes yes yes yes yes Adjustedcumulated

flows

Denmark yes yes yes yes yes yes yes Respondents

Germany yes yes yes yes yes yes yes Respondents(beginning (beginning (beginning (beginning

with figures with figures with figures with figuresat end 2002) at end 2002) at end 2002) at end 2002)

Spain yes no yes no Partially Partially Partially Stock(inward (inward exchangestock in stocks in prices

the MFI’s the MFI’ssector) sector)

France yes yes no - yes yes yes Securitydatabase+ otherpublicly

availableinformation

Italy yes yes yes no no no yes/no Perpetual(perpetual (perpetual (respondents inventoryinventory inventory can report method +method) method) book values respondents

yes yes if market(FDI survey) (FDI survey) values not

available)

Ireland yes yes yes no no yes yes Stockexchange

prices

The Netherlands no no yes no yes yes yes/no Respondents(respondentscan decide

for themselveswether theyuse book or

market value)

Austria yes no no - yes yes yes Stock(inward exchange

stock only) prices

Portugal yes yes no - yes yes no -

Finland yes yes yes no yes yes yes Respondents

Sweden no no yes no yes yes no -

United Kingdom no no no - yes yes no -

Greece yes no yes yes yes yes yes Stockexchange

prices

55


stocks


March 2004

available (case of outward stocks), thedistinction is made manually (e.g. AT) bymeans of internal security databases andpublicly available sources (mostlyfinancial press and stock exchange websites) to identify listed companies.

The TF-FDI considers that all these solutionsmay be deemed valid to obtain the split betweenlisted and non-listed companies and, thus, noprioritisation among them is provided in thisreport.

133. As regards the proportion of listedcompanies out of the total direct investmentstocks, few countries were able to provide data.The ratio of the number of listed directinvestment companies to the total number of DIcompanies varies widely, from 0,7% to 12,2%at the end of 2000 (inward FDI stocks). Inproportion of the total amount of the stock, thevariance is even greater (from 0,8% to 25%).Similar results were found concerning the stockof outward FDI.

APPLICATION OF THE “CONSOLIDATED SYSTEM”134. The possible use of consolidatedaccounts for the compilation of FDI statisticswas extensively covered in the previouschapter. The questionnaire only raised thequestion of the application of the “fullyconsolidated system” or of any other form ofconsolidation by Member States.

135. Eight countries (BE, DK, DE, ES, IT, FI,NL, SE) say they apply, at least partially, the“consolidated system” as described in theOECD FDI Benchmark Definition of ForeignDirect Investment, but few give precise answersregarding their methodology. GR does not fullyapply the FCS but, whenever indirect FDIrelations are identified, they are taken intoconsideration in the FDI figures.

136. In some cases, consolidated data arecompiled on the basis of accountingconsolidation (e.g. for inward FDI in Finland).However, the extent to which this is the case,the principles underlying the concept of

“consolidation” in each country and whether allthese facts may constitute a problem ofconsistency in the European aggregates couldnot be investigated sufficiently in detail on thebasis of the answers to the questionnaire.

VALUATION OF STOCKS IN NON-LISTEDCOMPANIES137. All but three countries, namely IT, GRand ES (partially), say they are able to compiledirect investment stocks in non-listedcompanies at book value, applying the WG-BP&ER agreed definition of own funds at bookvalue. It was not clear to the countries thoughwhether such a definition should apply toconsolidated or to non-consolidated balancesheets. Hence different applications bycountries may be a source of asymmetries. Theclarifications provided through the previoustwo chapters should help overcome suchasymmetries in the future.

VALUATION OF STOCKS IN LISTED COMPANIES138. Nine countries (BE, DK, ES, FR, IT, NL,AT, FI and GR) declare being able to compile,at least partially, direct investment stocks inlisted companies at market value.

Practical solutions

(i) Four countries (ES for inward FDI stockin the banking sector, FR, AT and GR) usean individual valuation method based onstock exchange prices and, in the case ofFR, the combination of an internalsecurities database + other publiclyavailable information.

(ii) Four countries (DK, IT, NL and FI) rely oninformation provided by respondents tocompile marked-to-market stocks, while onecountry (DE) plans to do so in the future.

(iii) In two cases though (IT, NL), it seems thatrespondents may report stocks at eitherbook or market value depending on theavailable information, which could impedethe compilation of consistent stocks usingone or the other valuation method.


(iv) In three cases (BE, IT and ES partially),this valuation is made using a perpetualinventory method ().

The TF-FDI considers than (i) and (ii) can bedeemed valid solutions, while (iii) and (iv) arenot recommended.

139. The two main obstacles for compilingmarked-to-market stocks for listed companiesare, first, the difficulty to identify listedcompanies among foreign direct investmentcompanies and, second, the difficulty to gainaccess to stock exchange prices for thesecompanies. The future Centralised SecuritiesDatabase – CSDB – may however help to solvethis problem, allowing an individual valuationof FDI stocks in listed companies.

140. Among countries that do not apply theSTC decision to compile marked-to-marketdirect investment stocks for listed companiesyet, three (BE, DE, ES) plan to change theircollection systems. One country (PT) says itwill rely on available sources and on newassessment exercises to comply. Othercountries (NL, SE, UK) do not have any plansas regards this issue.

141. Six countries (IT, BE, ES, NL, SE, UK)would have difficulties in providing FDI stockson the basis of two different valuationprinciples, i.e. book value and market value, forlisted companies. All of them stressed theadditional costs such a requirement wouldimply. The next section presents a more detailedanalysis on the feasibility of combining thesetwo calculations for FDI in listed companies’shares.

NATIONAL FEASIBILITY STUDIES ON HOW TOCOMPILE FDI IN LISTED COMPANIES’ SHARES ONTHE BASIS OF BOTH MARKET VALUES AND BOOKVALUES

INTRODUCTION142. One of the most significant difficultiesdeclared by EU countries at the time ofimplementing the STC agreements on thevaluation of FDI equity stocks was related tothe collection of FDI in listed companies. Theimplementation of the STC agreements requiredthat FDI in listed companies’ shares should bevalued twice, on the basis of both book values(based on the common definition of OFBV) andmarket values. For this reason, the TF-FDIinvestigated, on the one hand, how somecountries may currently collect this informationand, on the other hand, how the other countrieswould plan to change their collection systems toaccommodate this request.

143. Some countries participating in the TF-FDI decided to carry out individual nationalfeasibility studies (NFS) in order to determinewhether collecting two valuations for FDI inlisted companies was feasible and outline atentative assessment of costs, if possible. Forcountries already collecting this information,the intention was to seek ideas on how this canbe done and how costly/feasible it is.

144. Against this background, the countrieswhich volunteered to carry out these feasibilitystudies were classified into three differentcategories, on the basis of their current state ofplay:

(i) Countries currently compiling FDI data forlisted companies on the basis of bothmarket values and book values. FI, DK,FR and GR (the latter for inward FDI)pertained to this group.

(ii) Countries not currently compiling bothvaluations, but with solid plans to do so inthe near future. PT declared to be in thissituation.

57


stocks


March 2004

(iii) Countries neither compiling both values atpresent nor with concrete plans yet, butable to evaluate how feasible and costly itwould be. BE and ES volunteered toprepare a joint assessment from thisstarting point.

145. The next subsections introduce the resultsof these NFS. At the end, a global assessmentaddresses some overall conclusions on the basisof the feedback reported by the participatingcountries.

( I ) COUNTRIES CURRENTLY COMPILING FDI DATAFOR LISTED COMPANIES ON THE BASIS OF BOTHMARKET VALUES AND BOOK VALUES

DENMARK

Implementation

146. When we began compiling book value aswell as stock-exchange value (market value) forlisted companies we only had to make a fewchanges in our procedures:

– add two fields in our database, one for directinvestments in DK and one for Danish directinvestments abroad

– add two fields in our questionnaire and ourcorresponding excel-file

– adapt the changes in guidelines concerningthe questionnaire

147. In fact very small changes werenecessary. The costs were small, because weproduce an updated questionnaire, an updatedexcel-file and updated guidelines for the surveyevery year.

Practice

148. We ask the companies to provide us withinformation on the book value of FDI equitiesin all types of companies as well as on themarket value of FDI equities if the company islisted. Information on ownership share, name of

the stock exchange where the company is listedor the ISIN-code is not requested.

149. The questionnaire is sent to respondentsin March every year.

Problems

150. We do not have very detailed informationabout the listed companies and where thecompany is listed. We rely on the respondents’information.

Future plans

151. We plan to change our collection systemfrom reference year 2003 or 2004. Our plansinclude more detailed information about listedcompanies.

FINLAND152. In the annual direct investment surveys,the data on both the book value and the marketvalue of listed direct investment enterprises arecollected. The published time series for FDIposition data are still based on book values.

153. The data request for the market value wasadded to the inward and outward FDI surveysfrom the reference year 2000. The evaluation ofthe costs, related to the addition of market valuedata to the surveys, is not possible.

Inward investment

154. In the annual inward FDI survey, theresident listed direct investment enterprisesreport the total book value and the market valueof equity capital and the direct investor’sownership percentage.

155. The survey is addressed only to thedirectly foreign-owned enterprises.30 Theequity capital at book value is based on the

30 Therefore, it does not cover market value data of any possibleforeign indirectly-owned listed enterprise. Such possiblesubsequent investments are supposed to be considered by themarkets at the time of assessing the stock exchange price of thef irst-shot FDI company.


consolidated accounts of the directly foreignowned enterprise and the indirectly foreignowned enterprises are supposed to be coveredthis way. 31

156. The register of resident listed companiesmaintained by the resident stock exchangeauthorities is used to check the quality of thesurvey data. Both the information publiclyavailable on market prices of listed enterprisesand their annual reports are used to check thequality of the survey replies. With a few years’experience, respondents seem to report themarket value data with high quality.

157. As to inward investment for 2000, therewere 23 resident direct investment listedcompanies which represented 8.3 per cent of thetotal inward equity stock at book value.

Outward investment

158. In the annual outward FDI survey, theresident direct investors report the total bookvalue and the market value of the foreign listeddirect investment enterprises and the directinvestor’s ownership percentage.

159. The survey covers both directly andindirectly owned direct investment enterprisesand the data collection method gives us theopportunity to get data on the market value ofboth directly and indirectly owned listed directinvestment enterprises.

160. The respondents can provide data onequity capital by individual foreign directinvestment enterprise. They are also allowed togive consolidated sub-group replies, where oneforeign direct investment enterprise is theparent enterprise of the sub-group. If the directinvestment enterprise is listed, we insist ongetting the sub-group reply with this listedenterprise as the parent company.

161. Within these few years, the respondentshave not reported market value data with care.For the moment no appropriate quality controlmethods are available. We are dependent on

how carefully the respondents want to reply.Only the quality of the data on very largeinvestments mentioned in the financial presscan be checked.

162. As to outward investment for 2000, therewere 17 foreign direct investment listedcompanies which represented 7.9 per cent of thetotal outward equity stock at book value.

FRANCE163. FR currently collects just book valuesfrom reporters. The compiler subsequentlycalculates market values using other publiclyavailable sources. This method enables tocompile two different values for FDI equitystocks without increasing the reporting burdenweighing on respondents. It however entailssome shortcomings.

Current practices

164. The current process for compilingmarked-to-market FDI stocks is not fullyautomated yet, but should be in the future.Methods differ for inward and outward FDIstocks.

– Inward stocks

165. Data on inward stocks are compiled usingvarious databases, first to identify directinvestment companies and then to get theiraccounting data. There is no specific stocksurvey.

166. The population of resident directinvestment companies at the end of a given yearis defined as the population at the end of theprevious year, plus resident companies thathave been acquired by non-resident directinvestors during the year, minus directinvestment companies that have been sold bytheir non-resident direct investors. A databaseof resident direct investment companies,including data on shares of ownership, is

31 See previous chapter on consolidation.

59


stocks


March 2004

maintained by the Balance of PaymentsDirectorate of the Banque de France.

167. Once the whole population of residentdirect investment companies has beenidentified, balance sheet data are mostlydownloaded from an internal database onFrench companies maintained by anotherDirectorate of the Banque de France. These dataare used to compile inward direct investmentstocks at book value.

168. The distinction between listed and non-listed companies is currently made manually, byusing a security database and other publiclyavailable sources (mostly financial press). Thepossibility to use the national identificationnumber of each French company (“SIREN”) inan automated way to search for the ISIN code ofthe company’s shares (when it exists) has beeninvestigated and will be implemented in the dataprocessing system in order to be operational atthe end of 2004. When a company has beenidentified as listed, its market value is retrievedfrom the above-mentioned security database.

– Outward stocks

169. Outward direct investment stocks arecollected via an annual survey conducted byBanque de France branches, which gatherinformation on companies located within theirrespective areas. Respondents are asked toprovide us with the book value of their foreignaffiliates, following the common definition ofOFBV agreed by the WG-BOPER.

170. The distinction between listed and non-listed companies is here again made manually,on the basis of the names of the foreign directinvestment companies and using varioussources (security database, financial press orstock exchange web sites). Since this method isboth time-consuming and imperfect, thepossibility to collect information on the status(listed/non listed) of the direct investmentcompanies is currently under consideration.

171. When a company has been identified aslisted, its market value may be obtained fromthe above-mentioned sources.

Advantages / shortcomings of these methods

– Advantages

(vi) The system enables to compile both bookand market values for listed companies.

(vii) Very limited information is required fromrespondents. In fact, nothing is directlycollected from resident companies tocompile inward direct investment equitystocks.

– Shortcomings

(viii) The distinction between listed and non-listed companies is both time-consumingand imperfect. Because it is mademanually, thresholds are applied.

Future plans: possible ways of improvement

– Inward stocks

172. The process of distinguishing betweenlisted and non-listed companies will beautomated in 2004, using the link between theidentification number of French companies(“SIREN”) and the ISIN code in the securitiesdatabase.

– Outward stocks

173. The new survey (which will beoperational as of 2004) will contain a questionon whether or not the foreign direct investmentcompany is listed. Moreover, the survey willask directly two values.

GREECE174. In the annual direct investment survey (asof data corresponding to end-1997) therespondent enterprises provided informationonly on the basis of book values. GR startedcompiling both book and market values for


inward FDI on listed companies from end-2001positions.

175. The whole process is fully automated andconsists of, firstly, the identification of listedFDI companies and, subsequently, a specialcomputer program is applied taking into accountthe end of period stock exchange prices and theequity capital information provided byrespondents.

176. The set-up cost and the operational costfor calculating market values were small sincethese changes were part of a general project ofcomputerising the process of collecting andprocessing the i.i.p. data. So far there no specialproblems have been encountered in the wholeprocess.

177. As far as outward FDI data is concerned,such information is also collected through anannual survey using a business register. Therespondent enterprises report only book valuesand there is no distinction made between listedand non-listed companies but there are plans tomanually identify listed companies.

COUNTRIES NOT CURRENTLY COMPILING BOTHVALUATIONS, BUT WITH SOLID PLANS TO DO SOIN THE NEAR FUTURE

PORTUGAL178. Annual information on both book andmarket values is collected through the FDIstocks surveys. Concerning the series availablefor market valuation, no stability can be foundfor the outputs obtained since the type ofinformation requested has varied along theyears. For the time being, no control has beenmade to the answers provided.

179. Recently, in the context of the joint-workdeveloped within the Banco de Portugal for theWorking Group on Unquoted Shares (WG-US),and with a view to obtaining a first assessmentregarding the practical implementation of theSTC recommendation the TF-FDI is dealingwith, we have tried to develop a test exercise on

the answers provided under the last surveys andsome additional sources of information wereevaluated as well.

180. Assets and liabilities were assesseddifferently, provided their specifications,namely by ranking in a different way thesources of information.

Inward direct investment

181. Information on the market value and thepercentage of participation was asked in 2001,under the last inward stocks survey. Only banksand insurance companies were approached withthis aim, for data concerning 1999 and 2000.Replies to these questions were never checkedbefore, and therefore, for the time being, no usewas made of them.

182. Recently, under the test exercise made forthe WG-US, a new source of information wasadditionally tested for gathering the marketvalue of direct investment enterprises:information on quotations made available by theEuronext Lisbon (Stock Exchange).

183. As a result of the comparison exercisemade the last month, we can say that answersprovided in the survey by banks and insurancecompanies for their market value are of quitegood quality, when compared with informationprovided by the Stock Exchange.

Future plans:

– This issue was only tested once;

– Quality control on the replies to this type ofquestions in the FDI survey must beimproved;

– A methodology of production needs to bedefined;

– This type of questions still need to beextended for non-financial enterprises and re-defined for banks and insurance companies;

61


stocks


March 2004

– It requires the definition of new procedures interms of regular production;

– Supplementary sources of information shouldbe taken into account, namely news, annualreports of companies and publicly availableinformation on market prices.

Outward direct investment

184. Through the outward stocks survey, theresident direct investors report information onthe market value of their direct participationabroad, quotations, number of owned sharesand percentage of participation. Thisinformation is supplied to the Banco dePortugal since 1998. For the time being, repliesto these questions were not carefully checked,provided no supplementary information wasavailable to this end.

185. Regarding the market valuation of directinvestment enterprises located abroad, use wasalso made recently of the information availablefor some European Indexes, namely for theStoxx 600 companies and the Stoxx 50 indexcompanies, which were disseminated under thetest exercise of the WG-US. Additionally, somefurther investigation was made on theinformation available for some stock exchangemarkets of countries where there is a significantstock of Portuguese direct investment.

Future plans:

– Further investigation is needed on the wayhow to proceed;

– Accessibility to additional stocks exchangemarkets should be studied. However, effortswill be concentrated on the most importantmarkets evaluated in terms of PortugueseFDI;

– Quality control and check procedures have tobe defined;

– Supplementary sources of information shouldbe taken into account, namely news, annual

reports of companies and publicly availableinformation on market prices.

Additional comments (for both Inward andOutward)

186. The test exercise was performed on thedirect participation in equity. According to thedirectional principle, there are reverserelationships on equity, which were, in theexercise, excluded. Being howeverinsignificant, further definition is needed onthis issue.

187. The information collected through theFDI stocks surveys is based on the accounts ofthe resident direct investment company, forinward FDI, and the non-resident directinvestment company, for outward FDI, the lastbeing reported by the resident direct investor.

188. The book value of either inward oroutward FDI is calculated from the directparticipation, as collected through FDI stockssurveys. No indirect relationship is covered.Consolidated accounts are also requested in thesurveys for both outward and inward FDI, butno use is made of them.

(III) COUNTRIES NEITHER COMPILING BOTHVALUES AT PRESENT NOR WITH CONCRETE PLANSYET

BELGIUMPossibilities for the collection of the necessaryinformation

– Direct Investment in BE

In the survey it could be questioned whether anenterprise is listed and if so, where it is listed,for instance:– Is the resident company listed?

Yes No– On which stock exchange

EuronextNasdaq EuropeElsewhere namely


The market value can be calculated based on thenumber of shares (cf. CD-ROM “Data onstandardised annual accounts”) and the stockexchange value (financial papers).

– Direct Investment abroad

– Is the non-resident company listed?Yes No

– On which stock exchange?– Number of shares?– Stock exchange value on the last day of the

reference year ?

Timing

189. In April 2003, the survey is sent to collectdata related to the reference year 2002.

Problems

190. It will be difficult to check whether or nota company is listed in the case of directinvestment abroad.

Double reporting by listed companies

191. As mentioned before, the market value oflisted companies can be calculated based on theextra data that will be asked in the future. Infact, listed companies do not really have todouble report but report as in the past (just bookvalues) and deliver some extra information sothat we can calculate the market valueourselves.

SPAIN– Direct Investment in ES

192. In the absence of an FDI survey, FDIstocks in ES for non-financial sectors arecurrently being compiled by accumulatingb.o.p. flows.

Split between listed and non-listed companies:

“Other Sectors”. We plan to use the informationprovided by our future new data collectionsystem on tradable securities.

“MFIs”. We already have this informationavailable from accounting statements.

Double valuation:

“Other Sectors”. Only the market value wouldbe available.

“MFIs”. Market value and book value would beavailable.

– Direct Investment abroad

Split between listed and non-listed companies:

“Other Sectors”. We plan to use the informationprovided by our future new data collectionsystem on tradable securities.

“MFIs”. Information available from accountingstatements.

Double valuation:

“Other Sectors”. Only the market value wouldbe available.

“MFIs”. Market value and book value would beavailable.

Timing

193. For the MFI sector, the new sources ofinformation will be available next year. Theprocessing, checking and analysis of the newdata would require additional time and effort. Inthe case of the new data collection system fortradable securities, the data will not be availablebefore January 2004.

Problems

194. The new system for tradable securities willonly provide information on market values. Wehave no survey implemented. This makesalmost impossible to have information on bookvalues related to the non-financial sector ofthe economy. The evaluation of the costs and

63


stocks


March 2004

timing of implementing an FDI survey is, at themoment, not possible.

CONCLUSIONS195. The six countries which conducted theNFS may fairly represent the situation of alleuro area countries concerning the eventualcollection of data on FDI in listed companies onthe basis of two different valuation methods.Some of them already collect this information,while some others will have to introduce somechanges in their collection systems in order tocope with the need to produce the necessarydata.

196. The main lessons from the countriescurrently compiling this information can besummarised as follows:

– FR is currently the only country trying tocompile FDI stocks both at book value andmarket value without requesting two valuesfrom respondents and by using otheravailable information (security database andfinancial press). This process of compilingFDI stocks at market value is howeverimperfect and time-consuming, as it is notfully automated. For that reason, FR willmodify its collection system for outward FDIin order to collect directly the necessaryinformation. For inward FDI, the treatmentwill be automated. Both systems will beavailable at the end of 2004.

– Two countries directly collect the informationon both book and market values fromreporters, by including additional questionsin their FDI surveys. The cost of introducingsuch additional questions was not deemed toohigh (although FI could not provide a preciseassessment).

– For inward FDI, information collected fromreporters can be cross-checked with datagathered from the domestic stock exchange.

– Most difficulties are linked to theimplementation of plausibility checks to thestock exchange prices collected from

reporters for non-resident direct investmentcompanies (i.e. for outward FDI in listedcompanies), due to the lack of direct access toinformation on foreign markets’ quotations.

197. As regards the country not currentlycollecting this information but with plans to doso in the near future (namely PT), the mainconclusions could be the following:

– Annual information on both book and marketvalues can be collected through the FDI stocksurveys.

– For inward direct investment, information onthe market value and the percentage ofparticipation can be collected as part of thesurveys.

– The results can be checked with information onquotations in the domestic stock exchange.Such checkings have revealed that the answersprovided in the survey are of good quality.

– For outward direct investment, the surveymay get information on the market value ofdirect participation abroad, quotations,number of owned shares and percentage ofparticipation.

– Answers are not so easy to check due to thelack of supplementary information. The useof European Indexes as well as informationfrom additional stock exchange markets couldbe considered to this aim.

– Supplementary sources of information shouldbe taken into account, namely news, annualreports of companies and publicly availableinformation on market prices.

– Reverse relationships on equity couldconstitute a problem

198. Concerning countries neither currentlycollecting this information nor with concreteplans, the following conclusions maysummarise the outcome of the NFS conductedby the participating countries:


– The use of current information sourcesshould be promoted to the extent possible.

– The most feasible way of compiling theadditional information required could be theintroduction of additional questions to theFDI surveys. For inward FDI, this additionalinformation could be combined with datagathered from the domestic stock exchange.

– In the absence of FDI surveys, the use ofMFIs’ balance sheets may be an alternativesolution for the MFI sector’s FDI. Directinvestment by the “other sectors” would stillrequire an alternative solution, which doesnot seem straightforward without pure FDIsurveys.

From the outcome of the NFS conducted bythe three groups of countries, the TF-FDIadopted the following conclusions andrecommendations:

– The collection of FDI equity stocks for listedcompanies on the basis of two differentvaluation methods (market values and bookvalues on the basis of the common definitionof OFBV) can be deemed feasible forcountries running FDI surveys.

– For those countries, it does not imply addingtoo much to costs.

– The most feasible way to collect thisinformation would be the addition ofsupplementary questions to the FDI surveys.

– For outward FDI, where no access toquotations in foreign stock exchanges may bepossible, resident reporters should bedirectly questioned through the FDI survey.

– For those countries that, in the absence ofFDI surveys, would require collectingadditional information, the use of currentinformation sources could be promoted to theextent possible as a temporary solution untilFDI surveys may be introduced and producealternative results.

– For inward FDI, the availability of stockexchange quotations could be used as eitheran additional information source aimed atreducing respondents’ burden or to double-check the accuracy of the informationgathered from respondents.

65c ECB


4 Reinvestedearnings

INTRODUCTION

199. The TF-FDI was mandated to “…reviewthe practical aspects of the compilation methodsfor reinvested earnings (declaration byrespondents, calculation by the compilers,interim estimations etc.).”

200. The core objective, as stated in themandate, is to identify “best practices” andrecommend their implementation in order tominimise dissimilar treatments within EUMember States for balance of payments flowsas well as international investment positionstocks. Particular attention should be paid to theaccuracy of the geographical allocation.

201. As a further guidance, the ECB WG-BP&ER and the Eurostat BoPWG, at theirrespective meetings in March 2003, stated that“the Current Operating Performance Concept(COPC) should be the reference concept for thecompilation of reinvested earnings. It was alsostressed that for the practical application of thisconcept, some simplification may be needed inorder to avoid excessive costs. Additionally,the WG-BP&ER encouraged the TF-FDI tofurther examine the practicalities in compilingaggregates (in particular as regards‘extraordinary profits’ and any possibledistinction between financial and non financialcorporations).

202. Reinvested earnings (RIE) is oftenreferred to as a transaction category per se. Itis, however, a function of two variables: profitfrom current operations and dividends payable.What should be recorded in the current accountand which affects the GNI is the amount of totalprofits (both distributed and undistributed)from current operations. The calculation of RIEserves mainly one purpose for the b.o.p.,namely to estimate an offsetting entry in thefinancial account, FDI, which is necessaryfirstly to avoid net errors and omissions effectsand secondly to contribute to the reconciliationof stocks and flows. (There is of course also ananalytical interest to study the reinvested part ofthe profit in relation to the total profit.)

4 R E I N V E S T ED E A RN I NG S203. Any study of RIE will, as a consequence,focus on the two components rather than on theRIE residual. This is also the case for the workcarried out by this TF-FDI.

204. The chapter is structured as follows: itstarts by presenting international standards asregards reinvested earnings, total profits andpayable dividends as well as some clarificationsprovided by the TF-FDI to the contents of someof these concepts. The second section reviewscurrent practices so as to monitor to which extentinternational standards are applied as well as themain deviations observed. This review ofcurrent practices is carried out through thefindings of a questionnaire circulated to themembers of the TF-FDI on the characteristics oftheir current collection systems. Thereafter, anadditional section analyses the impact that theintroduction of new international accountingstandards (IAS) will most likely have in thecompilation of FDI statistics. The final sectionrecapitulates the main problems and causes ofasymmetries and presents some proposals basedon best practices.

205. Issues connected to the internationalrecommendations regarding consolidation havebeen treated in the first two chapters of thereport and are, thus, no further developed inthis chapter.

INTERNATIONAL STANDARDS CONCERNINGREINVESTED EARNINGS AND CLARIFICATIONSADOPTED BY THE TF-FDI

206. The definition of reinvested earnings canbe found in the OECD Benchmark Definition ofForeign Direct Investment (paragraph 28):– For subsidiary and associate companies: “thedirect investor’s share of the total consolidatedprofits earned by the company and itssubsidiaries and associates in the periodcovered, after allowing for tax, interest anddepreciation, less dividends due for payment tothe direct investor in the period even if thesedividends relate to profits earned in earlierperiods”.


– For branches: “[…] total profits earned by thebranch in the period covered, after allowing fortax, interest and depreciation, less earningsremitted in the period even if these relate toprofits earned in earlier periods”.

207. The Balance of Payments Manual (5thedition) of the IMF also provides a definition(§ 278):“Reinvested earnings comprise directinvestors’ shares – in proportion to equity held– of (i) earnings that foreign subsidiaries andassociated enterprises do not distribute asdividends and (ii) earnings that branches andother unincorporated enterprises do not remit todirect investors. (If that part of earnings is notidentified, all branch earnings are considered,by convention, to be distributed.) Thus,reinvested earnings may be calculated as theentrepreneurial income (net operating surplus)of the direct investment enterprise, plus anyincome or current transfers receivable, minusany income or current transfers payable. Thelatter include any current taxes payable onincome, wealth, etc.”

208. These definitions are consistent withESA95/SNA93. Appendix 1 contains asummary of the treatment of reinvested earningsin national accounts and data about the impactof net reinvested earnings on Member StatesGNI32.

209. Thus, for subsidiaries and associates,reinvested earnings are calculated as thedifference between total profits of the directinvestment enterprise and dividends distributedby the direct investment enterprise. Thetreatment of profits and dividends in FDIstatistics is developed hereafter.

210. Profits: the BPM5 and the Benchmarkrecommend the use of the Current OperatingPerformance Concept (COPC) to measure theearnings of direct investment enterprises.According to the BPM5, the COPC includesincome from normal operations of the enterprise(net of depreciation allowances and of othertransfers) and does not include any realised or

unrealised holding gains or losses arising fromvaluation changes and write-offs33.

211. Counter to the COPC, the so-called all-inclusive concept includes in the calculation ofincome all items (therefore also capital gains/losses and write-offs) causing any increase ordecrease in the shareholders’ or investors’interests during the period, other than dividendsand any other transactions between theenterprise and its shareholders or investor (BD,§31).

32 Neither BPM5 nor SNA 93 provide guidance on the application/non-application of the accruals concept with respect to directinvestment income. In particular, if a direct investmententerprise is sold at the beginning of the year (or acquired at theend of the year), this is normally not taken into account in thecalculation of total reinvested earnings, which are based onannual results and do not take into account how many months theenterprise was part of the total FDI stock. This may be a sourceof asymmetries, in particular vis-à-vis the United States whichdoes apply this principle for income associated with acquisitions/disposals recorded during the period under review.

33 See BPM5 §285 for a long list of examples: “Direct investmentearnings are measured on the basis of current operatingperformance. Operational earnings represent income fromnormal operations of the enterprise and do not include anyrealized or unrealized holding (capital) gains or losses arisingfrom valuation changes, such as inventory write-offs; gains orlosses on plant and equipment from the closure of part or all of abusiness; write-offs of intangibles, including goodwill, becauseof unusual events or developments during the period; write-offsof research and development expenditures; losses on the write-offs of bad debts or on expropriation without compensation;abnormal provisions for losses on long-term contracts; andexchange-rate-related gains and losses. Unrealized gains orlosses resulting from the revaluation of fixed assets, investments,and liabilities and any realized gains or losses resulting from thedisposal of assets or liabilities should be excluded from directinvestment earnings; that is, gains should not be added in andlosses should not be deducted. In addition, valuation changesresulting from unforeseen obsolescence, catastrophes, anddepletion of natural resources are treated as holding losses at thetimes that the decreases in values actually occur. Because datafor many countries are available only on an all-inclusive basis,when holding gains and losses and other extraordinary incomeare included in reported earnings, those countries that reportearnings on either an operating basis or all-inclusive basis shouldcollect and publish supplementary information on holding gainsand losses and other extraordinary items. This practice wouldenhance international comparability for both flows and stockpositions.”

67c ECB



212. Turning to the COPC, whereas theexclusion of holding gains and losses is clearlystated in the methodology34, the reference to„income from normal operations“ made in §285of the BPM5 may lead to ask the question if it ispossible to identify categories of income (asdistinct from holding gains) that should beexcluded from the COPC because they do notrefer to the normal operations of the enterprise.

213. The TF-FDI discussed the case of financialincome (interest and dividends being the maincategories) for a manufacturing firm. The BoPtextbook clarifies this point in §411, where itstates that reinvested earnings equal Operatingprofits (operating revenue minus operatingexpenses) plus current transfer receivable,interest receivable, dividends receivable and theenterprise’s share of reinvested earnings of anysubsidiary or associated enterprises, minustaxes due for payment, other current transferpayable, interest payable and dividendspayable35. Therefore financial income isincluded as a component of the COPC and theconcept of “income from normal operations”should not be equated with concepts such as“operating profits” or “operating surplus”.

214. The case of financial institutions was alsodiscussed. In particular, it was asked whether itis possible to isolate holding gains and losseson financial assets. Most of the members of theTF-FDI considered that when these gains andlosses are regarded as normal activity by thereporting financial company it is very difficultin practice to obtain separate information fromthe respondent. In addition, in the case of MFIs,the separate distinction of holding gains andlosses may prove extremely difficult in practice.

215. The TF-FDI also discussed the treatmentof goodwill, particularly as concerns write-offand depreciation of existing goodwill. In §285,the BPM5 explicitly lists “write-offs ofintangibles, including goodwill, because ofunusual events or developments during theperiod” among the cases that should beexcluded from the COPC because theyrepresent holding losses. Concerning the yearly

allowances for depreciation of goodwill, §286of the BPM5 refers to depreciation, withoutdistinction between tangibles and intangiblesassets, as an element that should be included inthe calculation of the COPC.

216. Dividends: according to BPM5 (paragraph287) dividends payable to the non-residentshareholders (direct investors and portfolioinvestors) must be recorded gross of anywithholding taxes. In practice, it is often thecompany itself that pays the taxes to the taxauthorities of the country in which it operatesand, subsequently, distributes the dividends tothe non-resident shareholders net of tax. Insuch cases a correction should be ideally madeso that dividends be considered as being paid infull to the non-resident shareholders byimputing a counter-entry (for the amount of tax)as a current transfer (credit). Likewise, acorrection should be made for dividendsreceivable.

217. Time of recording: dividends should berecorded as of the date they are declaredpayable. According to BPM5, dividends are tobe recorded “as of the date they are payable.”This recommendation could be misunderstoodto mean the dates on which dividends are “duefor payment.” This issue is addressed in the

34 Besides the paragraph quoted in the previous footnote, see also§409 of the BoP Textbook: “Capital gains and losses do notconstitute income and therefore are not included in thecalculation of enterprise earnings. Examples of capital gains arethe sudden discovery of natural resources; the revaluation offixed assets; and increases, which are due to changes inexchange rates or to higher stock exchange quotations for theseassets, in the market value of financial assets. (…)”. The generaldefinition of holding gains can be found in the SNA93 (3.62)“Positive or negative nominal holding gains may accrue duringthe accounting period to the owners of financial and non-financial assets and liabilities as a result of a change in theirprices. (…) Holding gains may accrue on assets held for anylength of time during the accounting period, not only on assetsheld at the beginning or end of the period”. The BD also specifiesthat the exclusion of all classes of realised and unrealised gains/losses applies also to banks and financial intermediaries (§33).

35 It is understood that interest paid and received from affiliatedenterprises (direct investor or direct investment enterpriseaccording to the direction) are recorded separately in thebalance of payments and should not be double counted whenmeasuring reinvested earnings (see also § 410 of the BoPTextbook)


IMF Balance of Payments Textbook which usesthe terms “declared payable”36, in accordancewith the accrual principle for the time ofrecording dividends. However, the BD (§28)recommends that dividends should be recordedas they are “due for payment”. The issue ofdiffering recommendations of the BPM5 andBD on the time of recording dividends wasaddressed in earlier discussions of the OECDWPFS where it was agreed to align the conceptswith the recommendations of BPM5, andtherefore that dividends should be recorded onthe date they are declared payable by thecompany distributing them.

218. However, in practice many countriesdeviate from this standard and record dividendsat the time they are paid. Point 16 of the lastrelease of the summary results of the IMF/OECD SIMSDI can be quoted here: “TheSIMSDI survey questionnaire focused ondetermining the number of countries that wereusing a paid (cash) basis for measuringdividends, as compared with the‘payable’ basis. The results indicate that themajority of OECD countries record dividendsas of the date they are paid. This departure fromthe accrual principle is largely a result of theprimary data source used for recording FDIincome data – a large number of these OECDcountries rely on an international transactionsreporting system for recording dividends flowsand, as this is basically a settlement-basedreporting system, it mostly providesinformation on a cash basis. Largely for thissame reason, about half of the OECD countriesrecord interest on a paid basis, instead of as it isaccrued”.

219. Provision of funds to cover losses:Another aspect discussed by the TF-FDI inrelation with dividends was the treatment offunds provided by parent companies to theiraffiliates to cover losses that some countriesrecord as negative dividends. The prevalentopinion was that these transactions should betreated in the financial account (under FDI/equity capital) rather than in the current account(either as negative dividends or in a separate

item). Losses are supposed to have had anearlier impact in the current account (as well asin the financial account and in the value ofequity stocks) as long as reinvested earnings(on a COPC basis) are calculated as a net figurebetween total ordinary profits (net of losses)and dividends paid. The provision of money tocover those losses makes the value of the directinvestment company return to its originalmagnitude, thus offsetting the temporary loss ofvalue resulting from the losses registered by thecompany. Therefore, the recording as a newdirect investment transaction seemsappropriate. The reference here is §374 of theBPM5.

220. Exceptional dividends: the TF-FDI alsodiscussed the treatment of dividends paid bymultinationals as a consequence of exceptionalcapital gains. Reference was made to the case ofliquidating dividends . According to BPM5§290, “liquidating dividends are excluded frominvestment income because such dividendsrepresent returns of capital contributions ratherthan remittance (distribution) of earnings.Therefore, liquidating dividends should berecorded in the financial account aswithdrawals of capital.”

221. In some countries, the treatment adoptedfor liquidating dividends is extended to somecases of dividends received and paid by largemultinationals. Such dividends are paid out asan extra bonus to shareholders and stem fromimmense capital gains after the sale orliquidation of an affiliate. These dividends arenot related to any form of operational incomeand are therefore not included in FDI income.Since the capital gain at the origin of thedividend is not recorded under total profitbecause of the application of the COPC, there isno impact on reinvested earnings.

36 Refer to paragraphs 397 and 406 of the IMF Balance of PaymentsTextbook, 1996.

69c ECB



222. Support for the treatment of dividends asdescribed above can be found in IMF’sGovernment Finance Statistics Manual (GFS),§5.87:

When payments are received from publiccorporations, it can be difficult to decidewhether they are dividends or withdrawals ofequity. Dividends are payments a corporationmakes out of its ongoing productive activities.A corporation may, however, smooth thedividends it pays from one period to the next sothat in some periods it pays more in dividendsthan it earns from its productive activities. Suchpayments are still dividends. Distributions bycorporations to shareholders of proceeds fromprivatization receipts and other sales of assetsand large and exceptional one-off paymentsbased on accumulated reserves or holding gainsare withdrawals of equity rather than dividends.

223. Further support for this treatment is alsofound in ESA 95 Manual on GovernmentDeficit and Debt (4.b The Notion of Dividend).

224. Other participants in the TF-FDImentioned some counter-arguments to thistreatment. In the first place, the identification ofexceptional dividends in the sense definedabove may not be easy and may contain someelements of arbitrariness. Secondly, forcountries in which these dividends are receivedrather than paid, it is likely to be impossible todistinguish the various cases and so thesedividends would be recorded in the currentaccount. This would give rise to anasymmetrical treatment of this type of dividendsbetween the two countries involved.

CURRENT COLLECTION AND COMPILATIONMETHODS FOR REINVESTED EARNINGS IN THEEUROPEAN UNION

225. The sub-group on reinvested earningsdesigned a questionnaire circulated to themembers of the TF-FDI as part of a fact-findingexercise on methods and practices used,problems encountered and plans for the future.

The main answers to the questionnaire aresummarised in the summary table next page.

226. The replies to the TF questionnaire onreinvested earnings indicated the followingsituation: FI, IE, the NL, SE and the UK applythe COPC, while AT, DK, FR, DE, IT and PTapply the all-inclusive concept. MS applyingthe COPC reported to base their calculation onsurveys conducted on the enterprise and onpossible contacts with the respondent aiming atclarifying the nature of particular entries. MSapplying this approach mainly concentrate onthe most important reporters. These companieshave the largest effect on extraordinary itemsand are, therefore, mostly approachedindividually. Smaller companies can also beapproached though.

227. BE and ES do not publish reinvestedearnings from collected data and therefore thequestions on the concept used for total profits atthe enterprise level did not apply. RIE for BE(BLEU up to 2002) have been estimated for2000 and 2001. ES covers RIE only for MFIsand is planning to estimate the remaining part.


Table 6 Questionnaire on reinvested earnings – summary table

Concept Relations Collection Collection Geo Revisionsof profits of break-

dividends down

Include Different Opera- Direct Census ITRS Available Final PlansRIE calculations tional (O) Indirect Sample Surveys, Estimated data for

inward/ All- Full direct according 2002outward inclusive (A) estimation reporting to stocks ready in:

Estimation Limited

Austria yes no A Direct Census ITRS Available 2004 June Improve timeliness

Belgium yes yes O (inward) Direct Estimate Estimate Limited 2003 Sep. Enlargement surveyA (outward)

Denmark yes no A Direct Census Survey/DR Available 2004 More detailed dataon profits/lossesand paid dividends

Finland yes no O Direct, Sample Survey/DR Available 2003 Sep.indirect

France yes no A Direct Census ITRS Estimated 2004

Germany yes no A Direct Census ITRS Available 2004 Feb.2005 Feb.

Italy yes no A Direct Sample ITRS Estimated 2003 May

Netherlands yes no O Direct, Census Survey/DR Available 2004 April 2003 directindirect reporting system

Portugal yes no A Direct Sample/cen ITRS Estimated 2004 Improve timelinessand reducefrequency of surveys

Spain no

Sweden yes no O Direct, Sample Survey/DR Estimated 2004 Jan. Dividends askedindirect in annual survey

United yes no O Direct Sample Survey/DR Estimated 2004 Jan.Kingdom

Greece no (until no A Direct Census ITRS Available Q1 2004 Calculate data for2002) 2002 and use of

COPC for 2003 data

Total profit

228. Most of the MS including RIE in theBalance of Payments calculate these data on thebasis of annual surveys on the (annual) totalprofit combined with short-periodic datacollection of (e.g. monthly) dividends. BE,however, uses estimation techniques withoutthe support of data collected.

229. The annual surveys are carried out eitherwith a census approach (with a cut-offthreshold applied) or as sample surveys.

71c ECB



230. BE bases the estimation of profits onbalance sheet data and structural businesssurvey for inward FDI and applies a ratio forthe estimation of outward.

231. FDI by individuals are included – at leastpartially/in principle – by AT, DE, IT, the NL,PT and SE.

DIVIDENDS232. Most countries collect the data ondividends on a monthly/quarterly basis, mostlywhen paid as a result of the settlement systemsused.

233. The SIMSDI summary table (Table 7)below shows that all EU MS record dividends atthe time they are paid, except SE and IE. In theTF questionnaire, FI specified that in the caseof “huge dividends” these are recorded at thetime they are declared payable.

Country Cut-off Direct/indirect relations

Austria EUR 72,000 – shareholding DirectDenmark DirectGermany mEUR 3 – balance sheet total DirectFrance DirectNetherlands Direct and indirectPortugal For non-financial enterprises, 90% of previous stock has to Direct

be covered, new investments since then are added

Table 7 Census/cut-off

Country Sampling basis Direct/indirect relations

Finland Size inward/outward stock, main economic activity Direct and indirectItaly Volume of assets, geographical area, economic sector DirectSweden Size inward/outward stock, main economic activity Direct and indirectUnited Kingdom Direct

Table 8 Sample surveys


Dividends as of the date they are:

Countries that record: Payable Paid Other

Austria no yes noBelgium no yes noDenmark no yes noFinland no yes noFrance no yes noGermany no yes noGreece no yes noIreland yes no noItaly no yes noLuxembourg 2) no yes noNetherlands no yes noPortugal no yes noSpain no yes noSweden yes no noUnited Kingdom no1) yes noTotal EU15 2yes 13yes -Total OECD (30)

Yes 6 24 1No 24 6 29

Response not available 0 0 0

1) “Yes” for distributed branch prof its of unincorporated enterprises.2) Separate annual transaction data on LU can be derived from the survey on FDI positions. Monthly transactions statistics are commonfor BE and LU.

Table 9 Time of recording of FDI income transactions

234. Two countries, AT and DE, include in FDIIncome contributions from parent companies tocover losses in their subsidiaries. AT recordsthese transactions as negative dividends,whereas DE regards them as a separate sub-itemof FDI Income. The majority of the MS treatsuch contributions as financial transactionssimilar to shareholders’ contributions.

THE CURRENT OPERATING PERFORMANCE (COPC)AND THE ALL-INCLUSIVE CONCEPT235. In compliance with the internationalstandards, FI, The NL, IE, SE, and the UKapply the COPC, whereas the other MS use theall-inclusive valuation of total profit. Thereason why the majority of the MS use the all-inclusive concept is partly that it is easy tocollect, simply asking for the net result of thefinancial year. In order to calculate the profitaccording to COPC, adjustments will have to bemade for write-offs, capital gains and losses,etc. Even though respondents may have thesedata available, most MS have chosen not to

collect them, mainly in order to limit the datacollection.

236. Furthermore, as a result of a request of theBoP WG, the TF-FDI discussed in more depthsome practical aspects linked to the problem ofidentifying items for holding gains and losses(as defined in the statistical methodology)within the accounts of the enterprises includedin the surveys. Some of the Member States thatadopt the COPC explained that thisidentification often requires a supplement ofwork with the respondent through directcontact. As a result, it is often the case that thisdeeper analysis is limited to the most importantrespondents.

237. The NL supplied more information on theconcepts used for obtaining the COPC from thesurvey used for data collection. In theexplanatory notes to the survey there is theindication that companies should report theirprofits excluding extraordinary items and

73c ECB



before profit appropriation. Extraordinary itemsin the Dutch guidelines for the annual reportsare profits/gains or losses which do not arisefrom normal operations. This means that, forinstance, what is regarded by the reporter as“normal” write-offs of fixed assets such asmachinery or plants is included as a deductionin the COPC. In some practical cases thedistinction between amortisation of fixed assetsand “normal” write-offs of these assets maybecome blurred.

Forecasts

238. In order to calculate RIE for the currentyear, most countries use forecasting methods toestimate the total profit (the other component ofRIE, dividends, is collected on a current basis),extending from rough estimates to modellingbased on different indicators.

239. AT, for example, uses profits extrapolatedcompany by company combined withinformation on GDP and exchange ratedevelopment. FR applies an exogenous growthrates hypothesis, whereas FI and SE base theirestimates on figures for the previous year,development of the economy, and (SE)exchange rates. DK applies an average ratio forinward/outward FDI, respectively, betweenprofit/loss and total equity. PT uses GDP trendsin PT for inward FDI and in the most importantrecipient countries for outward FDI.

Geographical breakdown

240. Depending on data collection methods, thegeographical breakdown is either collected orestimated on the basis of stock data.

241. Countries collecting geographicallydistributed data are AT, DE, DK, FI, and theNL. Estimations on the basis of stock data aremade by FR, IT, SE, PT and UK. BE has only alimited breakdown (intra/extra Eurozone).

242. The data on total profit (and RIE) iscollected on a consolidated basis by the UK, FI,IE, the NL, and SE. All countries except SE

allocate the data to the first directly ownedcompany, whereas SE makes a distribution onall levels of the chain of indirectly ownedenterprises.

Timeliness

243. Final RIE data are available only aftercompletion of the annual surveys, the time forwhich differs considerably between MS. Thepresent situation is seen below:

Country Date

Austria June 2004Belgium September 2003Denmark 2004Germany February 2004, revision in February 2005Finland September 2003France 2004Italy May 2003Netherlands 2004Portugal 2004Sweden January 2004United Kingdom January 2004

Table 10 Avai labi l i ty of f inal data onreinvested earnings (to end-2002)

INTERNATIONAL ACCOUNTING STANDARDS (IAS)

244. A preliminary discussion of the principlescontained in the new International AccountingStandards on the notion and recording ofincome was also held.

245. The main result was that the new IASrefers to a broader concept of income than theone adopted in the statistical manuals discussedabove. In particular, the IASB Framework37

clarifies that income encompasses “revenue andgains” (expenses and losses) and that the twocomponents are not distinguished in the IASgeneral framework. The relevant paragraphsfrom the Framework (§§74-76) are quotedbelow:

37 References were taken from the publication “InternationalFinancial Reporting Standards – incorporating InternationalAccounting Standards and Interpretations”, InternationalAccounting Standards Board, January 2003.


74 The definition of income encompasses bothrevenue and gains. Revenue arises in thecourse of the ordinary activities of anenterprise and is referred to by a variety ofdifferent names including sales, fees,interest, dividends, royalties and rent.

75 Gains represent other items that meet thedefinition of income and may, or may not,arise in the course of the ordinary activitiesof an enterprise. Gains represent increasesin economic benefits and as such are nodifferent in nature from revenue. Hence,they are not regarded as constituting aseparate element in this Framework.

76 Gains include, for example, those arisingon the disposal of non-current assets. Thedefinition of income also includesunrealised gains; for example, thosearising on the revaluation of marketablesecurities and those resulting fromincreases in the carrying amount of longterm assets. When gains are recognised inthe income statement, they are usuallydisplayed separately because knowledge ofthem is useful for the purpose of makingeconomic decisions. Gains are often reportednet of related expenses.

246. Therefore the application of the new IASby companies, even on a compulsory basis,would not imply that holding gains and lossesbe isolated in the standard presentation. Theanalysis of IAS n° 8 (on Profit and Losses)confirmed this conclusion. The TF-FDI alsoexamined an “improvement project”38 beingdeveloped in the framework of IAS and aimingat distinguishing various standard componentsof income. The TF-FDI considered that it wouldbe very useful for data compilers to implementthe reporting of these components by companiesin a harmonised way. The project is howeverstill at a preliminary stage of discussion in theIAS framework. The subject could be usefullystudied further in the framework of theEurostat/ECB “Accounting and statistics”TF-FDI.


247. It appears that a significant part of MS findit difficult to adapt to the internationaldefinitions as stated in BPM5 and OECD’sBenchmark Definition and consider a fulladherence to these principles to be hard toattain, at least in the short run. MS’ differentpractices naturally lead to inconsistencies on anEU/Euro area level. The problems of applyingthe standards in practice depend partly ondifficulties to obtain the necessary data, partlyon the data collection methods used and partlyon the information internally available torespondents.

PROFITS248. A significant part of MS do not apply theCurrent Operating Performance Concept; onlyfive countries have so far implemented theCOPC, namely FI, The NL, IE, SE, and the UK.Asymmetries caused by different practices willmainly affect the total Income on FDI Equity inthe Current Account. The different practicesused affect the total Income in the CurrentAccount and thus the calculation of the GNI forthe MS, causing asymmetries on the EU/euroarea aggregates.

249. In order not to deviate from internationalstandards, the same concept for the calculationof profits, namely the COPC, should be used byall MS. The TF discussed the problemsconnected with a change from the all-inclusiveto the COPC and the practices applied by theMS whose compilation practices alreadycomply with the standards.

250. According to the experiences of somecountries like the UK and DE, a reduced numberof domestic companies (in some cases, holdingcompanies39) involved in FDI transactions/

38 See http://www.iasc.org.uk/cmt/0001.asp?s=9198007&sc={61FA8D2B-B229-4738-95E6-B3FEE9D05804}&n=66. Theproject examined by the TF-FDI is “The Income Statement(Reporting Performance)”.

39 Such holding companies normally present very high profits/losses in the financial year compared with a relatively lowvolume of equity capital.

75c ECB



positions contribute to most of the totalextraordinary gains/losses in their countries.For the other companies, the all-inclusiveapproach often provides similar results to theCOPC.

251. The TF-FDI concluded that acceptablesolutions for the application of the COPCshould aim at covering at least the reducednumber of companies which contribute the mostto extraordinary results (namely the biggestrespondents and holding companies).

252. Two information sources were consideredfor the application of the COPC: (i) companies’public accounts; (ii) restricted information,internally available to the companies. Althoughthe split between ordinary and extraordinarygains/losses in accounting statements is notnecessarily consistent with statisticaldefinitions, it was considered by the TF to be animmediately available proxy. For the firstsolution (public annual accounts) to beconsidered as an acceptable proxy for theCOPC, additional information (internallyavailable to respondents) would be required(notably, the geographical breakdown).Therefore, a combination of both informationsources would be necessary in any case.

253. One aspect to be considered is the impacton the data collection from the introduction ofthe new IAS, which could be expected to affecta larger number of enterprises than wasoriginally expected. In the present draft versionof the IAS not all components necessary for aCOPC valuation are specified in the standardpresentation . Within the framework of IAS astudy has, however, been initiated aiming atdistinguishing various standard components ofincome.

254. The development of the new IAS willimply a more specific definition of thecomponents which may serve as a firm basis forthe harmonisation of MS’ application of theCOPC. However, the development of the newIAS may pose an additional difficulty forcompilers to properly apply the COPC, to the

extent that only very exceptional results will beexcluded from the ordinary profits and losses.

DIVIDENDS255. Collection systems are determining theapproach for dividends paid vs. payable. Todayonly SE, IE and FI collect data on dividendswhen payable. Countries using settlementsystems as the main source for dividends willnaturally record the transactions when paid. Itmay, however, be foreseen that the increaseduse of surveys (monthly or quarterly) will bringthe practices closer to the internationalstandards, as these surveys are likely to reflectaccounting data, based on accruals. Applicationof different practices will in this case mainlylead to inconsistent treatments due todifferences in time allocation.

256. Treatment of dividends stemming fromexceptional capital gains is a problem in so faras it affects the calculation of reinvestedearnings. The BoP effect of such dividends isseen in the FDI item of the Financial Accountand the subcomponents of Income; total profitis not affected by these transactions. The NLhave chosen to treat these dividends as financialtransactions (disinvestments) rather thanincome, as is the practice by other MS. Byrecording exceptional dividends in the FinancialAccount the same concept (current operations)will be used for both total profit and dividends.The other MS using the COPC for profits,record the total dividends (all-inclusive) asincome, which causes some confusionregarding the concept of reinvested earningsdata calculated on this basis. Therefore, the TFconcludes that dividends stemming fromexceptional capital gains should be recorded asFDI disinvestments. Even though suchdividends are easy to recognise, bilateralcontacts are recommended in the case of intraeuro area/EU flows so as to avoid asymmetries.

GEOGRAPHICAL BREAKDOWN257. Inconsistent treatments due to dissimilargeographical attribution may also be caused bythe different collection systems used. A census-type survey (with no grossing-up) will, for


example, in most cases “allow” a more detailedbreakdown on countries than a sample survey.In most cases stratification and sampling forFDI surveys is made on the basis of stocks andeconomic activity and thus not on geographicalvariables. In order to produce data with areliable and detailed country breakdown thesamples would have to be increased, whichleads to higher costs.

258. Also, the practices used by some MS toestimate the geographical allocation of profitson the basis of stocks will cause asymmetriesvis-à-vis countries collecting the data with acountry breakdown.

PROPOSALS

REINVESTED EARNINGS259. A fundamental need to reduce asymmetriesregarding reinvested earnings within the EU isthat all MS compile RIE, and they base theircompilation methods on similar concepts, whichis not the case at present. Not all MS have yetestablished a system for this purpose. Since thisis of basic importance, the TF-FDI proposesthat annual FDI surveys to collect the necessarydata be implemented in all MS as soon aspossible (in line with the proposal made for FDIequity stocks). This is the first priority.

TOTAL PROFITS260. In line with the clarified Mandate of the TFand international recommendations the TF-FDIproposes that the Current OperatingPerformance Concept (COPC) be used by allMS for the valuation of total FDI profit, i.e.exceptional results should be appropriatelyexcluded from the current account.

261. Furthermore, there is a need to follow thedevelopment of the new IAS framework asregards the income components, since aligningstatistical principles with the new accountingstandards to the largest extent possible seems tobe the most cost-effective (and perhaps the onlypossible) approach vis-a-vis respondents. Amore general statement concerning the close

monitoring of forthcoming changes in statisticalstandards and the development of the new IASis suggested as a follow-up action point inchapter 8.

COVERAGE OF INDIRECT RELATIONS262. The TF-FDI proposes that the coverage ofreinvested earnings be made in accordance withone of the simplification procedures addressedin chapter 2, namely (i) coverage of indirectlinks of ownership above 50%; or (ii) coverageof direct and indirect ownership links above10%, calculated as the product of thesubsequent ownership links along a chain.

DIVIDENDS

PAID/PAYABLE263. Since MS may increasingly collect databased on accounting, a “natural” transition mayalso take place to collect dividends “whenpayable”. This foreseeable changeover will,however, not be made in the short term. On theother hand, asymmetries will only occur inshort-periodic statistics, since the differencebetween payable and paid is usually only amatter of time allocation during a fairly limitedperiod. We suggest that MS address the issue inthe development of their future systems so as tobring their practices nearer to the internationalrecommendations.

CONTRIBUTIONS TO COVER LOSSES264. As to contributions to cover losses indirect investment enterprises, the TF-FDIproposes that these transactions, in line withinternational recommendations, be recorded inthe Financial Account, as additional investmentflows and not as direct investment income.

DIVIDENDS FROM EXCEPTIONAL CAPITAL GAINS265. The TF-FDI recommends that dividendsstemming from exceptional capital gains berecorded as FDI disinvestments rather thanunder Income in the Current Account. Such atreatment is supported by internationalrecommendations and will ensure a commonconcept to be used for the components ofreinvested earnings.

77c ECB


5 Issues relatedto other

capital in FDIstatist ics

INTRODUCTION

266. This report investigates the theory andcurrent practices regarding “Other Capital” inforeign direct investment flows and stocks ofMember States of the European Union asmandated in the Mandate of the Task Force onForeign Direct Investment (TF-FDI). Bycombining theory and current practices, bestpractices are being put forward for the problemslinked to Other Capital in foreign directinvestment (FDI) flows and stocks.

267. The report will focus on the directionalprinciple, the identification and impact of thecomponents of Other Capital in FDI, and thetreatment and classification of transactions/positions on Other Capital when Monetary andFinancial Institutions (MFIs) are involved. Thelatter mainly concentrates on the case ofsubordinated loans. In addition, the treatmentand classification of Other Capital when SPEsand other financial intermediaries (not MFIs)are involved and the valuation principlesconcerned with Other Capital are explored.

268. The report is constructed in the followingway: it starts with a description of the currentpractices in European Union Member States onthe treatment of Other Capital in FDI flows andstocks.40 It continues with the establishment ofsome conclusions drawn from the previousanalysis and with the identification of bestpractices. The chapter concludes with a shortrevision of some practical aspects related to theapplication of the directional principle,including a table summarising the magnitude ofthe financing by affiliates to parent companies(for both FDI flows and stocks) using theinformation available in Member States toassess the importance of a correct application ofthe directional principle. All definitional issuesregarding Other Capital in FDI, as contained ininternational standards, are described inAnnex 2, at the end of this report.

5 I S S U E S R E L AT ED TO OTH ER C A P I TA L I N F D IS TAT I S T I C S

CURRENT PRACTICES IN THE TREATMENT OFOTHER CAPITAL IN THE EU MEMBER STATES

269. The current practices of the treatment ofFDI-Other Capital in EU Member States areidentified by the use of the documents onCurrent Practices, submitted by Member Statesand circulated at the first meeting of the TF-FDIin April 2002 and by the use of a questionnaire,conducted in August. The questionnaire wascirculated to the participants of the TF-FDIbecause not all necessary and relevantinformation for this report was included in thedocuments submitted initially.

DESCRIPTION OF THE QUESTIONNAIRE270. The questionnaire was designedspecifically to address the issues raised in theTask Force Mandate relating to Other Capitaland was conducted in August 2002. Thisincludes the various forms taken by OtherCapital, the items included and the specialcircumstances of MFIs and SPEs. Theopportunity was also taken to inquire intomethods and use of the directional principle andvaluation methods as applied to Other Capital.

271. In November 2002 an other questionnairewas sent to the members of the Task Forcewhich apply the directional principle. Thisquestionnaire investigated the practical aspectsof the application within the institutions.

RESULTS FROM THE QUESTIONNAIRE COMBINEDWITH THE SUBMITTED DOCUMENTS272. 12 countries (out of 13) returnedcompleted questionnaires (IE also providedreplies at a later stage) and the results aresummarised below. No details are known forthe two remaining countries (LU and GR).

40 Current practices are also compared with the theoretical aspectscontained in the appendix to this chapter.


A1. Do you apply the directional principle? Number of Member States

Yes or partially 12No 1

Table 11 Application of the directional principle

273. Those Member States that declare theyapply the directional principle do so for bothflows and stocks in both inward and outwarddirections, although, further investigationsrevealed that most of them only partially do it,i.e. not for all elements of Other Capital.

274. The table summarising the answers to thequestionnaire, to be found in Annex 3 at the endof the report, shows that the coverage of thevarious items included in other capital differsamong countries. For instance, five countriesonly can include transactions/positions on debtsecurities between affiliates in other capital.

275. In most cases, special internal codes are usedto identify and process Other Capital transactionsand positions involving the directional principle.The process is generally automated in the MemberStates, with some having additional manualprocesses for quality checks mainly. MostMember States can only apply the directionalprinciple for direct relationships.

276. One Member State (BE) which does notapply the directional principle yet, will do thiswhen the new system is introduced. Theprocess, described by BE in the questionnaire,has many similarities with the overallapplication of the directional principle in theother Member States.

B1. Which items are included in Other Capital of regular companies? Number of Member States

Debt securities 6Trade credits 10Financial leasing/Leasing credits 11Financial derivatives 1Assets in intra-group accounts (intercompany loans) 13Transferable deposits 9Permanent debt 10Other 0

Table 12 Items included in Other Capital

277. Annex 5 (at the end of the report) presentsa synthetic table showing empirical evidence ofthe relative importance of the variouscomponents of other capital.

DEBT SECURITIES278. Six Member States include debt securitiesin FDI Other Capital. Five (IT, PT, SE, UK, IE)include both flows and stocks for inward andoutward directions. Three (IT, SE, IE) of thesealso collect long and short term data. FI collectsstock data for both directions and periodicitiesbut find that the reported values areinsignificant.

TRADE CREDITS279. Nine Member States include trade creditsin FDI Other Capital for both flows and stocksand both directions (DE, ES, FI, IT, NL, PT,SE, UK, IE). DK records trade credits as OtherCapital only for stocks.

FINANCIAL LEASING/LEASING CREDITS280. Eleven Member States include financialleasing in FDI-Other Capital. Nine (AT, DE,ES, FI, IT, PT, SE, UK, IE) include it in bothflows and stocks and in inward as well asoutward. BE only collects flows and DK onlystocks. No distinction is made between long-term and short-term financial leasing. In thenew reporting system (from April 2003

79c ECB




onwards), NL will also include financial leasingin their FDI figures (both stocks and flows).

FINANCIAL DERIVATIVES281. NL includes financial derivatives in OtherCapital but will be discontinuing this collectionafter the new collection system for directreporting is introduced (from April 2003onwards). and DK includes only stocks offinancial derivatives.

DEPOSITS282. Deposits are included in Other Capital bynine Member States (BE, DK, FI, FR, IT, NL,SE, UK, IE) for both flows and stocks in bothdirections.

INTERCOMPANY LOANS283. Intercompany loans (assets in intra-groupaccounts) are included in other capital by allthirteen responding Member States. All thirteenMember States include loans between affiliatedcompanies in both flow and stock data and ininward as well as outward data. Six MemberStates (DE, DK, FR, NL, SE, IE) report thatthey make a distinction between long-term (over1 year) and short-term (up to 1 year) assets andliabilities.

PERMANENT DEBT284. Ten Member States (AT, BE, DE, DK,ES41, FI, IT, PT, UK, IE) include permanentdebt in Other Capital.

285. Seven Member States (AT, BE, DK, FI,PT, UK, IE) have indicated that subordinatedloans between MFIs are included in OtherCapital. BE only includes these loans in flows;UK in stocks and the other five include them inboth flows and stocks. FR and DE include suchloans in equity capital rather than in othercapital, for reasons of confidentiality in the caseof DE.

286. One Member State (DK) includes flowsand stocks, such as deposits and other claimsand liabilities, between MFIs in Other Capital.

THE INCLUSION OF DATA ON SPES AND OTHERFINANCIAL INTERMEDIARIES287. Twelve Member States have indicated thatthey include data on SPEs and other financialintermediaries in Other Capital flows andstocks for both directions.42 Valuation methodsgenerally correlated with the overall valuationmethods given in the next paragraph. A numberof Member States reported that they didn’tspecifically identify SPEs and that suchenterprises were treated as any other companyfor FDI purposes.

E1. Do you include subordinated loans between MFIs in Other Capital? Number of Member States

Yes 7No 6

E2. Do you include deposits between MFIs in Other Capital?

Yes 3No 10

Table 13 Treatment of other capital data relating to MFIs

41 Only for “other sectors” and not for negotiable instruments.42 In the case of ES, the register of external loans enables the

identif ication of SPE’s located abroad and only engaged inraising funds. These financial flows/positions when are obtaineddirectly from a bank located abroad and routed through an SPEare not included in direct investment but in other investment.


288. For stocks, ten Member States (AT, BE,DE, DK, ES, FR, IT, PT, UK, IE) use bookvaluation, SE uses market valuation and FI andNL use both valuations.

OVERALL COMPARISON OF THEORY ANDPRACTICES43

289. Generally, as can be seen in the tablesabove, there already appears to be a closeconvergence between theory and practice acrossthe Member States for the main principles andtreatment of Other Capital. There is somedivergence in the more specialised areas but thisappears to be related to the importance of theseareas in the individual economies.

290. The directional principle is used for OtherCapital by the majority of Member States withone of the remainder planning to apply thedirectional principle in the future. Thus thetheory and practice of the directional principleappear to have been adopted by the majority ofMember States already.

291. A majority of Member States include dataon trade credits, financial leasing, deposits andpermanent debt in Other Capital. The totalityinclude intercompany loans. Less than 50 percent include data on debt securities. In the nearfuture no Member State will any longer includedata on financial derivatives, which is in linewith the recommendations of the IMF and ECB.

292. The treatment of Other Capital datarelating to MFIs is more erratic. Subordinatedloans are included in Other Capital by half ofthe Member States. The other Member Statesrecord such loans as “equity capital” or “otherinvestment”, frequently taking the view thatthey represent permanent debt. To fully meet the

theoretical treatment of FDI, Member Stateswould need to be able to distinguish betweenloans between MFIs and loans between MFIsand some other sector. Deposits between MFIsare considered to belong to Other Investment by75 per cent of the Member States which accordswith the OECD’s Benchmark.

293. All twelve responding Member Statesinclude data on SPEs but few specificallyidentify them. It is probably more accurate tosay that no Member State specifically excludesdata on SPEs. Without some method ofpositively identifying SPEs it seems doubtfulthat accurate data can be obtained to enableSPEs to be considered separately from anyother type of enterprise.

294. Valuation of flows is evenly split betweenbook and market bases, though, as has beenmentioned before, market valuation does nothave the same implications when consideringOther Capital. For stocks, most Member Statesuse book valuation or both book and marketvaluations.

CONCLUSIONS

THE DIRECTIONAL PRINCIPLE295. The directional principle is a primarycomponent of the accurate compilation offoreign direct investment. It is recommendedthat the directional principle is followedconsistently across Member States whenconsidering transactions in Other Capital.

43 Theoretical issues are presented in Annex 1.

C2. Which valuation principle is used for other capital stocks?

Book value 9 (10)Market value 1Both 2

Table 14 Valuation principles

81c ECB




296. In the case of companies which are fellowsubsidiaries – exclusively through theirrespective relation with their mother companyand with no relevant cross-participation, i.e.less than 10 percent (“sister companies”) – aninvestment should be recorded, by the economyproviding the investment, as direct investmentabroad and, by the economy of the enterprisereceiving the investment, as direct investment inthe reporting economy.

297. If there exists also a direct investmentrelationship (cross-participation of at least 10percent) between two fellow companies thedirectional principle should be appliedaccordingly as for any other FDI relationship,since such companies could no longer beconsidered as fellow companies.

298. In the case of companies with an FDIrelationship through an indirect link ofownership, the directional principle should beappropriately applied in accordance with theconclusions stated in chapter 1.

ELEMENTS OF OTHER CAPITAL WHEREAPPLICABLE TO NON-MFIS299. For some instruments used forintercompany financing the current practices ofMember States are already largely in accordancewith the internationally agreed methodology.However, for several other items, theirallocation to a certain category is apparently not

Item of Extra information IncludedOther Capital or excluded?

Debt securities Except for financial derivatives Included

Preferred shares Possibility of identifying non-participating preferred Excluded, unlessshares non-participating

Trade credits Included

Financial leasing Included

Financial derivatives Excluded

Deposits Included

Intercompany loans Included

Permanent debt Included

Table 15 Elements of other capital included in FDI statist ics

so obvious. The conclusions for the latter aremotivated hereafter. The recommendations forall considered instruments are presented in thefollowing table:

PREFERRED SHARES300. The following conclusions can be drawn:

1. Capital in the form of preferred shares isprovided on a permanent basis without finalmaturity. Preferred shares improve thefinancial stability of a company due to abroadening of its equity base.

2. The three main features of preferred sharesare voting power, claims on the company’sprofits (dividends), and participation in theresidual of a company.

3. As indicated above, preferred shareholdershave hardly any voting power. In return,preferred shares are endowed withprecedences compared to common shares.They are typically ranked ahead of commonequity in respect of dividends anddistribution of the value on dissolution ofthe incorporated enterprise.

4. The expression (non-)participating inrelation with preferred shares is not relatedprimarily to voting rights, but to theparticipation in the distribution of theresidual value on dissolution and to the


return on equity (participating preferredshares incorporate dividend paymentsdependent on the companies profit whereasthe payment of fixed dividends (interest) isrelated to non-participating shares). Thus, iffixed payments are given together withpayments on companies’ profits, then thepreferred shares have to be classified under“Equity Capital”.

5. Following a practical view on the issue ofthe classification of preferred shares andtaking into account their “equity nature”,preferred shares should be treated in generalas equity capital. Only in the case whenpreferred shares can be clearly identified asnon-participating according to all specifiedcriteria (which should be rather difficult)they should be treated as a debt instrumentand recorded as Other Capital.

FINANCIAL DERIVATIVES301. Although financial derivatives areincluded under debt securities in theBenchmark, the IMF and ECB have concludedthat these items should not be included. Asupplement to the BPM5 is dealing withfinancial derivatives as an own categoryseparated from Direct Investment and PortfolioInvestment.

302. Therefore financial derivatives should notbe included in Other Capital.

PERMANENT DEBT303. Permanent debt shows in some respectsimilarities to equity capital. This raises thequestion whether permanent debt should beincluded either in Other Capital or in EquityCapital.

304. Permanent debt is typically characterisedby a higher risk for the creditor compared to0147ordinary0148 loans due to a morecomprehensive liability. In the case of thedebtor’s insolvency, permanent debt willusually not be repaid until all other debts arepaid off. The same applies to interest paymentswhich could be suspended in the case of a

liquidity squeeze. However, permanent debttypically incorporates no voting rights unlikecommon equity capital. Moreover, in generalinterest is paid on permanent debt indicating itscloseness to debt instruments. However, thesame can be said for capital endowment ofbranches, which is nonetheless classified underequity capital.

305. In the case of securitised permanent debt(perpetual bonds) there are further facets to beconsidered. Perpetual bonds are tradeable andfrom the point of view of the creditor, adisinvestment of principal through the capitalmarkets is possible. Perpetual bonds as a formof debt securities might be interpreted to becloser to Other Capital, or rather, more remotefrom equity capital, than subordinated loans.However, no general statement about theranking of securitised and unsecuritisedpermanent debt between equity capital andOther Capital can be made. Moreover, theimportance of securitised permanent debt as afinancing instrument between affiliatedenterprises is not significant for most countries.

306. Consequently, from the theoretical point ofview there is no clear indication whetherpermanent debt is closer to equity capital or toOther Capital. The majority of Member Statesrecord permanent debt (securitised andunsecuritised) as Other Capital. Thus, from apractical point of view, the inclusion ofpermanent debt in Other Capital is preferablefor the sake of simplicity and in order to avoidasymmetries.

307. Overall, practical aspects are decisive inrecommending the inclusion of permanent debt(securitised and unsecuritised) in Other Capital.

OTHER CAPITAL TRANSACTIONS BETWEEN MFIS308. When both parties involved are MFIs,financial intermediaries or financial auxiliaries,permanent debt should be included in OtherCapital. The positions of the TF-FDI membersas to whether permanent debt should berecorded under equity capital or under othercapital in FDI statistics were almost evenly

83c ECB




split. In the end, one of the main reasons to takethis option (recording in Direct Investment/Other capital) was the consistency with thecurrent treatment in money and bankingstatistics.

309. All other transactions involving OtherCapital should be excluded.

310. It should be mentioned that this conclusionis not unproblematic for practical reasons.According to this recommendation permanentdebt is the only remaining item in Other Capital.Therefore, the inclusion of permanent debt inOther Capital might raise confidentialityconcerns in some Member States; in particulardue to the fact that the granting of permanentdebt to affiliated companies in the bankingsector is usually rather limited. Therefore, insome countries where there could be a risk ofdisclosure, caution will have to be observed atthe time of publication/communication of thedata, e.g. these data may have to be flagged asconfidential when transmitted to the ECB.

OTHER CAPITAL TRANSACTIONS OF SPES311. Where they can be separately identified,the activities of SPEs should be included inOther Capital unless their activity solely relatesto the MFI and financial intermediation sector,when they should be treated as previouslyexplained.

312. Where they cannot be separately identified,their activities should be included in OtherCapital as would any other enterprise engagedin FDI.

VALUATION313. Where valuation specifically applies toOther Capital, market valuation should be usedwhere applicable but it is acknowledged thatOther Capital does not always lend itself tomarket valuation in the true sense, or rather thatthere is usually little difference between bookvalue and market value for other capitalstocks.44

METHODS OF COLLECTING DATA ON OTHERCAPITAL314. In general, methods of collecting datashould be considered in terms of FDI as a wholeentity, rather than just one particular element.There is a danger that there may be differentrecommendations for different topics,presenting an incoherent approach across allFDI elements. With this proviso the followingconclusions seem applicable to other capital.

SURVEY-BASED SYSTEMS315. There appear to be two ways in whichsurvey-based systems can collect all elementsof other capital. Firstly, questions can be addedto the survey form asking for each element ofother capital separately, taking into account thedirectional aspect of the investment. Secondly,the notes accompanying the survey form can beamplified to specify the elements required,giving examples where possible. This optionwould seem to be cost-effective once set up.

SETTLEMENT-BASED SYSTEMS316. The settlement-based system uses a seriesof codes for different transactions. In countriesnot currently fully compliant with the standardsagreed for FDI, these codes should be expandedto include the elements of other capital required.They should also include information on thedirection of the investment to satisfy therequirements of the directional principle. Theaccompanying documentation should beexpanded to specify the requirements.

THE APPLICATION OF THE DIRECTIONALPRINCIPLE: PRACTICAL ISSUES AND EMPIRICALEVIDENCE

PRACTICAL ISSUES IN IMPLEMENTING THEDIRECTIONAL PRINCIPLE

PRACTICAL ISSUES IN SURVEY-BASED SYSTEMS317. Surveys typically ask for aggregate datawith a geographical breakdown. Therefore noindividual information of the group structure

44 Just in the event of write-offs / write-downs.


can be derived. However, a separation ofinward and outward direct investment allowsthe classification of claims and liabilitiesaccording to the directional principle also forcountry-aggregated claims and liabilities. Theaccuracy of the data with respect to thedirectional principle crucially depends on therespondent’s available information about theintra-group structure and its effort incompleting the survey. At this juncture thenotes and instructions to the survey given bythe compiler should be designed as a usefulguideline for the respondent with cleardefinitions and requirements.

PRACTICAL ISSUES IN SETTLEMENT-BASEDSYSTEMS318. The application of the directional principlefor equity capital is for most countriesapparently more difficult than for Other Capital.This might be related to the fact that reverseinvestments below the 10% threshold are oftennot considered by the system under directinvestment but under portfolio investment.However, in the following we will concentrateon aspects related to Other Capital.

319. In countries not currently fully compliantwith the standards agreed for FDI, the existingcodes could be expanded to include the elementsof Other Capital required. They could alsoinclude information on the direction of theinvestment to satisfy the requirements of thedirectional principle. As in 4.6.1 theaccompanying documentation should beexpanded to specify the requirements. Theincorporation of separate codes in settlement-based systems should enable the compiler todistinguish between three types ofrelationships, e.g. if the resident enterprise actsas a

– parent company (above the foreign affiliate inthe investment chain)

– subsidiary (below the foreign affiliate in theinvestment chain)

– fellow company (without relevant cross-participation).

320. For outward investment parent companiesshould usually be aware of their status. Forinward investment the resident company mayoften not assess the information if the foreigncounterpart is a (grand)parent or a sistercompany. When a clear distinction is notpossible the foreign affiliate should be treatedas a fellow company which implies that thetransaction is compiled according to the asset/liability principle. However, the implementationof an extension of the reporting system as afirst-best solution is apparently not veryrealistic in many countries.

321. When the extension of the existing systemby additional codes is not possible some otherpossibilities can be regarded. The systematicregistration of transactions in equity capital canprovide an opportunity to identify the directinvestment relationships between affiliatedcompanies. A database on FDI relationships ofresident companies, as in Austria, seems to be auseful tool to ensure that any other capitaltransaction involving affiliated companies canbe classified with respect to the links in equitycapital and therefore effectively recorded underdirect investment according to the directionalprinciple. However, according to Austriancompilers the maintenance of such a companyregister and taking care of its completeness andaccuracy requires comprehensive humanresources.

322. An alternative or additional source to theestablishment or cross-check of own companyregisters are databases maintained by externalproviders. The “WorldBase” database, acommercial database for company interlinkagesprovided by Dunn & Bradstreet is deemed to bevery helpful for identifying linkages for the useof the directional principle. De NederlandscheBank (NL) and the ONS (UK) have asubscription to this database. They use it for allkind of information, such as the directionalprinciple, for the identification of new FDIrelationships which have not been covered inthe system yet and thus to keep the nationalregisters complete.

85c ECB




323. Furthermore, also the systematicexploration of the information provided by theown system may be helpful. The direction of atransaction can be classified for example in thecase of pure subsidiaries, e.g. companies at thebottom of the direct investment chain with noparticipation in foreign companies and in thecase of pure parent companies, e.g. enterprisesat the top of the direct investment chain inwhich no foreign investor has a participation of10% or more.45 In this connection also theinformation from stock statistics can be useful.More precisely, when for example only outwardstocks in equity capital are reported by acompany all transactions in Other Capitalshould be recorded as direct investment abroadand if there exist only inward equity stocks alltransactions in Other Capital should be treatedas direct investment in the reporting economy.

324. Unfortunately, also a geographicalbreakdown of direct investment flows/stockswithout specifying the type of thecorresponding direct investment relationshipscan not provide much additional informationwith respect to the directional principle. Theinformation about the counterpart country inorder to detect the investment relationship vis-à-vis a certain country (pure subsidiary orparent company status) is only useful in thecase when the whole group structure is known,because parent companies and subsidiariescould be located in the same counterpartcountry. However, indirect links (to companiesdownward or upward the direct investmentchain) are usually not covered.

325. In addition to the exploration of thecompany status, the assignment of special codeswhen a loan – granted by a company identifiedas a subsidiary or parent company – was firstconsidered could be helpful. In order toexemplify this approach, the Dutch system isdescribed in Annex 4 of this report, which alsocontains a description of the application of thedirectional principle in Germany.

EMPIRICAL EVIDENCE ON THE IMPORTANCE OFTHE DIRECTIONAL PRINCIPLE326. With a view to assessing the impact that acorrect application of the directional principlecould entail with regard to total FDI figures,Member States which had declared they appliedthe directional principle were requested toprovide detailed information on the differentcomponents of direct investment. 11 countriesprovided figures for flows, while ten alsoprovided information for stocks. Table 16below shows the outcome of this empiricalexercise.

45 This approach is applied in the German b.o.p. for short-termintercompany loans.


Euro area I.I.P. (as published in the ECB Monthly Bulletin) at end-December 2000 (EUR millions)

Other capital

FDI abroad 353,700

FDI in the reporting economy 300,700

Table 16 Impact of the appl ication of the directional principle for “other capital”

(EUR millions)

B.O.P. data for the year 2001

Other capitalEuro area European Union National

Intra Extra Intra Extra

FDI abroad 22,558 -19,300 16,872 -13,612 3,257 claims on affiliated enterprises -9,225 -16,573 -12,381 -13,416 -25,799 liabilities to affiliated enterprises 1) 31,784 -2,727 29,253 -196 29,056

FDI in the reporting economy 32,283 -3,226 24,215 4,841 29,057 claims on direct investors 1) -5,252 -1,953 -6,374 -840 -7,205 liabilities to direct investors 37,535 -1,273 30,589 5,680 36,262

Net 54,841 -22,526 41,086 -8,772 32,314

1) For FR, only long term loans.2) No f igures for IT.3) No figures for FR.

I.I.P. data at end-December 2000

Other capitalEuro area European Union National


FDI abroad 2) 8,308 165,068 25,485 68,903 94,387 claims on affiliated enterprises 3) 55,559 187,441 78,053 85,959 164,012 liabilities to affiliated enterprises 3) 47,252 22,373 52,569 17,057 69,625

FDI in the reporting economy 2) 165,316 125,259 188,047 66,426 254,472 claims on direct investors 3) 28,495 50,523 43,183 35,835 79,017 liabilities to direct investors 3) 193,810 175,782 231,230 102,260 333,489

Net -157,008 39,809 -162,562 2,477 - 160,084

327. As it can be seen in these table, themagnitude of the financing flows and stocks byaffiliates to their parent companies can bedeemed rather substantial in comparison withthe financing flows and stocks in the oppositedirection (i.e. in the usual direction of FDIflows/stocks, namely from parent companies totheir affiliates). On this basis, the TF-FDIconcluded that the non-application of thedirectional principle could cause a substantialand artificial overestimation of the gross assetsand liabilities (i.e. inward and outward FDI)flows and stocks registered in the b.o.p. and thei.i.p. respectively.

87c ECB


6 Identificationand treatment

of special purposeentities (SPEs)

INTRODUCTION

328. The development of multinationalenterprises and the increasing integration of theworld economy, in the process of globalisation,are reflected in a remarkable increase of cross-border financial transactions. The establishmentof complex intra group structures is more andmore determined by financial and fiscal aspects.Against this background Special PurposeEntities (SPEs) are playing an important role inconducting group financing and in providingfinancial services. Moreover, profit shifting inorder to minimise tax burdens is clearly animportant determinant for the establishment ofSPEs in countries with a convenient taxenvironment for foreign companies. Regardingthese activities, not only the growing amountsof intra group cross-border capital flows, butalso the increased complexity of suchtransactions may raise problems in thestatistical treatment of this phenomenon in somecountries.

329. For the time being the recommendations inBPM5 related to SPEs provide no specialtreatment of SPEs except para. 372, which saysthat only transactions in equity capital andpermanent debt between affiliated banks andaffiliated financial intermediaries – includingSPEs with the sole purpose of serving asfinancial intermediaries – should be included inFDI, whereas other capital transactions shouldbe excluded. With the exception of SPEsmentioned in BPM5, para. 372, most memberstates have no special approach for theidentification and treatment of SPEs within FDIstatistics. On the one hand, the absence ofnational legal provisions or specificrecommendations in the national statisticalframework prevents a coherent approach indealing with SPEs. On the other hand, thenecessity for a separate identification of SPEsseems often not to be obvious. Indeed, onecould ask about the reasoning of defining andidentifying SPEs. The discussion within theOECD Workshop on Financial InvestmentStatistics has shown that several compilers donot really recognise the identification of SPEs

6 I D EN T I F I C AT I ON AND T R E ATMENT O F S P E C I A LPURPO S E EN T I T I E S ( S P E s )

within the population of FDI enterprises as atop priority issue. Therefore, some argumentswill be presented hereafter in order to providesome rationale for the specific importance ofSPEs.

REASONING FOR THE IDENTIFICATION OF SPEsAND PROBLEMS IN DEALING WITH SPEs

WHY SHOULD SPEs BE IDENTIFIED SEPARATELY?330. First of all, there is a good reason toidentify SPEs for analytical purposes. Users ofFDI statistics may be interested in the nature ofFDI activities, e.g. whether they stem from realoperational business or merely from fiscalcalculus. In this context a key issue is thereforeto question the economic incentives related toSPE activities.

331. The term SPE comprises a large variety ofenterprises with more or less specific activities.In general, SPEs are primarily engaged ininternational activities but in few or no localoperations. Their functions frequently dependon the legal environment and taxation regime intheir host country. SPEs acting as financialvehicles can be economically more accuratelycharacterised as intermediaries rather than asdirect investors or direct investment targets.The impact of SPEs on the economy they areoperating in is often negligible. In particular,holding companies, shell companies, etc.typically have no real economic activity, noturnover and no significant employment.

332. The so-called green field investments areregarded as an important indicator for theeconomic attractiveness and competitiveness ofregions and countries competing for directinvestment. Policy makers and economists doregularly refer to foreign direct investmentfigures as an indication for economic conditionsand the quality of economic policy. Thesediscussions are mainly focused on inward flowsand stocks. The increasing amounts of cross-border financial transactions arising from theformation of SPEs – together with theaccelerated M&A activities – have extended


gross financial flows significantly. However,these transactions have hardly any real impacton the domestic economy and do commonly notaffect the creation of new production capacitiesand employment opportunities. Therefore, theinterpretation of gross flows and theconclusions drawn from this might bemisleading, if comprehensive gross directinvestment transactions arise from fundschannelled through third countries rather thatdirectly from the investor to the final target.

333. There can also be deducted a connectionwith the calculation of direct investmentincome. International standards congruentlyrecommend the application of the currentoperating performance concept including onlyprofits and losses arising from operationalbusiness in FDI income. In this context onemight argue that all foreign direct investmenttransactions and positions should be identifiedaccording to their “economic background”. Inthe case of holding or shell companies, theassignment of FDI flows and stocks accordingto the ultimate investor and the ultimate target,respectively, would facilitate a geographicalallocation in agreement with real economicincentives.

PROBLEMS IN DEALING WITH SPEs334. In the context of providing meaningfulFDI statistics, severe problems related to SPEsarise from:

– the identification and classification of SPEs,in particular holding companies, shellcompanies, financial intermediaries andfinancial auxiliaries. Detailed informationabout the business segment are essential foran assessment of the economic rationalebehind the investment activities. In thisrespect an industry breakdown might oftennot be sufficient for an unambiguousdistinction between SPEs and other FDIenterprises. Moreover, information about thebusiness segment or industry are oftenlimited to domestic entities. In the case ofoffshore centres one might assume that SPEsplay a major role in these countries, simply

classifying enterprises located in offshorecentres as SPEs. However, there are evidentdifferences between the group of offshorecountries. Regarding offshore centrespossessing a broad economic basis, e.g.Hong Kong or Singapore, inward andoutward direct investment might also bemotivated by real economic activities ratherthan merely by fiscal incentives.

– the geographical allocation: When funds arechannelled through a SPE, the real economicactivity might often not be located in thecountry where the SPE is resident, but inthird countries where the final target of FDIis located. When only direct FDIrelationships are taken into account, FDItransactions and positions might be biasedtowards host countries of SPEs. In thiscontext the identification and coverage ofindirect relationships or the possibility of“looking through” holding and shellcompanies is a critical issue. Severalcountries have already developed approachesto tackle this problem. For example, in theU.S. FDI statistics, foreign shell companiesincorporated abroad having all their physicalassets or operations in a second foreigncountry are treated as incorporated foreignaffiliates in the second country where theirphysical assets or operations are located.However, for the implementation of such atreatment, comprehensive informationprovided by reporting companies arerequired; and for the time being such data areapparently not available in most countries.

– asymmetries between national statistics:according to international standards, theactivities of SPEs integrated in amultinational group structure should betreated as direct investment. However,various national practices in treating SPEs, inparticular financial intermediaries andfinancial auxiliaries, are causingasymmetries. Moreover, the sectorclassification might be differing amongmember states.

89c ECB




PREVIOUS INVESTIGATIONS BY THE OECD335. The analysis of the subgroup on SPEsshould also shed some light on the differentapproaches in dealing with SPEs and off-shoreenterprises among the member states. Acategorisation of country specific approaches indealing with SPEs was already attempted withinthe scope of the OECD Working Party onFinancial Statistics in the recent past.46 Sincewe do not intend to duplicate this exercise wefocused on an empirical investigation, whichwas designed to analyse the present stage ofpossibilities in identifying SPEs in nationalstatistics and to evaluate the order of magnitude(of parts) of SPE activities.

336. There is still no common approach in orderto tackle these open issues among Europeancountries. Though the results of the OECDstudy are still preliminary and will be reviewedin a second stage, the OECD paper represents auseful reference providing an overview of thevarious treatments of SPE activities in differentcountries. The results indicated that only a fewcountries possess a legal provision regardingthe definition of SPEs and that there rarelyexists any specific recommendation for thetreatment of SPEs in FDI statistics. As thediscussion within the OECD Workshop onInternational Investment Statistics showed, itproved hard to categorise the treatment of SPEsamong countries when most countries haveeven not established recommendations for aspecific treatment of FDI enterprises.

337. What can be learnt from the OECDstudies? The large variety of answers andcomments provided by compilers indicates themost important obstacles in identifying andanalysing SPEs and their activities:

– there is hardly any definition,recommendation or treatment of SPEs appliedby at least two countries in the same mannerexcept for those countries not distinguishingbetween ordinary FDI enterprises and SPEsat all.

– there is no country of the euro area collectingdata from resident enterprises on their SPEsabroad.

– information about tasks and activities of SPEsare not collected systematically by mostcountries. In this context, also the use ofshort-term instruments is still not analysedcomprehensively. However, in particularshort-term intra-group financing causessubstantial cross-border capital flows.

338. Against the background of the follow-upwork of the OECD secretariat, the subgroup onSPEs did not intend to replicate these results byan own survey, but to focus on an empiricalexercise exploring the extent of SPE activitiesthat can already be identified for the time being.It might be helpful to identify a least commondenominator with respect to the category ofSPEs which can already be separated by someMember States.

DEFINITIONS OF SPEs IN INTERNATIONALGUIDELINES

339. This section presents the conclusionsderived from the examination of some selecteddefinitions of SPEs available in the maininternational guidelines. Mainly the IMF 5th

Manual, the IMF BOP Textbook, the IMF BOPcompilation guide, the OECD Benchmark, theESA95, and the International FinancialReporting Standards (IFRS-2003 Edition) havebeen investigated. The results of theseinvestigations are presented in greater detail inthe stand-alone document that served as thebasis for this chapter (“Report of the sub-groupon Special Purpose Entities”).

340. The exploration of various internationalguidelines has put forward the difficulty forcompilers to define SPEs in simple words.However, some common and internationally

46 See “Report on Special Purpose Entities and Off-ShoreEnterprises”, note by the OECD secretariat prepared for theWorkshop on International Investment Statistics, 5-6 March 2003in Paris, DAFFE/MC/STAT(2003)4.


agreed characteristics that define an SPE couldbe helpful for the identification and theharmonisation of the treatment of SPEs. Thefollowing conclusions rely mainly on the fact-finding exercise done in IMF or OECDManuals. Information relating to SPEs in theIFRS manual deals mainly with the issue linkedto the consolidation of SPEs and the concept ofcontrol in accounting standards.

341. SPEs are defined in different ways ininternational guidelines, the OECD benchmarkfocusing more on the purpose or the structure ofSPEs whilst the IMF puts forward the conceptof their engagement in few or no localoperations. But the essential point is that allthese definitions do not contradict each other.The OECD, in its report on Special PurposeEntities and Off-shore enterprises (p.9),merged both its own definition with the IMFone to present a common understanding ofSPEs according to international guidelines,which can be summarised as follows47. SPEsare:

– generally organised or established ineconomies other than those in which theparent companies are resident.

– engaged primarily in internationaltransactions but in few or no local operations.

– SPEs can be a financing subsidiary, a holdingcompany, a base company or a regionalheadquarter.

– SPEs can act as a sale and regionaladministration, a management of foreignexchange risk, or have the purpose offacilitating the financing of investment for thewhole group.

342. In practice, the concept of “few or no localoperation” seems to be the one which is more inline with national definitions currently availablein few Member States (mainly DK, the NL and,to some extent, IE).

47 Report on Special Purpose Entities and Off-shore enterprises –Workshop on International Investment Statistics, 5-6 March2003, paragraph 15 p9.

343. The definitions expressed above, mergingboth IMF and OECD definitions and notcontradicting the current national definitionswill constitute the basis for any future work thatcould be enhanced by Eurostat or the ECB in thefield of SPEs.

344. Regarding residency criteria, both theOECD and the IMF consider that offshoreenterprises, including SPEs, have to beconsidered as residents of the economy inwhich they are established.

345. Regarding the statute of SPEs and thetreatment of their FDI transactions, both theOECD and the IMF recommend to considerSPEs as resident direct investment enterprisesif they meet the general criteria defining FDIrelationships, therefore to treat their FDItransactions accordingly.

346. For SPEs with the sole purpose of servingin a financial intermediary capacity, allinternational guidelines recommend to restricttheir FDI transactions only to equity capital andpermanent debt (or fixed assets for branches).

347. The IMF Textbook (but not the 5th Manual)is the only guide defining SPEs according totheir operational presence in the host country(the OECD only raises the question in the “Taxauthorities’ perception” paragraph):

– SPEs are generally established in economiesother than those in which the parentcompanies are resident.

– SPEs are engaged primarily in internationaltransactions but in few or no local operations.

348. But nothing is said regarding the evolutionof the SPEs’ population of enterprises (whatshould be done in case of an expansion of theactivities on the local market).

91c ECB




349. Both the OECD and the IMF recognise thatsome country may net out transactions throughSPEs, but they all recommend that nationalcompilers should provide information on agross basis, for international comparison.

350. All international guidelines agreed on thefact that SPEs can take the form of holdingcorporations. ESA95 has a precise definition ofholding corporations (§2.14), – institutionalunits whose main function is to control anddirect a group of subsidiaries. Theidentification of holding companies is neededfor the sectoral classification of SPEs in S11,S122 to S125. Such an identification isnecessary to indicate which FDI treatmentshould be adopted (the general FDI one or theone defined for SPE acting only as a financialintermediary).

351. Holding SPEs have to be classified in thesector which better reflects the activity of thegroup of subsidiaries as a whole (measured onthe basis of value added, ESA95 §2.23-e,2.40-e, 2.43). ESA95 states that a holdingcorporation controlling a group of subsidiariesconsisting predominantly of insurancecorporations and pension funds has to beclassified in the financial intermediation sector(S123), instead of the insurance one (ESA95§2.43 and 2.63-b). However, on this specifictopic, paragraph 70 of the IMF “Monetary andFinancial Statistics Manual” does not mentionthis exception. This Manual states that(extract):

– “A holding corporation is classified asfinancial if the preponderant activity of thegroup of corporations as a whole isfinancial.”

– “Similarly, financial holding corporationsshould be allocated to subsectors according tothe type of financial activities mainly carriedout by the group they control.”

352. For an SPE not having the statute ofholding, it is necessary to check if the SPE isacting as a financial intermediary or simply as a

treasury provider (ESA95 §2.34, 2.37, 2.55-f,IMF Textbook examples §543, 544, OECDBenchmark §69).

353. ESA95 states that financial intermediationdoes not include institutional units providingtreasury services, unless subject to financialsupervision.

– Units acting as treasury services, and notsubject to financial supervision, are allocatedto the sector better reflecting the predominantfunction of the group.

– Finance vehicle corporations holdingsecuritised assets have to be classified underS123, or S122 if they are MFIs.

IMPORTANCE OF SPEs AND OFFSHORE COUNTRIESIN EUROPEAN STATISTICS

354. The TF-FDI highlighted the fact that notall the MS identify SPEs in their regularstatistics. Mainly the NL and, to some extent,DK and IE reserve a specific treatment/definition for transactions involving them. TheNL does not include the gross transactions ofSPEs in the national FDI statistics. The CentralBureau of Statistics (CBS) of the NL arguesthat transactions with SPEs do not affect theDutch economy, and therefore are not relevantfor National Accounts purposes (no effect onthe net worth). Despite the fact that Eurostatand the ECB compile European aggregatesfollowing the same method, the results obtainedobviously diverge, which is not satisfactoryfrom an analytical point of view, and from theusers’ one.48 Investigations done within theTF-FDI confirmed that the other Member Statesinclude transactions with SPEs both in theirdata transmission to ECB and Eurostat.

48 Even though, from the purely national viewpoint, the exclusionfrom national statistics of SPE’s transactions may make sense,since, in some cases, national statistics could otherwise beblurred by the volume of financial transactions between non-resident entities channelled through domestic SPE’s.


355. Eurostat is not able to evaluate directly theglobal contribution of SPEs using the currentEurostat/OECD questionnaire. Within thisquestionnaire, information relating to eitherFDI capital invested abroad by resident directinvestors acting as financial/managementholding companies, or invested in residentfinancial/management holding companies, canbe found. But it cannot be assumed that allholding companies could be considered asSPEs. Furthermore, non-holding SPEs couldexist if we refer to all the internationaldefinitions expressed in the previous section.For this reason, Member States have been askedby the TF-FDI to provide an estimation of thedirect contribution of SPEs in FDI statistics.

356. The Eurostat/OECD questionnaire asksinformation about FDI relationships with theso-called offshore centres, a group constitutedby “small” economic countries (in term ofGDP). FDI transactions are recorded accordingto the first shot criteria. It is questionablewhether large amounts (if not all) investeddirectly in the EU by offshore centres wouldneed to be reallocated if the UBO criteria had tobe applied. Reversibly, EU direct investmenttowards offshore centres could be seen asinvestments for which we do not know exactlythe final destination.

357. Given the relative “disconnection”between their GDP size and the amount of FDIcapital they generate, it seems suitable to go onhaving a close follow up of this specificpopulation. Another reason rests on a possiblecorrelation between resident SPEs and offshorecountries, an assumption clearly stated inAnnex 3 of the OECD Benchmark definition(“Location of SPEs”, p.45, 1st sentence). Butthis assumption will be difficult to check, atleast on European data.

358. One of the purposes of the followingtables is to try to convince Member States toinvestigate whether it is necessary to have anin-depth analysis about SPEs and offshorecentres. All the following tables should beanalysed, bearing in mind that:

i) the Eurostat current list of the offshore(financial) centres49 cover in fact “non-European” offshore (financial) centres.

ii) this list has been updated recently (inOctober 2002, therefore mainly for the 2001and 2000 FDI data), which might altercomparisons over time. But the formerEurostat list already contained the mostsignificant offshore countries. Therefore, itcould be expected that these (recent)changes do not influence significantly thelevel of FDI transactions with offshoreentities.

OVERVIEW OF THE IMPACT OF OFFSHORECENTRES ON EUROPEAN AGGREGATES359. Table 17 presents the evolution of EUFDI transactions with offshore centres since1995, both in absolute value and with regard tothe total extra EU transactions. The non-symmetry between outward and inward flows isquite obvious: whilst the evolution is rathervolatile on the outward side, EU inflowsreceived from offshore centres have regularlyincreased since 1997, from €4.9 bn to €13.4 bn.Furthermore the issue linked to offshore centresseems to be more relevant on the inward side:on average over the 1995-2001 period, EUinflows received from offshore companiesaccounted for 9.1% of the total extra EUinflows. Two reasons could explain the smallerpercentage on the outward side:

i) the Eurostat current list of offshore entitiesdoes not contain any European countries. Itmight be possible that EU direct investorsare mainly dealing with similar entities closeto the EU border (inside or outside the EU),

49 Eurostat current list of offshore financial centres (31 countries):Antigua and Barbuda, Anguilla, Netherlands Antilles, Barbados,Bahrain, Bermuda, Bahamas, Belize, Cook Islands, Dominica,Grenada, Hong Kong, Jamaica, St Kitts and Nevis, CaymanIslands, Lebanon, Saint Lucia, Liberia, Marshall Islands,Montserrat, Maldives, Nauru, Niue, Panama, Singapore,Turks&Caicos Islands, Saint Vincent and the Grenadines, VirginIslands (UK), Virgin Islands (US), Vanuatu and Samoa.

93c ECB




whilst EU direct investment companiesreceive FDI capital from foreign groupslocated in other continents (Caribbean areaor South East Asia zone).

ii) inward and outward flows related to SPEsreported from the NL are not available in theEurostat database. May be it could fill inpart of the gap of overall flows to and fromoffshore countries.

360. For direct investors of the Eurozone, therelative interest for investing in offshorecountries is quite similar to the one observed atthe EU level (3.4% of extra Eurozone outflows,see Table 18 below). But the relative weight ofinflows from offshore companies is estimatedto 4.1% of the total extra Eurozone inflows, 5percentage points less that the weight observedon EU aggregates. Despite this gap, it seems

reasonable to consider that percentagesobserved on the inward side are quitesubstantial, which justify the need of furtherinvestigation for analytical purposes.

361. The relative importance of the group ofoffshore countries is also confirmed by the FDIpositions data, however with a better balancingbetween assets and liabilities: 5.1% and 7.4%respectively, for the European Union as awhole, 3.3% and 3.9% for the Eurozone as awhole entity (see Table 19).

362. Table 4 of the supplementary document“Special Purpose Entities” presents theclassification of the main extra EU FDIpartners, both on the assets and liabilities sides.At the moment, Eurostat doesn’t have thepossibility to offer a full breakdown of EUdirect investment vis-à-vis all extra EU

Table 17 EU FDI f lows transactions with the rest of the world and offshore centres

(EUR million)

1995 1996 1997 1998 1999 2000 2001 1995-2001

Outward

Extra EU -62,407 -68,665 -109,802 -218,754 -302,395 -408,925 -234,800 -1,405,749Offshore centres -3,337 -5,561 -6,811 -1,830 -8,971 -7,351 -12,332 -46,194% 5.3 8.1 6.2 0.8 3.0 1.8 5.3 3.3

Inward

Extra EU 42,464 36,509 50,160 96,432 102,118 150,407 118,470 596,559Offshore centres 3,427 2,583 4,932 5,816 10,234 13,628 13,403 54,023% 8.1 7.1 9.8 6.0 10.0 9.1 11.3 9.1

Table 18 Euro area FDI f lows with extra Eurozone and offshore centres

(EUR million)

1999 2000 2001 1999-2001

Outward

Extra Eurozone -310,409 -411,863 -252,341 -974,613Offshore centres -5,523 -11,205 -16,241 -32,969% 1.8 2.7 6.4 3.4

Inward

Extra Eurozone 174,459 343,118 157,474 675,052Offshore centres 8,409 7,788 11,788 27,985% 4.8 2.3 7.5 4.1

Note: Transactions involving Dutch SPEs are excluded from Euro area calculations.


Table 19 FDI posit ions with the rest of the world and offshore centres

(EUR million)

European Union Euro areaValues at end 2000 Values at end 2000

Assets

Extra EU/Eurozone 1,517,186 1,335,073Offshore centres 77,236 44,493% 5.1 3.3

Liabilities

Extra EU/Eurozone 890,709 1,112,495Offshore centres 65,800 43,790% 7.4 3.9

Note: Transactions involving Dutch SPEs are excluded from EU and Euro area calculations.

individual countries. The current list involves50 individual extra-EU countries, and it hasbeen necessary to introduce some groupingzones (eight in all) to better approximate a fullcoverage of the extra EU area. None of theseadditional groups contains any of individuallisted countries and, apart from the “GulfArabian countries” (see footnote 5), this listalso avoids the possibility of having one non-listed country being included in more than oneeconomic group.

363. When looking at the regional distribution,the results clearly illustrate the dominantpositions of offshore companies, in the field ofFDI:

i) assets side: offshore financial centres,excluding Singapore and Hong-Kong,hosted altogether €41 billion of EUexternal FDI assets, ranking in the top 6countries’ list just behind Argentina andbefore Australia.

ii) liabilities side: offshore financial centres,excluding Singapore and Hong-Kong,were altogether responsible of €52 bn ofEU external FDI liabilities, ranking in the3rd position just behind Switzerland andbefore Japan.

iii) if we add Singapore and Hong Kong, thenit could be seen that the group formed byoffshore financial centres, as defined in

the current Eurostat list, is the 3rd EUpartner both on the assets and liabilitiesside (€77 bn on the assets side, €66 bn onthe liabilities side).

364. From this table it could also be possible toportray also the impact linked to a possibleextension of the offshore list. For the moment,no official “offshore” list has been established/approved by the International Institutions.Nevertheless, the “Other European countries”group50 defined by Eurostat involves countrieshaving (more or less) similar characteristics tothose from the Eurostat offshore list. If we hadto include them in the Eurostat list, then the totalamount of EU external FDI assets located inoffshore countries would have jumped from€77 to €115 billion, in the 2nd position behindthe United States and far ahead of Switzerland.

365. These results put forward the powerfulrole played by offshore centres in the field ofFDI, if we consider them as a unique entity.With the perspective of improving internationalcomparison it is questionable whether thepresence of offshore entities alter significantlythe regional distribution picture.

366. The extra EU FDI capital structure couldalso point out the specificity of these entities, asit is shown in table 20: On the liabilities side,the capital distribution between “equity capital

50 Andorra, Faroe Islands, Guernsey, Gibraltar, Isle of Man, Jersey,Moldova, Macedonia, San Marino and Vatican City State.

95c ECB




and RIE” and “other capital” is around twothird/one third respectively. For extra EU FDIliabilities vis-à-vis foreign offshore companies,the FDI capital distribution profile divergessignificantly from the extra EU average: 43% ofthe total assets held in the Union by offshorecountries were constituted by inter-companydebt stocks (see Table 20).

FOCUS ON US DATA367. Would the application of UBO criteria alterthe importance of offshore entities? A firstattempt to this answer is given by the analysisof US figures, shown in Annex C of the “Reportof the sub-group on Special PurposeEntities”.51

368. The main comments are synthesised in thethree following points:

(i) The application of the UBO criterion doesnot alter the large influence of offshorecountries52 in US statistics. Surprisingly,US figures indicate even a reinforcementof this group, when going from first shotto UBO.

(ii) However, if offshore countries areanalysed individually, the role played byeach of them is drastically affected by theUBO country reallocation.

(iii) There are apparently large differencesbetween offshore centres subject to theireconomic background. It seems that themove from first shot to UBO does notalways enable the identification of the

51 See supplementary document “Special Purpose Entities”.52 Offshore centres estimated with Bahamas +Bermuda

+Netherlands Antilles +Panama +UK Islands, Caribbean +OtherOWH +Liberia +Lebanon +Hong Kong +Singapore.

UBO owners’ nationality. In fact, theconstraints encountered by the BEA for afull application of their UBO definition arenot known.

RESULTS OF THE EMPIRICAL EXERCISE ON THEIMPORTANCE OF SPEs IN EU MEMBER STATES

369. An empirical investigation concerning therole of SPEs was conducted by the TF-FDI. Theoutcome of the empirical exercise on the impactof SPEs in b.o.p. and i.i.p. data illustrates thedifficulties of such an exercise in the absence ofa specific definition of SPEs in a majority ofcountries: five countries replied that they couldnot provide any information and four (BE, DE,FR and GR) provided data, essentially byapproximating the definition of SPEs with oneor more of the following: co-ordination centres,financial holding companies, managementholding companies and “other services forenterprises”.

370. The answers received are attached inAnnex D of the “Report of the sub-group onSpecial Purpose Entities”. No meaningfulaggregation has been deemed feasible, given thedisparity of the proxies used to compile thedata.

371. The examination of the figures providedshow that the amounts involved both for flows

Table 20 Extra EU FDI posit ions capital structure at end 2000

(%)

Equity cap. & RIE Other Capital Total

Assets

Extra EU assets, exc. Offshore centres 79 21 100Offshore centres 92 8 100

Liabilities

Extra EU liabilities, exc. Offshore centres 65 35 100Offshore centres 57 43 100


and for stocks are quite significant, especiallyin DE. As the four countries which haveprovided data for the exercise are not known tobe particularly attractive for SPEs, (contrary toNL), this would indicate that the validity of theapproximation SPEs = holding companies isquestionable.

372. As already said, SPEs cannot be identifiedas such within the NACE classification ofactivities. In international guidelines, SPEs areidentified not only according to their maineconomic activity, but also through theirstructure’s analysis (holding or non-holdingcorporation).

373. As some Member States proceed, Eurostatapproximated the importance of some type ofSPEs using a correspondence table to link theOECD definition (holding company, regionalheadquarters, management of foreign exchangerisk etc.) with the main concerned classes/groups (of the NACE rev1) for hosting SPEs.Annex E of the above-mentioned report gives anoverview of this linkage together with anestimation of the impact of these groups, at theEuropean level.

374. The fact that, apart from the NL, DK and,to a lesser extent, IE, there exists no specificdefinition of SPEs in EU countries may alsoindicate that there is no real need for such adefinition because the issue is not important inthese countries.

SPECIAL TREATMENTS OF SPEs IN EU MEMBERSTATES375. As already mentioned SPEs play animportant role in a few countries and thereforespecial treatments were established in thesecountries. As the net flows of SPE transactionsthrough the NL are close to zero and hardlyaffect the national economy, the gross flows arenot included in the national FDI statistics of theNL. DK has still not decided if certain SPEactivities should be excluded for nationalreporting and IE identifies separately a part ofits SPE population. The treatment of SPEs in

DK, IE and the NL is described in Annex F ofthe “Report of the sub-group on SPEs”.

SUMMARY

376. Offshore financial centres are importantFDI partners, having a significant impact onEuropean statistics. For internationalcomparison, it would be necessary to have anagreed list of offshore financial countries,which is also stable over time.

377. The US figures show that a moving fromfirst shot to UBO criteria does not “globally” alterthe importance of offshore financial centres. Butit seems necessary to distinguish betweenoffshore centres with a real economic backgroundand offshore centres whose attractiveness ismerely based on fiscal incentives. However, ahigh sensitiveness to change in criteriaclassifications has been observed for eachoffshore country taken individually. The use ofUBO implies a geographical breakdown of FDIdata better reflecting economic reality. But resultsobtained on US figures leave some “openquestions” in the sense that it is not sure whetherthey have identified the real UBO company:between the first shot identified partner and thereal UBO one, FDI capital could be in the hand ofseveral intermediary offshore entities, and we donot know the real reasons for assuming directFDI relations between EU SPEs and “offshoreSPEs”.

378. The restricted analysis on holdingactivities, using the Eurostat/OECDquestionnaire by activity, has put forward the“potential” impact of SPEs in FDI statistics,both on European aggregates and at the MemberStates level. At the moment, a more accuratemeasure of SPEs’ transactions is not feasible:none of currently available definitions –international guidelines or the national one –considers all holding companies as SPEs.

97c ECB




CONCLUSIONS

379. The analytical interest of a separateidentification of transactions/positions related toSPEs was exposed at the beginning of thischapter. However, various problems in dealingwith SPEs restrict the separate identification ofSPEs. The empirical investigations on theimportance of SPEs among member states hasrevealed that even the identification of domesticSPEs is at present obviously not possible for mostmember states. At best SPEs could roughly beapproximated by certain sectors potentiallyincluding SPEs. Moreover, it is clear that a fullanalysis of the transactions/positions involvingSPEs would also require the identification of non-resident SPEs when they are counterpart to atransaction/position of a resident company: thiswould imply the existence of a world-wide database on SPEs, which seems an impossible goal toachieve in the absence of a universal definition ofSPEs and of a general recognition of the necessityof a separate identification of the activities ofSPEs.

380. The importance of SPEs varies significantlyamong member states. Consequently, practicesand efforts in identifying SPEs and theapplication of special treatments differ betweencountries. Potential asymmetries betweennational statistics respectively, differences in datadissemination to international organisationsshould be avoided, or be properly exposed.

The TF-FDI recommends the inclusion oftransactions/positions of/with SPEs or SPE-likecompanies in b.o.p./i.i.p. reporting concerningthe contributions to the euro area/EUaggregates.

Notwithstanding all the practical andconceptual difficulties previously stated, theTF-FDI recommends that the possibility tocollect separate statistics for SPEs continuebeing assessed by both working groups and inthe framework of ad-hoc workshops in thefuture. To this aim, co-ordination should beensured with the related work currently beingdeveloped in the OECD.

381. Additionally, chapter 553 recalled thedecision of the IMF, in co-ordination with theECB’s WG-BP&ER and the OECD’s WorkingParty on Financial Statistics (WPFS),concerning inter-company loans betweenaffiliated MFIs. In particular, there is anexplicit reference in the IMF resolution to SPEsprincipally engaged in financial intermediationfor a group of related enterprises.

According to the IMF decision, SPEs principallyengaged in financial intermediation for a groupof related enterprises should be included in thecategory of affiliated financial intermediariesand, therefore, inter-company loans with anyother institution included in this category shouldbe excluded from direct investment and should berecorded in other investment54.

53 See Annex 1.54 Permanent debt should still be recorded in direct investment.


382. In line with the suggestion made in theBD (§45), the TF FDI investigated into thecalculation of inward stocks data broken downby both the country of the first foreign parentand by the country of the UBO (ultimatebeneficial owner). The allocation by country ofUBO allows allocating FDI stocks according tothe country of residence of the entity thatexercises control on the capital stockconsidered. Data on the impact of the UBOcriterion as concern the geographical allocationof inward FDI stocks were collected from AT,DK and DE.

CONCEPTUAL FRAMEWORK

INTERNATIONAL STANDARDS383. The OECD BD does not give a definitionof the UBO. During the discussion at the FATSWG, it was accepted that for FATS purposes thefollowing definition would be used:

The ultimate beneficial owner is the first personproceeding up the chain (beginning with andincluding the first foreign owner) that is notcontrolled by another person.

384. This definition was taken from the oneused by the BEA for USA UBO inward FDIstocks, with only difference that the Americandefinition says “majority-owned” instead of“controlled”. The use of the term “controlled”was considered preferable for FATS becausemajority ownership is seen as a practicalcriterion to approximate the target concept ofcontrol at each stage in the chain. In the lastproposal of the draft regulation for FATS theUBO has been in fact renamed “ultimatecontrolling institutional unit” and the newproposal is as follows (article 2):

“Ultimate controlling institutional unit of aforeign affiliate” shall mean the institutionalunit, proceeding up a foreign affiliates’ chain ofcontrol, which is not controlled by anotherinstitutional unit.”

7 A L LO C AT I ON O F F D I I NWARD S TO CK S B YCOUNTRY O F “ U LT IMAT E B ENE F I C I A L OWNER”(UBO )

385. This choice might not appear appropriatefor FDI, which measure the lasting interest evenfor cases in which control on the directinvestment enterprise is absent. However, itshould be also clarified that the UBO in theabove definitions is not necessarily the onlyunit having rights on the income generated bythe direct investment capital considered, as itcould be suggested by the term UBO. Incomereceived is proportional to the percentage ofindirect ownership, calculated by multiplyingthe percentages of ownership in the chain goingup from the direct investment enterprise. Thispoint was clarified during the FATS WGbecause of the existence of an old definition inthe SBS framework based on the % of indirectownership. This definition has now beendeleted from the SBS set of definitions and it isnot known if it has ever been applied by anycountry in the context of FDI statistics.

EXAMPLESExample1: The resident direct investmententerprise (DIE) is held 100% by the firstforeign parent, but then there is a 51% chainleading to the UBO. The other non-residentminority intermediate owners receive someincome from the resident DIE, but they are notto be considered UBOs, even if they are notcontrolled by another person. This kind ofcalculation would be practically impossiblebecause various different chains would need tobe followed.

386. Here §42 of the BD can be quoted: “Forinward direct investment it is possible toestimate earnings and the stock of net assets dueto the immediate investing country and toreanalyse this by country of ultimate control.The share of the earnings and net assetsattributable to the ultimate parent company willnot normally be known. This is because the hostcountry does not know the percentageshareholdings in the various intermediarycompanies between it and the ultimate parent.”

387. As said before, the main differencebetween FATS and FDI is that for FDI alsoshares in resident direct investment enterprises

99c ECB


7 Allocationof FDI inward

stocks bycountry of“ultimatebeneficial

owner” (UBO)

that are between 10 and 50% should beconsidered. Therefore, for one directinvestment enterprise, there can be more thanone UBO. An example discussed at the FATSWG in respect of the use of the rule based on the% of indirect ownership can illustrate this case.

Example 2: B is the resident enterprise while theothers are located each in a different country. Wholds indirectly 52% of B in the examplereproduced below (Chart 8). However, X(holding 48%) should be considered as thecontrolling unit by the application of the step-by-step majority ownership rule. For FDI, the60% of the stock should be allocated to thecountry of residency of X and the remaining40% to the country of residency of W.

388. In conclusion, it seems preferable tocontinue using the term UBO in FDI and notadopt the UCIU terminology introduced forFATS. There are differences with FATS in thatalso minority shares are considered. However,the selection of the UBO in terms of indirectownership shares is not correct.

389. The fact that there can be multiple UBOssuggests clarifying the definition given in thebeginning as follows:

The ultimate beneficial owners are the firstpersons proceeding up the chains(beginning with and including the first

foreign owner of each FDI position of adirect investment enterprise) that are notcontrolled by another person.

PRACTICAL APPLICATION IN SOME EU COUNTRIES

RESULTS OF THE SURVEY ON INDIRECTRELATIONS390. The questionnaire on indirect FDIrelationships mentioned in chapter 2 requestedinformation on, inter alia, the availability ofdata concerning the ultimate beneficial owner offoreign direct investment flows and stocks. Theanswers to this question have been summarisedin Table 21:

Name of ultimate mother 7 (+1)Country of residence of the UBO 10 (+1)% of direct ownership of the UBO 2 (+2) 2)

% of indirect ownership of the UBO 2 (+2)Details of levels in between (1)

Table 21 UBO principle: countries in whichthe information is avai lable 1 )

1) Numbers in brackets represent countries in which theinformation is only partially available.2) These data are only known in case an organisation chart isavailable.

391. Methodology and information sources:countries normally request the identification ofthe UBO to the respondents in the FDI survey.If no UBO is mentioned, the directly linkedforeign owner is considered as the UBO. Forsome countries, more than one single UBOcould exist (each one holding a share of morethan 10%).

392. The production of additional statisticsbased on the UBO was considered feasible for alarge majority of respondents (10). However,the feasibility should be subject to thefollowing conditions:

– All countries should agree on a commonapproach.

– The methodology to be applied should beclearly defined. For instance, in cases wheremore than one single UBO exists, the

X

B

YA

W

Chart 8 FDI case with more than one singleUBO


criterion to allocate the data should beestablished.

– Suitable software application to extract datashould be available;

– In some cases, more detailed data would needto be asked to the resident enterprise;

FURTHER INFORMATION ON COUNTRY PRACTICES393. From Table 22 below reproduced from theSIMSDI results, it can be seen that 8 EU MS(AT, Belgium, DK, DE, IE, Luxembourg,Portugal and Sweden) compile inward FDIstocks also according to the country of UBO.

Countries compiling geographical breakdown in respect of the:

Immediate host/investing country Ultimate host/investing country

OECD countries Inward Outward Inward Outward

Australia IM IMAustria IM ULBelgium IM IM UL ULCanada IM IMCzech Republic IM IMDenmark IM IM UL ULFinland IM IMFrance IM IMGermany IM IM ULGreece IM IMHungary IM -Iceland IM IMIreland IM IM UL -Italy IM IMJapan IM IMKorea IM IMLuxembourg IM IM UL ULMexico IMNetherlands IM IMNew Zealand IM IMNorway IM IMPoland IM IMPortugal IM IM ULSlovak Republic IM IMSpain - - - -Sweden IM IM UL ULSwitzerland IM ULTurkey - - - -United Kingdom IM IMUnited States IM IM UL

Total OECD (30)Immediate host/investing 24 21Ultimate host/investing 9 5Response not available or not applicable 2 5

Table 22 Al location of geographical FDI posit ion data

The remaining EU MS do not have plans forcalculating stocks on the UBO principle.

394. Descriptions of the method used and datacomparing FDI stocks with first shot and UBOallocation have been supplied by AT, DK andDE. These contributions are given in Annex 6,at the end of this report.

395. As concerns the methods, the maindifference is that in DE and in DK only thecountry of the UBO is known. The informationis asked to the respondents and if the reply isnot supplied, the first shot and the UBO areassumed to coincide. Also in AT the

101c ECB




owner” (UBO)

information is asked to the respondents, butwith more detail including the name and the %of ownership in the intermediate steps of thechain of control. Another important differenceis that in AT and in DE the allocation by UBO ismade on total FDI stocks (including othercapital) whereas in DK equity capital only isconsidered. The detailed information suppliedby the three countries is annexed to the report.

396. The countries participating in the subgroupalso sent data that can be used to compare thegeographical allocation of FDI stocks under thefirst shot and the UBO criterion. The impact ofthe UBO criterion on the intra-EU vs. extra-EUallocation is very important in DE and in evenmore so in DK. In DE, FDI stocks controlled bythe extra-EU at the end of 2000 increase by33,690 million EUR on a total (first shot, extra-EU) of 106,101 million EUR (i.e. 32%). In DK,the extra-EU stocks according to the UBOcriterion are 29,271 million EUR against 10,141EUR million according to the first shotcriterion. In AT, the impact was lower: 7,315with the UBO vs. 6,967 with the first shot.

397. Within the extra-EU aggregate, theallocation to main partners is also affected. Inthe case of DE, for instance, the USA has66,602 EUR million with the first shot criterionand 87,273 EUR million with the UBO. In DK,the FDI stock held by the USA is 4,175 EURmillion with the first shot and 19.829 EURmillion with the UBO. In AT the effect is13,325 EUR million with the first shot versus23,923 EUR million with the UBO.

EU AGGREGATES AND UBO ALLOCATION

398. The national data calculated by MemberStates (see Annex 6 for the tables with data)cannot be aggregated directly to produce Extra-EU aggregates according to the UBO. If there isan enterprise in DE held by an Austrian andthen ultimately by the USA, both DE and ATwould record the USA as the UBO (for AT theUSA is also the first shot). EU inward stocksheld by the USA would be inflated.

399. To avoid double counting, only the inwardstocks directly held by an extra-EU countryshould be reallocated according to the countryof the UBO. As a result, the size of the inwardstocks controlled by the extra-EU cannotincrease for the effect of the reallocation fromthe first shot to the UBO country of ownership.It would decrease in those cases when the firstshot is extra-EU, but the UBO is intra-EU.

400. However, the value of the FDI liabilitiesdirectly and indirectly controlled by extra-EUcountries (the consolidated Austrian andGerman enterprise FDI parts in the exampleabove) would be correctly assessed only if theFDI stock recorded by the point of entry (theAustrian enterprise in the example) are recordedaccording to the consolidated system.


401. The TF-FDI discussed some definitionalaspects of the UBO. Firstly, the mainmethodological sources do not contain adefinition of the UBO. Secondly, the presentformulation of the methodology in the BDrefers to net assets. Therefore, it was discussedwhether the allocation of FDI inward stocks bycountry of UBO should be limited to equitycapital (including reserves) or should concernalso other capital.

– Concerning the definition of the UBO, theTF-FDI recommends the following definitionthat takes into account the possibility thatdifferent UBOs exist for the same directinvestment enterprise. The ultimate beneficialowners are the first persons proceeding upthe chains (beginning with and including thefirst foreign owner of each FDI position of adirect investment enterprise) that are notcontrolled by another person.

– Concerning whether or not FDI other capitalstocks could/should also be allocatedaccording to the UBO principle, the TF-FDIcould not find a way out towards aconvergent opinion.


402. Special cases. In the chain above the firstforeign owner, there can be an entity that is notmajority-owned by another, although controlmay be exercised by another person with aminority share of ownership. In this case thereare 3 alternatives: 1) Try to identify thecontroller (UBO). 2) Use the majority-ownership criterion and assign the UBO to thecountry of the unit in question because it is notmajority-owned in its turn. 3) Split the countryof UBO among the minority-owners of the unitaccording to the respective percentage ofownership. The last treatment is applied by AT.

403. The TF-FDI also discussed the practicalmethods applied in 3 MS to collect informationabout the UBO. In all 3 MS, the information iscollected from the respondents. DK and DE askinformation about the country of residence ofthe UBO, whereas AT requires more detailedinformation about intermediate owners andpercentages of ownership.

404. Both approaches are acceptable, as theonly information needed for the statistics is thecountry of the UBO. A problem ofcomparability may arise with less informationasked, because respondents may apply differentcriteria for identifying the UBO in cases inwhich there not a clear link of control. Therequest of additional information is certainlyhelpful also for validation and quality checks ofthe information, but implies a higher reportingburden.

405. As a possible source for the informationon the UBO, the TF-FDI discussed a shortreport on the state of advancement of the recentwork on enterprise groups from Eurostat (unitD1) and Member States in the framework of thebusiness register. Information about the grouphead is to become available from the businessregister.55

– The group head, apart from minordifferences, is equivalent to the Ultimatecontrolling institutional unit considered inFATS. The concept of UBO applied in FDIstatistics is also the same, but it differs

because it is applied to FDI stocks and not tothe whole enterprise. In the case of minorityparticipation this source would therefore notbe useful. The same conclusion applies toprivate databases giving information aboutthe ultimate controller of an enterprise.

406. The TF-FDI also analysed the datasupplied by the three MS of the sub-groupabout the effect of the reallocation by UBO ofFDI stocks.

– The impact of the UBO criterion on the intra-EU vs. extra-EU allocation is very importantin DE and is even more so in DK. It is lessimportant in AT, although also in AT mainpartners such as the USA are substantiallyaffected.

407. Finally, the TF-FDI discussed obstaclesand possible solutions for the calculation of EUaggregates based on the UBO allocation.

– An aggregate based on the UBO could not becalculated by summing up the respective partsof the MS because the same UBO can becounted more than once at the EU level.

– A possibility for calculating EU aggregatescould be based on data re-allocated by UBOonly for the part of direct investmententerprises of each MS that are directlyowned by an extra-EU country. However, thevalue of intra-EU stocks controlled by extra-EU countries would be correctly assessedonly if these data reflect the consolidatedvalue of the FDI stocks that is controlled inthe intra-EU by these units.

408. The TF-FDI did not discuss the practicalimplementation of data collection on the UBOprinciple at the EU level. As seven MS do notcollect or plan to collect UBO data, this type ofdiscussion was not possible. Furthermore, the

55 Complete documents on the enterprise group methodology(Chapter 21 of the Business Register Recommendations Manual)is available on Business Methods web site in Eurostat’s CIRCA atthe address:http://forum.europa.eu.int/irc/dsis/bmethods/info/data/new/embs/registers/embs1_5.html

103c ECB




owner” (UBO)

theoretical discussion held about the calculationof EU aggregates showed that this topic shouldbe further analysed in the light of theconclusions and further actions taken in respectof the consolidated system for FDI statistics.


INTRODUCTION

409. This final chapter aims at condensing allconclusions and recommendations that the TF-FDI has addressed along the previous sectionsof this report, while providing with a certainprioritisation for the measures to be takenhereafter.

410. At the time of elaborating follow-upproposals, the TF-FDI always borne in mind thefairly demanding context of changes incollection systems in which most member statesare currently embarked. However, this specificsituation was the main reason why the TF-FDIconsidered this as the right moment to addressall possible minimum requirements that anyfuture system will need to accomplish in theforeseeable future.

411. As it has been the case of previous taskforces, the TF-FDI started its work by trying toidentify the most significant problemsexperienced by national systems at present.With a view to finding the best solutions tothose problems, the TF-FDI undertook adetailed analysis aimed at identifying bestpractices from the methods currently in place inmember states so as to promoting their widespread use across European Member States.

412. The analysis of current practices offered afairly dissimilar picture across individualMember States. The coexistence of differentsystems may be to some extent at odds with theneed to produce consistent European aggregatesand guarantee a certain degree of homogeneity.For this reason, the TF-FDI is of the opinionthat sharing common solutions to commonproblems as well as general shifts towardsfurther standardisation should be deemedpositive steps which may provide substantialbenefits.

413. The TF-FDI mandate encompassed fivemain issues, namely: (i) valuation of stocks; (ii)reinvested earnings; (iii) other capital; (iv)SPEs; and (v) indirect FDI relationships andstatistics based on the ultimate beneficial owner

8 GENERA L CONC LU S I ON S ANDRECOMMENDAT I ON S

(UBO) principle. Across the analysis carriedout by the TF-FDI, it became obvious that thefifth item (mostly in relation to the coverage ofindirect FDI relationships) required a higherprioritisation since it influenced how tointerpret the conclusions of most other items.For that reason, item (v) was split into threeparts, the first two referring to indirect FDIrelationships (from the conceptual and thepractical viewpoints, respectively) and the thirdone referring to the UBO.

414. The most significant issues encountered inthe analysis of these individual items are brieflypresented after this introduction. Subsequently,following the request from the StatisticsCommittee, a prioritisation of the measuresrecommended is provided, in which the mosturgent problems to be solved are highlighted.Finally, those issues that have not been coveredby the TF-FDI (as they were not part of itsmandate) and for which some follow-up workseems recommendable are listed in the finalsection of this chapter.

CONCLUSIONS AND RECOMMENDATIONS FORINDIVIDUAL ITEMS

THE FULLY CONSOLIDATED SYSTEM AND THECOVERAGE OF INDIRECT FDI RELATIONSHIPS415. This chapter aimed at providing a uniqueinterpretation of international standards asregards whether or not and how indirectFDI relationships should be covered by FDIstatistics, purely on conceptual grounds. Thepractical application of this uniqueinterpretation was developed in the followingchapter.

416. Firstly, the TF-FDI agreed that indirectrelationships should undoubtedly be covered byFDI statistics.56

56 This recommendation is very relevant for all FDI items, namelyequity capital, reinvested earnings and other capital. For a moredetailed analysis of the conceptual treatment suggested for thedifferent FDI components, please refer to chapter 1 of thisreport.

105c ECB


8 Generalconclusions

andrecommendations

417. Concerning indirect FDI relationships,two different types were identified: (i) parentcompany – affiliate (without any reciprocaldirect links of ownership above 10%); and (ii)sister/fellow companies, i.e. pertaining to thesame group, but without either direct or indirectlinks of ownership. As a very general summaryof the main recommendations, for the firstgroup of companies, flows and stocks should beclassified by the parent company as outwardFDI and by the (indirectly owned) directinvestment enterprise as inward FDI. In thecase of sister companies, flows and stocksshould be classified as outward FDI by thecountry which provides the investment orgrants the loan and as inward FDI for thecountry receiving the investment/loan.

418. Additionally, both the book-value-basedFDI equity stocks and reinvested earningsshould incorporate indirect relationships to theextreme described in chapter 1. Since allpractical measures to be taken henceforth inconnection with this issue are addressed in thenext section, the TF-FDI just recommends thatall countries agree and put into practice themethodology described in chapter 1 so as toavoid asymmetries and inconsistent treatments.

PRACTICAL ASPECTS RELATED TO THE COVERAGEOF INDIRECT FDI RELATIONSHIPS

SIMPLIFICATION PROPOSALS TOWARDS THECOVERAGE OF INDIRECT FDI RELATIONSHIPS419. The common interpretation of internationalstandards established in chapter one impliesthat indirect FDI relationships should beundoubtedly covered by FDI statistics.Therefore, the TF-FDI investigated into thedifficulties currently faced by member states tocomply with standards. The review of currentpractices in the compilation of FDI statisticsrevealed that EU member states are essentiallysplit into two groups: those that only coverdirect FDI relationships at present and thosethat also cover indirect FDI relationships to alarger or fewer extent in practice.

420. The TF-FDI considers that a fullapplication of the fully consolidated system(FCS) by all countries is virtually unfeasible onpractical grounds. The possibility to restrictFDI statistics to only cover direct FDIrelationships has the advantage of simplicityand a lower burden on respondents (which areobliged to report less information) and, in samecases, also on compilers. In addition, andmainly from the point of view of thecompilation of supranational aggregates, thisapproach would suffer from fewer problemsconcerning the geographical allocation of flowsand stocks. However, this approach would notbe compliant with international standards andthe outcome would offer a lower analyticalvalue, specially considering the increasing roleof special financial vehicles, clearing centres,etc. in the investment strategy of multinationalgroups.

421. In order to find out an alternative thatcould be consistent with international standardsbut also easier to apply in practice (than the fullapplication of the FCS), the TF-FDI exploredother solutions. The most important difficulty ishow to find practical ways for collecting thenecessary information, since the longer thechain of links between companies, the moredifficult it is to get access to the balance sheet offoreign subsidiaries with no direct link to thedomestic mother company.

422. Against this background, the TF-FDIsuggests a simplification of the rules describedin the FCS as the minimum with which allcountries should be compliant. Such a minimumcommon approach would narrow down the riskof asymmetries and would reduce the impact onthe European aggregates of the differentmethodologies applied in member states.

The two admissible simplifications that shouldconstitute the bottom line for all practices at theEU level would be as follows:


(i) The coverage of indirect links ofownership above 50%.57, 58

(ii) The coverage of direct and indirect linksof ownership above 10%, calculated asthe simple product of the subsequent linksof ownership along a chain.

GEOGRAPHICAL DISTRIBUTION OF FDI FLOWS/STOCKS RELATED TO INDIRECT FDI LINKS423. The use of non-fully harmonised criteriafor the geographical allocation of countrycontributions to the European aggregates (inrelation to the existence of indirect FDIownership links) implies a high risk of doublerecording and/or missing information. With aview to avoiding such a risk, the TF-FDIrecommends that, for both reinvested earningsand FDI equity stocks, all (indirect) FDItransactions/positions should begeographically allocated to the immediateaffiliate (direct investment abroad) orimmediate parent company (direct investmentin the reporting economy), i.e. the one withwhich the investor/direct investment enterprisemaintains a direct link of ownership.

424. It is acknowledged that this criterion mayresult in less valuable statistics from theanalytical viewpoint. For this reason, theTF-FDI would encourage countries to collectand publish additional information on thegeographical allocation of FDI flows and stocksbased on the residence of the ultimate beneficialowner, whenever such information were not toodifficult to obtain (see also the UBO section).

EUROPEAN DATABASE ON OWNERSHIPSTRUCTURES425. The TF-FDI analysed the issue from twodifferent points of view: (i) as potential dataproviders; and (ii) as users of the information.

426. From the point of view of potential dataproviders, four countries (DE, FR, ES and IT)investigated their ability to contribute to aEuropean database with the information onownership structures for multinational groups

available in the statistical departments of theirrespective NCBs.

427. The main findings of the TF-FDI pointedout that, while the existence of a centraliseddatabase with information about the structure ofmultinational groups would be seen as a veryuseful tool for the compilation of FDI statistics,the provision of the necessary informationwould imply a number of significant problems,such as (i) a need for additional (monetary andhuman) resources would exist; (ii) therespective national legal backgrounds wouldforce to ensure confidentiality; (iii) there couldbe technical problems to adapt the informationof the database to the structure of multiple anddifferent national collection systems; (iv) theavailability of the data would remain a concern,inter alia, since most public sources only coverlisted enterprises; (v) the specific situation ofeach individual group at different time periodsshould be maintained in the database so as toallow the compilation of statistics across time;(vi) the existence of individual identificationcodes (e.g. ISIN) for each enterprise would bean absolute must.

428. From the point of view of potential users,the TF-FDI suggests that a harmonised andmultilateral solution should be highly welcome.In this context, the TF-FDI received late noticeabout the ongoing project on a EuropeanBusiness Register, currently being developedby Eurostat in collaboration with the ECB. Inthis regard, it is suggested that other bodies, forinstance, the ECB’s WG-BP&ER and theEurostat’s Balance of Payments WG, elaboratethe list of user requirements which wouldpermit that the final product could be used forthe compilation of FDI statistics.


58 Some testimonies in the TF-FDI pointed towards the significantproportion represented by FDI relationships above 50% over thetotal FDI f igures. Additionally, this information is more easilyavailable to domestic respondents in those cases in which there isan obligation to compile consolidated accounts. Therefore, it wasconcluded that efforts should aim at appropriately covering atleast this kind of links.

107c ECB



andrecommendations

VALUATION OF FDI EQUITY STOCKS429. The TF-FDI considered practical problemsfor the implementation of the STC decisionsconcerning how to value FDI equity stocks. Noother alternative methods were considered inthis analysis. The decisions of the STC could besummarised as follows:

1. FDI in listed companies’ shares shall bevalued on the basis of stock exchange pricesin the euro area i.i.p.

2. FDI in non-listed companies’ shares shallbe valued on the basis of book values in theeuro area i.i.p.

3. Book values consist of the application ofownership percentages to the sum ofselected accounts extracted from theliabilities side of the target FDI company’sbalance sheet (according to the commondefinition of OFBV).

4. Two (four) memorandum items will becompiled on a centralised way: inward andoutward euro area FDI based on marketvalues and book values for all types ofcompanies, respectively (with nogeographical or sector details).

5. To this aim, inward and outward FDI equitystocks should be reported to the ECB withan split between listed and non-listed FDIcompanies, and FDI stocks in listedcompanies’ shares should be reported on thebasis of both market values and bookvalues.

430. In reviewing all possible practicalproblems that the implementation of all theseproposals could entail, the TF-FDI consideredthe absence of FDI surveys for the compilationof stock statistics as a major difficulty. Such aproblem has implications on the ability ofcertain member states to implement thedecisions adopted by the STC as regardsvaluation of FDI equity stocks.

Concerning this issue, the TF-FDI is of theopinion that the compilation of FDI equitystocks should be based on information collectedvia FDI surveys, at least on an annual basis. 59

The provision of annual FDI stocks based onaccumulation of b.o.p. flows should bediscontinued as soon as possible.60

In relation to this subject, the TF-FDI ranksthis issue as the first priority for any follow-upwork subsequent to the delivery of this report.

431. Concerning practical solutions to collectthe necessary information to comply with theSTC agreements, the TF-FDI makes thefollowing recommendations:

DISTINCTION BETWEEN LISTED AND NON LISTEDCOMPANIES432. The TF-FDI considered the following aspossible and acceptable information sources:(i) registers of (resident) listed companiesmaintained by stock exchange authorities; (ii)information provided by respondents; (iii)manual distinction based on internal databasesand/or publicly available sources (e.g. financialpress, stock exchange web sites, etc.)

VALUATION OF STOCKS IN LISTED COMPANIES433. The TF-FDI, on the basis of the results ofthe individual national feasibility studiescarried out by six Member States, has come tothe conclusion that:

The collection of FDI equity stocks for listedcompanies on the basis of two differentvaluation methods (market values and bookvalues on the basis of the common definition ofOFBV) can be deemed feasible and not toocostly for member states running FDI surveys.

434. Good/acceptable practices: the mostfeasible way to collect market values and bookvalues for listed companies’ shares is through

59 Some additional information such as a market valuation of FDI inlisted companies could be compiled by other means, aspreviously stated in this report.



the information provided by respondents via theaddition of supplementary questions to the FDIsurveys. Additionally, individual valuationmethods based on stock exchange pricescombined with internal databases and publiclyavailable information have also proved to be aviable way to get this information, especially inthe case of inward FDI. This is also a verypowerful tool to double-check the consistencyand accuracy of the information provided byrespondents, although it is less feasibleconcerning outward FDI (at least, until a fullyoperational Centralised Securities Database isavailable).

435. In addition, the TF-FDI has identified thefollowing non-acceptable practices:

1. To leave the choice to the respondents on thevaluation criterion (market values or bookvalues) they wish to use to report FDIstocks. This can neither ensure theprovision of the necessary information tothe ECB nor guarantee the compilation ofconsistent FDI equity stocks.

2. Application of perpetual inventory methods/accumulation of b.o.p. flows.61 This relieson the reasons previously explained.

COMPILATION OF FDI STOCKS AT T+9 MONTHS436. The TF-FDI concluded that, in thecurrent situation, only four member states arealready in a position to provide pure stocksdata, based on surveys, within the requiredtimeliness. The others can only accumulateflows to the last available stocks (perpetualinventory method), usually adjusted forexchange rate changes, and in few instances forprice changes. The provision of figures with therequired geographical breakdown (shortly on astep-3 basis) does not seem to pose significantproblems for most countries.

REINVESTED EARNINGS437. A review of current practices revealed thatsome member states have not established yetany system to calculate/estimate reinvestedearnings.

The TF-FDI deems the non-inclusion ofreinvested earnings as the most relevantproblem in this area. This difficulty seems to beclosely connected with the lack of FDI surveys,which should be resolved promptly, in line withthe proposal made for FDI equity stocks.

438. All other TF-FDI recommendations arebasically determined by how reinvestedearnings (RIE) are calculated. RIE arecalculated as the difference between twovariables: total profits from current operationsand dividends payable. The first component isnormally available later than the second oneand, hence, RIE (or rather total profits) areoften temporarily estimated from the projectionof total profits as presented in the last availableFDI survey.

TOTAL PROFITS439. International standards prescribe theapplication of the Current OperatingPerformance Concept for the measurement oftotal profits, i.e. the exclusion of extra-ordinarygains and losses. Only five EU countries arecompliant with this so far. In considering theneed to adapt systems to the application of theCOPC, the TF-FDI suggests two types ofinformation sources, both connected with theaccounting statements of the respondents: (i)companies’ public accounts; and (ii) restrictedinformation internally available to thecompanies.

440. Although the split between ordinary andextraordinary gains/losses in accountingstatements is not necessarily consistent withstatistical definitions, it was considered by theTF-FDI to be an immediately available proxyfor the time being. The first information source(public annual accounts) on its own cannot beconsidered as an acceptable proxy for the COPCwithout additional information internallyavailable to respondents, notably, thegeographical breakdown of the information.


109c ECB



andrecommendations

Therefore, a combination of both informationsources would be necessary in any case.

441. The development of the new IAS willimply a more specific definition of thecomponents which may serve as a firm basis forthe harmonisation of MS’ application of theCOPC. However, the development of the newIAS may pose an additional difficulty forcompilers to properly apply the COPC, to theextent that only very exceptional results will beexcluded from the ordinary profits and losses.

The TF-FDI concluded that inconsistenttreatments caused by different practices implyserious distortions for the euro area/EUcurrent account. Therefore, since the datanecessary for a COPC valuation of profit isavailable from the respondents’ accounting:

The same concept for the compilation oftotal profits, namely the COPC, should beused by all MS, i.e. exceptional resultsshould be appropriately excluded from thecurrent account.

As current practices within the EU indicatehow difficult this may be on practicalgrounds, the TF-FDI concluded thatacceptable solutions for the application ofthe COPC should aim at covering at leastthe reduced number of companies whichcontribute the most to extraordinaryresults.

442. Concerning the second recommendation,the experience of some member states (e.g. UKand Germany) is that a reduced number ofcompanies involved in FDI relations (in somecases, holding companies62) contribute to mostof the extraordinary gains/losses. For the othercompanies, the all-inclusive approach may beapplied, since it often provides similar results tothe COPC. An acceptable practice wouldtherefore be to apply the COPC, as a minimum,only to such companies (namely the biggestones plus holding companies) in each MS, andto collect the rest on an all-inclusive basis.

DIVIDENDS443. Time of recording: international standardsprescribe the recording of dividends whenpayable rather than when they are paid. Theforeseeable increase in the use of directreporting through surveys may bring thepractices closer to international standards, asthey are likely to reflect accruals-basedaccounting data. This changeover though willnot take place in the short term. Nevertheless,asymmetries will only occur in short timespans, since the difference between payable andpaid is usually only a matter of time allocationduring a fairly limited period.

444. The treatment of dividends stemming fromexceptional capital gains may be a problem in sofar as it affects the calculation of reinvestedearnings. While exceptional results are notincluded in total profits (according to the COPCdefinition), it is questionable whether or not,once payable, they should be recorded in thecurrent account as income on direct investment.

445. As regards the provision of funds toaffiliates to cover losses, some countries recordit as negative dividends, while some others do itin the financial account.

Conclusions

– Concerning the time of recording, the TF-FDI suggests that member states keep onwith their current practices for practicalreasons and due to the limited impact on theresulting statistics in prolonged time spans.However, member states are requested toswitch from paid-up dividends to payabledividends (where relevant) if they movetowards direct reporting systems.

– Concerning payment of dividends stemmingfrom exceptional capital gains, inaccordance with international standards, theTF-FDI recommends their recording in the

62 Such holding companies normally present very high profits/losses in the financial year compared with a relatively lowvolume of equity capital.


financial account as FDI disinvestments(thus not entering in the calculation of RIE).

– As to contributions to cover losses in directinvestment enterprises, in line withinternational recommendations, the TF-FDIproposes that these transactions should berecorded in the Financial Account, asadditional investment flows and not as directinvestment income.

446. Finally, as a reference to therecommendations related to the treatment ofindirect FDI relationships, it should also benoted that the coverage of reinvested earningsgenerated by indirectly related direct investmententerprises should at a minimum meet one of thefollowing simplified rules: (i) coverage ofindirect links of ownership above 50%63; or (ii)coverage of all (direct and indirect) ownershiplinks above 10%, calculated as the simpleproduct of the subsequent ownership linksalong a chain.

OTHER CAPITAL447. The TF-FDI tried to seek clearer guidanceon the inclusion/exclusion of some borderlinecases within FDI other capital. In particular, theTF-FDI addresses the followingrecommendations:

– Preferred shares should be excluded fromother capital and recorded as DirectInvestment/Equity capital unless they take theform of non-participating shares.64

– Permanent debt (e.g. subordinated loans,perpetual bonds, etc.) should be included inDirect Investment/Other capital, regardlesswhether or not it takes the form of securities.

– Trade credits, financial leasing, and any othertype of intercompany loans should beincluded in Direct Investment/Other capital,while financial derivatives (in accordancewith final agreements between the ECB andthe IMF) should be excluded from FDIstatistics.

– When both parties involved in lendingactivities are MFIs, financial intermediariesor financial auxiliaries, only permanent debtshould be included in Direct Investment/Other Capital. This recommendation couldraise some confidentiality concerns in someMember States, as the granting of permanentdebt to affiliate companies in the bankingsector is usually rather limited. In such cases,the contributions to the European aggregatescould be flagged as confidential.

448. Additionally, the TF-FDI tried to findpractical solutions to collect the necessaryinformation from reporting agencies. Theidentification of best practices necessarilyentailed an analysis of general collectionsystems for FDI statistics notably to avoid therisk that different recommendations could beaddressed for different topics, presenting anincoherent approach across all FDI elements.

In this framework, the TF-FDI is of the opinionthat the two main problems concerning FDIother capital are: (i) the incomplete coverage ofboth transactions and stocks between affiliatedcompanies, such as securities and tradecredits, lending activities between fellowcompanies (i.e. companies with the sameultimate parent company but not belonging tothe same ownership chain), etc. and (ii) thepartial application of the directional principleby some Member States.

449. The general collection methods for FDIflows and stocks can be split into two maincategories: survey-based and settlement-basedsystems. The practical solutions acknowledgedby the TF-FDI for a consistent application ofthe directional principle could most probablynot be deemed very innovative, but no otheralternatives have been found.


64 Non-participating shares do not provide the right to participate inthe distribution of the residual value on dissolution and to thereturn on equity, but rather to the payment of f ixed dividends(interest). They should be recorded under Direct Investment/Other capital.

111c ECB



andrecommendations

450. For survey-based systems, the TF-FDI isof the opinion that the most effective way tocollect the necessary information would be theaddition of questions to the survey form,requesting separately each component of othercapital and taking into account the directionalaspect of the investment (i.e. from parentcompanies to their affiliates and from affiliatesto their parent companies, separately). Onealternative to the direct request of separateinformation from reporters could be to instructthe reporters on how to reclassify (from inwardto outward FDI or vice versa) the fundsprovided by affiliates to their parent companies.This would require that the instructionsaccompanying the survey form wereappropriately amplified to clearly specify inwhich cases and how this reclassificationshould be made and the elements required,giving examples where possible. This secondoption could be more cost-effective (from thestandpoint of both reporters and compiler) onceappropriately set up. However, it could also bemore incomplete and could not suffice to fulfilall requirements (e.g. would not provideseparate information on the split between loansprovided by mother companies to their affiliatesand loans provided by affiliates to their mothercompanies).

451. For settlement-based systems (which,according to a previous recommendation of theTF-FDI, are only suitable to collect FDI flows),the codes used to collect information fromreporters should be expanded (wherenecessary) to include the elements of othercapital required. They should also includeinformation on the direction of the investmentto satisfy the requirements of the directionalprinciple. The TF-FDI also recommends thatinstructions to reporters should also beexpanded to specify the requirements. Adatabase on FDI relationships of residentcompanies, as in Austria, is a useful tool toensure that any other capital transactioninvolving affiliated companies is effectivelyrecorded under direct investment, although itsmaintenance normally requires a significantamount of resources.

SPECIAL PURPOSE ENTITIES (SPEs)452. Due to the increasing role of thesecompanies in the provision of intra-groupfinancing and other services, the TF-FDIexamined the appropriateness of collectingseparate statistics for this type of companiesand, in particular, of an alternative treatment fortransactions and positions in which SPEs areinvolved. More specifically, the TF-FDIconsidered the possibility of “passing through”this kind of enterprises (i.e. do not record eitherassets or liabilities) in those cases in whichSPEs do not carry out any real economicactivity in the territory in which they arelocated.

453. The TF-FDI disregarded both options (i.e.a different treatment and the collection ofseparate statistics) on the grounds thatinternational standards recommend treatingSPEs as any other FDI enterprise (exceptionmade of some special cases65) and do notrequire any separate statistics for this kind ofinstitutions. Additionally, the non-existence ofa single harmonised definition of SPE wouldhamper their identificantion as well as theapplication of different rules for the recordingof transactions and positions in which theseentities are involved.

454. At present, most countries do notdistinguish transactions/positions with non-resident SPEs from those with any other foreigntransactor. Furthermore, most member statesdo not separately identify domestic SPEs intheir regular statistics.66 A change in themethodology applied to these companies wouldbe confronted with the difficulty to calculateconsistent historical series.

65 For instance, in the case of holding companies (for which areclassification in the economic sector of activity isrecommended) or SPEs with a financial nature (for which it isrecommended excluding from FDI statistics intra-group lendingand borrowing vis-à-vis other related corporations with afinancial nature).

66 One country excludes SPEs from national statistics, since, if thatwere not the case, national statistics would be blurred by thevolume of financial transactions between non-resident entitieschannelled through domestic SPE’s.


– The TF-FDI recommends the inclusion oftransactions/positions of/with SPEs or SPE-like companies in b.o.p./i.i.p. reportingconcerning the contributions to the euroarea/EU aggregates.

– Notwithstanding all the practical andconceptual difficulties previously stated, theTF-FDI recommends that the possibility tocollect separate statistics for SPEs continuebeing assessed by both working groups andin the framework of ad-hoc workshops in thefuture.67

– Following the latest decisions of the IMF andthe ECB, SPEs principally engaged infinancial intermediation for a group ofrelated enterprises should be included in thecategory of affiliated financial intermediariesand, therefore, inter-company loans with anyother institution included in the category ofMFIs / affiliated financial intermediariesshould be excluded from direct investmentand should be recorded in otherinvestment.68

ALLOCATION OF FDI INWARD STOCKS BYCOUNTRY OF “ULTIMATE BENEFICIAL OWNER”(UBO)455. Following its mandate, the TF-FDIanalysed the possibility and implications ofclassifying inward FDI stocks by the country ofthe UBO.69 The compilation of FDI statisticsbased on the UBO principle implies allocatingFDI stocks according to residence of the entitythat exercise control on the capital stockconsidered.

456. The TF-FDI assessed the impact ofapplying the UBO principle on the intra/extra-EU allocation and concluded that such impactwas significant in most of the cases analysed(namely on two out of the three countries forwhich this information was available).

DEFINITION OF UBO457. Ultimate beneficial owners are the firstpersons proceeding up along the chain that arenot controlled by any other company. This

definition should be applied at least to equitycapital.

PRACTICAL METHODS TO COLLECT INFORMATIONBASED ON THE UBO PRINCIPLE458. Two approaches were identified by the TF-FDI to apply the UBO through the FDI surveys:(i) direct collection of UBO-based FDI stocksfrom respondents; and (ii) calculation by thecompiler on the basis of more basic information(e.g. on all intermediate owners pluspercentages of ownership) collected fromrespondents.

CALCULATION OF EU AGGREGATES BASED ON THEUBO ALLOCATION459. In the case of the EU aggregates, theapplication of the UBO criterion may implysome double recording related to the inwardFDI stocks held by non-euro area countries. Forthis reason, the TF-FDI recommends that theUBO should only be applied in those cases inwhich EU direct investment companies aredirectly owned by an investor located in anextra-EU country. However, the value of theFDI equity stocks controlled by extra-EUcountries should also reflect the consolidatedvalue of the group, including other affiliates(inside or outside the EU/euro area), in linewith the recommendations related to thetreatment of indirect FDI relationships.

460. The TF-FDI did not hold a conclusivediscussion on the practical ways in which theseproposals could be implemented in practice.Therefore, it is proposed that some further workin this area should be part of the follow-up tothe TF-FDI.


68 Permanent debt should still be recorded in direct investment.69 The mandate did not explicitly request assessing the possibility of

a mandatory collection of statistics based on the UBO principleand, therefore, the TF-FDI did not analyse such a possibility.Being of a more political nature, such a discussion should mostprobably take place in both working groups.

113c ECB



andrecommendations

PRIORITISATION AND TIMING FORIMPLEMENTATION OF THE TF-FDIRECOMMENDATIONS

461. The TF-FDI was requested by theStatistics Committee to provide an appropriateprioritisation of its recommendations, with aclear emphasis on those actions which wereconsidered more urgent for the quality of FDI

statistics. Following this request, the TF-FDIhas divided its recommendations into threecategories according to the potential distortionsthat departing from its recommendations couldentail for the European aggregates: high,medium and low importance, respectively.

462. Additionally, a second dimension refers tothe effort that each individual action would

Importance

Timeframe High Medium Low

Short-term • All countries should start compiling • Contributions toFDI equity stocks and reinvested cover losses ofearnings on the basis of the results direct investmentof FDI surveys, at least annually. 1) enterprises should

• FDI equity stocks should be collected be recorded in theseparately for listed (both book 2) and financial account.market values) and non-listed companies.

• All indirect FDI relationships 3) shouldbe conceptually treated in accordancewith the interpretation of standardsoutlined in chapter 1.

• All (indirect) FDI transactions/positionsshould be geographically allocatedto the immediate affiliate or parentcompany.4)

Medium-term • The COPC should be used by all MS. 5) • Contribute to the development of a• The components of other capital should European database with information

be identified on the basis of the about the structure of multinationalrecommendations provided in chapter 6. groups.

• Payment of dividends from exceptionalcapital gains should be recorded in thefinancial account (thus not entering inthe calculation of RIE).

Long-term • Indirect FDI relationships 6) should • Dividends shouldcover in practice (as a minimum) be recorded wheneither (i) indirect links of ownership payable ratherabove 50%; or (ii) direct and indirect than when paid.links of ownership above 10%,calculated as the product of thesubsequent links of ownership alonga chain.

• The directional principle shouldbe (fully) applied by all member statesfor FDI flows and stocks.

1) Exception made of provisional results to be provided at T+9 and real-state investments. The following non-acceptable practices shouldbe abandoned: (i) to leave the choice to the respondents on the valuation criterion (market values or book values); and (ii) the applicationof a perpetual inventory method/accumulation of b.o.p. flows to compile stocks.2) Based on the common definition of own funds at book value.3) To the extent that they can be identified, considering the practical difficulties existing at present, as addressed in chapter 2 of this report.4) For both reinvested earnings and FDI equity stocks.5) MS may focus on a reduced number of companies (the biggest ones and/or holding companies) to perform the distinction betweenordinary and extraordinary gains and losses.6) For all elements of FDI statistics (namely equity capital, reinvested earnings and other capital).

Table 23 Matrix of conclusions: priorit isation and timing for implementation of theTF-FDI recommendations


require from member states and the time lagwith which the application of itsrecommendations could be reasonablyexpected. On that basis, the proposed actionscould also be classified in three additionalcategories: short, medium and long-term. TheTF-FDI did not intend to define (in terms ofmore specific timing) the deadlinescorresponding to each slot, since this wasconsidered out of its mandate.

463. By combining both dimensions (i.e.importance and timeframe), the most significantrecommendations of the TF-FDI have beenintegrated in a matrix for illustrative purposes.

ISSUES FOR FOLLOW-UP WORK

464. Due to its limited time horizon, the TF-FDIcould not expand the topics defined in itsmandate with other subjects identified in thecourse of its investigations. Additionally, theTF-FDI did not hold discussions on morestrategic issues. Therefore, it is proposed thatsome work could follow the delivery of thisreport in the following areas:

– Elaborate an implementation calendar withspecific deadlines to put therecommendations of the TF-FDI in practice.

– Monitor on a regular basis theimplementation status of the TF-FDIrecommendations as well as other mattersrelated to FDI (e.g. exchange of experiencesand information on FDI) through, forinstance, the regular meetings of the workinggroups and/or ad-hoc workshops. Among thedifferent issues to be considered in the future,the TF-FDI recommends that the possibilityto collect separate statistics for SPEscontinue being assessed in the future.70

– Develop a twofold monitoring task, whichshould focus on: (i) the definition of the newinternational accounting standards; and (ii)the update of the IMF Balance of PaymentsManual. This monitoring task should aim at

promoting further convergence betweenstatistical and accounting standards, whilekeeping in mind that FDI statistics shouldalways be able to serve analytical needs fromthe macroeconomic viewpoint.71

– Explore practical ways to put the proposals tocompile statistics based on the UBO inpractice.

– Elaborate the list of user requirements for theEuropean Business Register project currentlybeing developed by the Eurostat’s BusinessStatistics Directorate. Such a contributionshould ensure that the final product will havethe necessary features for the compilation ofFDI statistics.


71 In particular, the TF-FDI discussed two alternatives to try toapproximate statistical rules to accounting standards: (i) changethe 10% rule defining all FDI relationships to a 20% criterion;(ii) consider only indirect FDI relationships over 50% (i.e.restrict the coverage of indirect relationships to cases ofmajority control). The TF-FDI tentatively expressed apreference for the second option, which is already a practicalsimplification addressed in this report.

115c ECB


AnnexesL I S T O F ANNEXE S


465. The issue of the compilation of FDI stocksat T+9 months was raised by the CMFB inconnection with the compilation of theinternational investment position (i.i.p.)statistics of which they are an integral part.I.i.p. statistics have to be provided to Eurostatand to the ECB within nine months of the end ofthe year to which the data relate. In theframework of the Special Data Disseminationstandard (SDDS) of the IMF, the timelinessrequired for the i.i.p. is nine months after theend of the reference year, provided that SDDSsubscribers publish quarterly external debtstatistics; otherwise, the timeliness for the i.i.p.is six months.

466. During the meeting of the TF-FDI inFrankfurt in February 2003, a brief stock-taking exercise was conducted among theparticipants on the possibility of providingannual FDI stocks statistics within this nine-month time-frame.

467. The outcome of the discussion was that inthe current situation, only four Member Statesare already in a position to provide pure stocksdata, based on surveys, within the requiredtimeliness. The others can only continueaccumulating flows to the last available stocks(perpetual inventory method), usually adjustedfor exchange rate changes, and in few instancesfor price changes.

468. The country details are as follows:

– FR uses the perpetual inventory method, withan adjustment for exchange rate and marketprice changes.

– ES collects pure stocks only for the MFIsector and for the other sectors’ externalloans; these are available within theprescribed timeliness. The perpetualinventory method will be applied to the equitycapital position of the other sectors,

– GR runs a survey system in whichquestionnaires are sent out in April and haveto be returned one month later, together withthe balance sheet, the profit and loss

statement and the distribution of profitsstatement of the company. Economic sectorand geographical breakdowns are available.The T+9 deadline can be met.

– DE uses the perpetual inventory method, withan adjustment for exchange rate changes andfor some other changes. Pure stocks for thei.i.p. are available at T+15 months.

– IE runs a quarterly survey of stocks and hastherefore no problem to provide annual databy end-September.

– NL prepares an i.i.p. statement within sixmonths, along the same line as DE and FR.

– BE: in the present system, the i.i.p. isproduced as in DE, at T+9, except for theMFI and general government sectors forwhich it is available at T+6. Adjustments aremade for exchange rate and price changes. Inthe future, BE intends to implement aquarterly survey.

– PT: the situation is equivalent to that of FR.– FI runs an annual survey for equity capital

and monthly and annual surveys for othercapital (a perpetual inventory is used for theequity component of the quarterly i.i.p.). Theresults of the annual surveys are available atT+9.

– DK: the results of the annual surveys areavailable at T+9.

– IT uses the perpetual inventory method withexchange rates and price adjustments in thecurrent system.

– SE runs an annual sample survey whoseresults are only available at T+10/11 monthsand will have to make estimates to compileFDI stocks at T+9.

– AT: since survey results are available only atT+18, stocks for T+9 can only be calculatedby adding annual flows to the latest availablestocks. A breakdown by individual countrieswould be available.

469. Regarding the geographical breakdown ofthe FDI stocks, a majority of countries willhave no difficulty for providing the level 1details of Eurostat at T+9, albeit on a veryprovisional basis. ES, however, can onlyprovide an intra/extra euro area/EU breakdown.

1 COMP I L AT I ON O F F D I S TO CK S AT T+9 MONTH S

117c ECB


Annexes

SE pointed out that the fact of using a samplesurvey makes it difficult to capture informationconcerning small counterpart countries.

470. On the whole, it appears that the provisionof FDI stocks at T+9, with a level 1geographical breakdown, i.e. according to theECB’s narrow list for step-3, is not seen ascausing major difficulties. However, this wouldbe in many cases on the basis of flowsaccumulated to the last available stocks.


471. In this annex, the definitions andrecommendations concerning Other Capital inForeign Direct Investment flows and stockswill be described. These definitions andrecommendations are based on the IMF’sBalance of Payments Manual, 5th Edition(1993; BPM5), the OECD BenchmarkDefinition of Foreign Direct Investment, 3rdEdition (1996; Benchmark) and ad hocrecommendations by international institutionssuch as the IMF, the OECD, the ECB, etc. Thedefinitions and recommendations construe thebasis for the determination and identification ofbest practices in Sections 3.

472. First, the broad concepts of foreign directinvestment capital and other capital transactionswill be defined. Second, the directionalprinciple, an important item of Other Capital, isdiscussed. Third, different items of OtherCapital, such as debt securities and tradecredits, are described. Fourth, intercompanytransactions involving Monetary and FinancialInstitutions (MFIs) are investigated.

DIRECT INVESTMENT CAPITAL AND OTHERDIRECT INVESTMENT CAPITAL

473. In general, according to BPM5, “Directinvestment capital” is:

“capital provided (either directly orthrough other related enterprises) by adirect investor to a direct investmententerprise or capital received from a directinvestment enterprise by a direct investor.

(…) Direct investment capital does notinclude funds provided by, or receivedfrom, any other source – including fundsfor which the direct investor merely makesthe arrangements or guarantees repayment(e.g., loans from outside parties to anincorporated direct investment enterprise.”(BPM5, paragraph 368)

“The components of direct investment capitaltransaction, which (…) are recorded on adirectional basis (i.e., resident direct investmentabroad and non-resident direct investment in therecording economy), are equity capital,reinvested earnings, and other capitalassociated with various intercompany debttransactions.” (BPM5, paragraph 369).

474. Other direct investment capital (orintercompany debt transactions) covers

“the borrowing and lending of funds –including debt securities and suppliers’ credits– between direct investors and subsidiaries,branches, and associates. The borrowing andlending are reflected in intercompany claimsand liabilities (receivables and payables),respectively. Both loans to subsidiaries fromdirect investors and loans from subsidiaries todirect investors are included. In contrast to thetreatment of other investment, no distinction ismade between short- and long-terminvestment”. (BPM5, paragraph 370)

475. The OECD recommends that:

“inter-company flows – with the exceptionof certain flows between affiliated banks,affiliated financial intermediaries (e.g.security dealers), and Special PurposeEntities (SPEs) with the sole purpose ofserving as financial intermediaries – beencompassed within the scope of foreigndirect investment transactions. Such flowsinclude those flows routed through SPEs( . . . ) .

OECD also recommends that loansprovided by all subsidiaries to parentcompanies be included in foreign directinvestment transactions, with theexception of financial loans provided bybanks and other financial intermediaries tothe parent bank. Recognising that somecountries may not include (or may net out)transactions through SPEs, and/or inter-company debt flows, and loans providedby subsidiaries to parents within direct

2 D E F I N I T I ONA L I S S U E S R E L AT EDTO F D I / O TH ER C A P I TA L

119c ECB


Annexes

investment transactions, OECDrecommends that countries provideinformation on gross flows to facilitateinternational comparability of directinvestment data.” (Benchmark, paragraph39).

DIRECTIONAL PRINCIPLE

476. The application of the directionalprinciple is one of the main items of theMandate of the TF-FDI. The IMF states thefollowing on reverse investments and thedirectional principle:

“Reverse investment in the form of otherinstruments (other than equity capital)should be recorded, under direct investmentin reporting economy – other capital ordirect investment abroad – other capital. Incases in which the equity participation is atleast 10 percent in both directions, twodirect investment relationships areestablished. Such transactions are recordedas direct investment claims and liabilities inboth directions; that is, as direct investmentin reporting economy and as directinvestment abroad, for each economy asappropriate.” (BPM5, paragraph 371)

“Direct investment is classified primarily ona directional basisresident direct investmentabroad and non-resident investment in thereporting economy and is subdivided intoequity capital, reinvested earnings, andother capital. (paragraph 330). Unlike otherfinancial investment, direct investment isnot recorded in the BOP on a straight asset/liability basis. Instead, direct investment isrecorded on a directional basis – residentdirect investment abroad and non-residentdirect investment in the reporting economy.”(IMF’s Balance of Payments Textbook,paragraph 529).

477. This means that all transactions in othercapital between a direct investor and its directinvestment enterprise should be recorded under

direct investment abroad in the BOP of thedirect investors country and as directinvestment in the reporting economy in the BOPof the country where the direct investmententerprise is resident. The investment directionis thus determined by the direction of the equityinvestment. Consequently, from the point ofview of the economy of the direct investor,loans to direct investment enterprises aretreated as direct investment abroad as well asclaims by the direct investment enterprise on itsinvestor. This application of the directionalprinciple may lead to negative FDI stocks incases where loans by the direct investmententerprise to its direct investor outweigh theamount of equity capital and loans provided bythe direct investor.

478. With respect to transactions when sistercompanies are involved, the Balance ofPayments Textbook states:

479. “When a direct investment enterpriseinvests in an enterprise related to its directinvestor, this investment is recorded, by theeconomy providing the investment, as residentdirect investment-abroad and, by the economyof the enterprise receiving the investment, asdirect investment-reporting economy.” (IMF’sBalance of Payments Textbook, paragraph 529).

ITEMS IN OTHER CAPITAL480. This paragraph describes certaindifferent items, which can be included in OtherCapital. The following items are explored:

– Debt securities (including financialderivatives)

– Preferred shares– Trade credits– Financial leasing– Deposits– Intercompany loans (assets in intra-group

accounts)– Permanent debt


481. According to BPM5, paragraph 370 (seeabove), intercompany debt securities and trade(or suppliers’) credits must be included underFDI-Other Capital. The other items listed aboveare not covered explicitly in the chapter onForeign Direct Investment in the Manual(Chapter 18), though intercompany transactionsincluding these items can be viewed upon asFDI-Other Capital.

DEBT SECURITIES

482. Intercompany debt securities are includedin FDI-Other Capital (BPM5, paragraph 370)and comprise several types of instrumentsthat are defined in BPM5 under “PortfolioInvestment”:

483. “Debt securities cover

1. bonds, debentures, notes, etc.;2. money market or negotiable debt instruments,

and3. financial derivatives or secondary

instruments, such as options, that usually donot extend to actual delivery and are utilisedfor hedging of risks, investment, andtrading purposes.)” (BPM5, paragraph 389.)

484. In 2000, the IMF published a new bookleton financial derivatives and modifications to theexisting material on financial derivatives inBPM5 (“Financial derivatives, a supplement tothe Balance of Payments Manual”, IMF, 2000).Initially, intercompany financial derivativeswere classified as direct investment. In themeantime, however, the IMF and the ECBrecommended excluding transactions infinancial derivatives between affiliatedenterprises from direct investment.

485. The summary of the March 2002 meetingof the Working Group on Balance of Paymentsand External Reserves (WG-BP&ER) states that“the WG-BP&ER (…) “supports the approach(...) to report financial derivatives betweenaffiliates under the functional category of“financial derivatives” rather than under “FDI-

Other Capital”, noticing that in practice someFDI transactions/positions may, however, stillbe recorded under FDI.”

PREFERRED SHARES

486. Preferred shares are a hybrid form offinancing, combining features of debt andcommon shares. In the event of liquidation, apreferred shareholder’s claim on assets comesafter that of creditors but before that of commonshareholders. Usually, this claim is restricted tothe par value of the share (equals the maximumsettlement of the principal amount). Preferredshares have the following features:

– A cumulative feature, providing for unpaiddividends in any one year to be carriedforward. Before a company can pay adividend on its common shares, it must paydividends in arrears on its preferred shares.

– A participating feature, allowing preferredshareholders to participate in the residualearnings of the company. Essential is thatpreferred shareholders have a prior claim onincome and an opportunity for additionalreturn if the dividends to commonshareholders exceed a certain amount.

– Voting power: because of their priority claimon assets and income, preferred shareholdersnormally are not given a voice inmanagement. The voting power that preferredshareholders may be granted is usuallyrestricted to issues relating to their status.

487. Preferred shares have no maturity. Thisdoes not mean, however, that most preferredshares issues will remain outstanding forever,because provision for retirement of the sharesinvariably is made. For instance, manypreferred shares issues provide call prices, asinking fund, and convertibility.

488. A straight preferred stock issue has nofinal maturity; in essence, it is a perpetual loan.From the standpoint of creditors, preferred

121c ECB


Annexes

stock adds to the equity base of the companyand thereby strengthens its financial condition.The additional equity base enhances the abilityof the company to borrow in the future. Theimplicit cost of preferred stock financing, fromthe standpoint of investors penalizing theequity-capitalization rate of the common stock,may be somewhat less than that of debtfinancing. To the extent that investors areapprehensive over legal bankruptcy, they wouldregard debt as a riskier form of leverage. Unlikecreditors, preferred stockholders cannot force acompany into legal bankruptcy. Therefore, thepotential costs of bankruptcy are not a factor asthey are with debt financing.

489. The mix of perpetuity and the optionfeatures, makes preferred shares a hybridform.72

490. Non-participating preferred shares can bedescribed as a type of preferred share where theholder has no entitlement to a share in any of theexcess returns of the issuing entity. Return onequity comprises only a dividend payment(usually at a fixed rate) which is calculatedaccording to a pre-determined formula. Paymentof interest is not dependent upon earnings of theissuer.

491. The treatment of preferred shares in BPM5is discriminated between participating and non-participating preferred shares:

“non-participating, preferred shares […] aretreated as debt securities and included underdirect investment-other capital.” (BPM5 ,paragraph 369).

492. The BPM5 is dealing with preferred sharesin other paragraphs as well:

“Shares, stocks, participation (..) usuallydenote ownership of equity. Preferred stockof shares, which also provide forparticipation in the distribution of theresidual value on dissolution of anincorporated enterprise, are included.”(Paragraph 388 in the BPM5)

493. Paragraph 390 in the BPM5 is dealing withinstruments that

“usually give the holder the unconditionalright to a fixed money income or contractuallydetermined variable money income. (..) Withthe exception of perpetual bonds, bonds anddebentures also provide the holder with theunconditional right to a fixed sum as arepayment of principal on a specified data ordates. Included are non-participatingpreferred stocks or shares (...).”

TRADE CREDITS

494. In BPM5, trade credits are included inOther Capital and are defined in the chapter on“Other Investment”:

“Trade credits consist of claims and liabilitiesarising from the direct extension of credit bysuppliers and buyers for transactions in goodsand services and advance payments for work inprogress (or to be undertaken) that is associatedwith such transactions (loans to finance tradeare not included as these are classified underloans). In the absence of actual data, tradecredits may be measured by the differencebetween entries for the underlying transactionsin goods and services, which are recorded as ofthe dates when ownership changes, and theentries for payments related to thesetransactions. Although frequently short-term innature, trade credits and advances aresubdivided into short- and long-termcategories’. For trade credits between affiliatedenterprises recorded under other capital, nomaturity breakdown is required. (BPM5 ,paragraph 414)

72 This information is extracted from “Financial Management andPolicy” by James Van Horne, Prentice Hall, Tenth Edition, 1995,pages 576-580.


FINANCIAL LEASING

495. In BPM5, financial leases are brieflydescribed in the chapter on “Other Investment”.In case of financial leasing between affiliatedcompanies, this paragraph can be applied to FDIflows and stocks.

“Financial leases are included under loansbecause such arrangements are taken aspresumptive evidence that a change in theownership of goods has occurred (see BPM5,paragraph 206). The financial leaseessentially is a method by which the lesseefinances the purchase of goods. The financiallease entails a financial claim, which is theasset of the lessor and the liability of thelessee. At the time the imputed change inownership occurs, the market value of thegoods is recorded and counterpart entries, asassets or liabilities, are made in the financialaccount. In subsequent periods, the actuallease payment is divided into interest, whichis recorded in the current account as incomepayable or receivable, and principal (debt)repayment, which is recorded in the financialaccount and reduces the value of the asset ofthe lessor and the liability of the lessee.”(BPM5, paragraph 417)

496. A very comprehensive description offinancial leasing can be found in theBenchmark:

“Wherever an operator (the lessee) employsassets which are held under financial lease (asdistinct from operating lease) rather thanbeing owned outright, the legal owner of theassets (the lessor) should be regarded asmaking a loan to the lessee, which the lesseeuses to buy the assets. If this arrangement isbetween a direct investor and its branch,subsidiary or associate, the loan should beregarded as direct investment, and be treatedaccording to the provisions of this benchmarkdefinition, in the same way as a conventionalloan would be regarded and treated.

497. In BPM5 (paragraph 206), it is stated thatfinancial leases are lease arrangements

“for a capital good for most or all of itsexpected economic life”, during which time“the lessor expects to recover most or all ofthe cost of the goods and the carryingcharges”. Therefore, a financial leasearrangement is to be taken as presumptiveevidence that a change of ownership isintended. A change of ownership is imputedbecause the lessee assumes all rights, risks,rewards, and responsibilities of ownership inpractice, and from an economic point of view,can be regarded as the de facto owner.

498. The financial lease essentially is a methodof financing the purchases of the good by thelessee (as opposed to taking out a loan for thepurchase). The OECD defines a financial leaseas

“a contract involving payments over a basicperiod (during which the agreement cannot beterminated) sufficient in total to cover in fullthe capital outlay of the lessor together withall subsidiary or financing cost and to givesome profit to him. This obligatory perioddoes not exceed the estimated useful life ofthe asset. The asset is selected by the lessee(...). The cost of maintenance and repair, onthe subject of the lease, and all risksconnected therewith, are borne by the lessee.At the end of the primary lease, the contractmay be extended for a further (secondary)period at a much reduced and perhaps purelynominal rental, or the asset may be sold to thelessee at a very low and perhaps purelynominal price, or the asset may be sold to athird party with some, or perhaps most, of theproceeds from the sale being passed on to thelessee as a “rebate of rental”.” It is hoped thatin due course a common definition offinancial leasing will be established.”(Benchmark, paragraph 57)

123c ECB


Annexes

DEPOSITS

499. Deposits are described as follows in theManual:

“Deposits comprise both transferable andother deposits. Transferable deposits consistsof deposits that are exchangeable on demandat par without restriction or penalty, freelytransferable by check or giro order, andotherwise commonly used to make payments.Deposits may be denominated in domestic orforeign currencies. Other deposits include allclaims reflecting evidence of deposit otherthan transferable deposits. Typical examplesare non-transferable saving deposits; timedeposits; and shares (evidence of deposit) –which are legally (or practically) redeemableon demand or on short notice – in savings andloan associations, credit unions, buildingsocieties, etc.” (BPM5, paragraph 421)

INTERCOMPANY LOANS (ASSETS AND LIABILITIESIN INTRA-GROUP ACCOUNTS) AND PERMANENTDEBT

500. Though the following items are underinvestigation in the survey, there are no cleardefinitions and recommendations describingthem. These items are intercompany loans(assets in intra-group accounts) and permanentdebt. Intercompany loans are, in the survey,defined as current accounts between groupcompanies.

501. Permanent debt is typically provided for anindefinite period of time without a fixedrepayment obligation. It can be seen as part of acompany’s own funds as it is presented on theliabilities side of the company’s balance sheeton a permanent basis. It can take the form ofdebt securities or loans.

502. Subordinated loans are loans that rankbehind debt senior to these loans with respect tothe claim on assets. In the event of liquidation,subordinated debenture holders usually receivesettlement only if all senior creditors are paid

the full amount owed to them. These holdersstill would rank ahead of preferred shareholdersin the event of liquidation.

INTER-COMPANY TRANSACTIONS BETWEENAFFILIATED MFIs AND AFFILIATED FINANCIALINTERMEDIARIES

503. The Manual describes inter-companytransactions between affiliated MFIs andaffiliated financial intermediaries as follows:

“Inter-company transactions betweenaffiliated banks (depository institutions) andaffiliated financial intermediaries (e.g.,security dealers) – including SPEs with thesole purpose of serving as financialintermediaries – recorded under directinvestment capital transactions are limited tothose transactions associated with permanentdebt (loan capital representing a permanentinterest) and equity (share capital) investmentor, in the case of branches, fixed assets.Deposits and other claims and liabilitiesrelated to usual banking transactions ofdepository institutions and claims andliabilities of other financial intermediaries areclassified, as appropriate, under portfolioinvestment or other investment. The stock offoreign assets and liabilities of banks andother financial intermediaries (internationalinvestment position) should be treated in aparallel manner.” (BPM5, paragraph 372).

504. Because there was much confusion aboutthis paragraph, the IMF Committee on Balanceof Payments Statistics (BOPCOM) has releasedsome final decisions, after consulting with theECB’s WG-BP&ER and the OECD’s WorkingParty on Financial Statistics (WPFS), on thesubjects described73.

73 See the OECD Document DAFFE/MC/STAT(2002)2, released atthe Working Party on Financial Statistics’ meeting on 20 and 21February 2002.


Present position

S122 S123 S124 S125 Other

S122 Excluded Excluded Included Included IncludedS123 Excluded Excluded Included Included IncludedS124 Included Included Included Included IncludedS125 Included Included Included Included IncludedOther Included Included Included Included Included

Proposed position

S122 S123 S124 S125 Other

S122 Excluded Excluded Excluded Included IncludedS123 Excluded Excluded Excluded Included IncludedS124 Excluded Excluded Excluded Included IncludedS125 Included Included Included Included IncludedOther Included Included Included Included Included

Table 1 Coverage of capital transactions (other than in equity and permanent debt)in FDI statist ics

505. The final decisions on the recommendedtreatment of transactions with affiliatedfinancial intermediaries are as follows:

“The BPM5 definition of the scope ofenterprises included under “banks and otherfinancial intermediaries such as securitydealers” should be clarified as beingequivalent to the following SNA93 financialcorporations sub-sectors: other depositorycorporations (other than the central bank);other financial intermediaries, exceptinsurance corporations and pension funds;and financial auxiliaries. As a result, SPEsprincipally engaged in financialintermediation for a group of relatedenterprises would be encompassed in thatdefinition.

The implications of the above clarification arethat financial (and investment income)transactions between two affiliatedenterprises that are part of (1) otherdepository corporations (other than thecentral bank); (2) other financialintermediaries, except insurance corporationsand pension funds; or (3) financial auxiliaries

would be excluded from FDI except fortransactions in the form of equity capital orpermanent debt.” (DAFFE/MC/STAT(2002)2,paragraph 5)

506. A simplified version of the IMF matrixchart of this proposal shows the present andproposed situation.

507. To further dissipate any confusion, it isclear from the preceding tables that e.g.deposits by non-banks (S125+Other) withaffiliated banks (S122+S123+S124) have to berecorded under Other Capital. Loans by banksto affiliated non-banks have to be recordedsimilarly.

Note: S122 MFIs other than central banks S123 Other financial intermediaries (exc. 125) S124 Financial auxiliaries S125 Insurance andpension funds 507. To further dissipate any confusion, it is clear from the preceding tables that e.g. deposits by non-banks (S125+Other)with affiliated banks (S122+S123+S124) have to be recorded under Other Capital. Loans by banks to affiliated non-banks have to berecorded similarly.

125c ECB


Annexes

OTHER CAPITAL AT BE DE DK ES FI FR1) IT NL PT SE UK 12DIRECTIONAL PRINCIPLEA1 At least partially 2) AT DE DK ES FI FR IT NL PT SE UK 11

No BE 1A11 Technology BE 1

Theory 0Other 0

A2 I/O AT DE DK ES FI FR IT NL PT SE UK 11In 0Out 0

A21 In F/S AT DE DK ES FI FR IT NL PT SE UK 11Flows 0Stocks 0

A22 Out F/S AT DE DK ES FI FR IT NL PT SE UK 11Flows 0Stocks 0

A23A24

A3 Flow Yes AT DE DK ES FI FR IT NL PT SE 10No UK 1

A31 Yes UK 1No

A4 Stock Yes AT DE DK ES FI FR IT NL PT SE UK 11No 0

A41 Yes 0No 0

A411

A5 Comment

A6 All items? Yes AT DK FI FR IT SE UK 7No DE ES NL PT 4

A7 Comment

Table 1 Summary of the answers to the questionnaire on Other CapitalDirectional principle

1) As of 2004.2) From the discussions on the answers to the questionnaire, it appears that the initial “yes” answers regarding the application of thedirectional principle should be understood in many cases as meaning “yes, at least partially”.

3 SUMMARY O F TH E AN SWER S TO TH EQUE S T I ONNA I R E ON O TH ER C A P I TA L


Table 2 Summary of the answers to the questionnaire on Other CapitalItems included in other capital

ITEMS INCLUDEDIN OTHER CAPITAL AT BE DE DK ES FI FR IT NL PT SE UK 12B 1 Debt Sec FI IT PT SE UK 5

Trade Cr DE DK ES FI IT NL PT SE UK 9Fin Lease AT BE DE DK ES FI IT PT SE UK 10Derivatives DK 1I/C AT BE DE DK ES FI FR IT NL PT SE UK 12Deposits BE DK FI FR IT NL SE UK 8Perm Debt AT BE DE DK ES FI IT PT UK 9Other 0

B 21 Debt Sec F/S IT PT SE UK 4Flows 0Stocks FI 1

B211B212B22 I/O FI IT PT SE UK 5

In 0Out 0

B23 L/S F/S FI IT SE 3L/S Flows 0L/S Stocks 0L/S No PT UK 2

B231B232B233B24 Bonds IT PT SE UK 4

Notes IT PT SE UK 4Prom Notes IT SE UK 3Oth Instr IT SE UK 3Other 0

B25 Bonds IT PT SE UK 4Notes IT PT SE UK 4Prom Notes IT SE UK 3Oth Instr IT SE UK 3Other 0

B 31 Trade Cr F/S DE ES FI IT NL PT SE UK 8Flows 0Stocks DK 1

B311B312B32 I/O DE DK ES FI IT NL PT SE UK 9

In 0Out 0

B 41 Fin Lease F/S AT DE ES FI IT PT SE UK 8Flows BE 1Stocks DK 1

B411B412B42 I/O - BE DE DK ES FI IT PT SE UK 10

In 0Out 0

B43 L/S Yes 0L/S No - BE DE - ES FI IT PT SE UK 10

B 51 Derivatives F/S NL 1Flows 0Stocks DK 1

B511B512B52 I/O DK NL 2

In 0Out 0

B53 L/S Yes 0L/S No DK NL 2

B 61 Interco F/S AT BE DE DK ES FI FR IT NL PT SE UK 12Flows 0Stocks 0

B611B612B62 I/O AT BE DE DK ES FI FR IT NL PT SE UK 12

In 0Out 0

B63 L/S Yes DE DK FR NL SE 5L/S No ES FI IT PT UK 5

B 71 Deposits F/S BE DK FI FR IT NL SE UK 8Flows 0Stocks 0

127c ECB


Annexes

Table 3 Summary of the answers to the questionnaire on Other CapitalValuation principles, and other capital data on MFIs

VALUATION PRINCIPLES AT BE DE DK ES FI FR IT NL PT SE UK 12C1 Flows Book BE DK ES IT UK 5

Market AT DE FI FR PT SE 6Both NL 1

C11

C2 Stock Book AT BE DE DK ES FR IT PT UK 9Market SE 1Both FI NL 2

C21

C3

SPEs and Other Financial IntermediariesD1 Flows Yes AT BE DE DK ES FI FR IT NL PT SE UK 12

No 0D11 Book BE DK ES IT UK 5

Market AT DE FI FR PT SE 6Both NL 1

D111D12 I/O AT BE DE DK ES FI FR IT NL PT SE UK 12

In 0Out 0

D13 Debt Sec FI IT PT SE UK 5Derivatives DK NL 2Deposits BE DK FI FR IT NL SE UK 8Fin Lease BE DE DK ES FI IT PT SE UK 9Other AT BE DK ES FR NL PT 7

D2 Stock Yes AT BE DE DK ES FI FR IT NL PT SE UK 12No 0

D21 Book AT BE DE DK ES IT PT UK 8Market SE 1Both FI NL 2

D211D22 I/O AT BE DE DK ES FI FR IT NL PT SE UK 12

In 0Out 0

D23 Debt Sec FI IT PT SE UK 5Derivatives DK NL 2Deposits BE DK FI FR IT NL SE UK 8Fin Lease DE DK ES FI IT PT SE UK 8Other AT BE DE DK ES FR NL PT 8

D3

TREATMENT OF OTHER CAPITAL DATA ON MFIsE1 Sub Loans F/S AT DK FI PT 4

Flows BE 1Stocks UK 1No DE ES FR IT NL SE 6

E11E12E13

E2 Deposits F/S DK ES 2Flows 0Stocks PT 1No AT BE DE FI FR IT NL SE UK 9

E21E22E23

E3


THE GERMAN SYSTEM

508. German balance of payments statistics areup to now only partially designed according tothe directional principle. Therefore onlyapproximate information can be derived fromexisting data sources, and only in case of loansbetween affiliated companies.

509. In case of shares and bonds it is notpossible to distinguish “reverse transactions”,therefore these transactions can only be classi-fied as portfolio investment according to theasset/liability-principle.

DIRECTIONAL PRINCIPLE FOR SHORT-TERMLOANS

510. Short-term loans in the b.o.p. are derivedfrom the corresponding stocks. Germanresidents are required to report all claims on andliabilities to non-residents arising fromfinancial and trade credits if total claims or totalliabilities exceed €5 mill (DM 3 mill prior to2002) at the end of the month (“Z 5 stocksstatistics”).

511. Consequently, in the case of short-termloans granted by German non-banks to foreignnon-banks it is possible to differentiate in a firststep between:

– short-term loans between non-affiliatedcompanies (these loans are classified as otherinvestment)

– and short-term loans between affiliatedcompanies (these loans are classified as directinvestment).

512. Moreover, German residents have to reportassets held abroad and non-residents’ assets inGermany are required to be reported by Germanenterprises in which non-residents hold sharesor voting rights (“capital links statistics”).These data are requested by separate forms on ayearly basis. Referring to the capital linksstatistics all reporting companies are classified

4 A P P L I C AT I ON O F T H E D I R E C T I ONA L P R I N C I P L EI N D I F F E R EN T COUNTR I E S

according to their inward and outward directinvestment relationships, e.g. the share inequity capital held by non-residents andwhether the reporting company has aparticipation in non-resident companies. Thus,resident enterprises reporting outward andinward direct investment relationships can beidentified as so-called double-statusenterprises. An identification code – which isthe same for the capital links statistics and theZ 5 statistics – and a code indicating theinvestment relationships are registered in abusiness register.

513. Though the relationship between thereporting company and affiliated foreignenterprises is not precisely specified in the Z 5statistics, we are able to identify the hierarchicalstructure – derived from the capital linksstatistics – automatically via the businessregister. For example, if a mere Germansubsidiary of a foreign parent company grants aloan to a foreign affiliated enterprise thisforeign company cannot be a subsidiary of thereporting company and this loan is treated asreverse investment.

514. Consequently, with regard to short-termloans between affiliated companies we candistinguish between:

– loans granted by German companies, inwhich foreign investors have a participationof more than 10% and which have noparticipation in foreign companies (meresubsidiary status), to foreign affiliatedcompanies

– loans granted by German companies, inwhich foreign investors have a participationof up to 10% or no participation, to foreignaffiliated companies

– loans received by German companies, inwhich foreign investors have a participationof more than 10%, from foreign affiliatedcompanies

129c ECB


Annexes

– loans received by German companies, inwhich foreign investors have a participationof up to 10 % or no participation, fromforeign affiliated companies.

515. These information enable us to classify

– as direct investment in the reporting country,claims on direct investors

– as direct investment abroad, claims onaffiliated enterprises

– as direct investment in the reporting country,liabilities to direct investors

– as direct investment abroad, liabilities toaffiliated enterprises

DIRECTIONAL PRINCIPLE FOR LONG-TERM LOANS

516. Whereas short-term transactions arederived from monthly stock statistics,transactions in long-term loans are reportedtransaction by transaction. Due to the requireddeclaration of the object, e.g. the name of thenon-resident counterpart, it is possible to checkevery reported transaction manually with regardto the classification according to the directionalprinciple. Thus, also transactions betweenGerman fellow subsidiaries of foreign parentcompanies and foreign fellow subsidiaries andtransactions between foreign fellowsubsidiaries of German parent companies andGerman fellow subsidiaries, respectively, canbe separated. Furthermore, loans granted byforeign special purpose entities (SPE’s) actingas financial intermediaries have been separatedto exclude them from direct investmenttransactions.

THE DUTCH SYSTEM

Present situation:

All companies in the Balance of Paymentssystem of De Nederlandsche Bank have a so-

called micro-section number (MSNR). In theRegister used for the compilation of the Balanceof Payments, all MSNRs of the companieswhich have direct or indirect relationships, arelinked together. If the account managers andcompilers are not certain, the database of theChamber of Commerce or the Dunn andBradstreet database are consulted or thecompany involved is approached to checkwhether the right interlinkages are used.

In the Register, all mother companies,subsidiaries, fellow companies, etc. areseparately identified. However, there are nospecific MSNRs for the identification of mothercompanies, subsidiaries or fellow companies.The account managers of the companiesinvolved know how the company is built up andwhat the interlinkages are.

For the application of the directional principle, aspecific register is used. In this Register allcompanies (with their specific MSNR) involvedin intercompany reverse flows or positions areregistered. Every month, this IntercompanyRegister (ICR) is put together with all thetransactions in Direct Investment in the Balanceof Payments. When a loan from daughtercompany to its mother company is registered bythe reporting system, the ICR will recognizethis as a directional principle loan and will putthis particular loan on a different reporting codethan an “ordinary loan” while compiling theBalance of Payments. On a macro level, thesedirectional principle loans are then deductedfrom the total of intercompany loans.

For example, when a loan is granted from amother company to a subsidiary, the loan isgiven a specific reporting code. Theamortization of the loan will also follow thesame MSNRs and reporting codes and theamortized amounts are deducted from theprincipal amount to obtain the loan stocks.

When the subsidiary provides a loan to themother company the same MSNRs andreporting codes are used. With the conversionof the single reports into macro-economic


statistics, the reports are first compared with theICR (no manual work involved). Whencompanies grant loans to their parent companyand these companies are recognised by the ICR,the particular loans will be given a specificdirectional principle code. At the end of themonth, the amounts on the “directionalprinciple” are deducted from the amounts on the“ordinary” loan codes to obtain the macroamount on intercompany loans. The sameprocedure is applied for stocks.

New system: (new direct reporting system fromApril 2004 onwards):

In the new reporting system, the application ofthe directional principle is also doneautomatically by the system. On the reportingforms, companies have to fill in how much theyhave borrowed from their foreign daughters,affiliates, etc (for outward FDI) or how muchthey have lent to their foreign mother company(for inward FDI). The applicable codes areimmediately generated by the system; noconversion to other codes is necessary. The restof the procedure described above is the same asin the new situation.

131c ECB


Annexes

517. The results of the questionnaire aredifficult to interpret owing to the variety ofresponse rates for the different cells of thetable.

518. For transactions/positions on debtsecurities between affiliated companies, onlyItaly and the United Kingdom have provideddata concerning flows and only the UnitedKingdom regarding stocks. In this particularcase, the position on debt securities betweenaffiliates is very important in the outwardstocks as compared to the total of other capital(nearly 75%). However, evidence from a singlecountry is not sufficient. Finland, for instance,mentioned that some information is available inthis respect in their annual stocks survey but

5 QUE S T I ONNA I R E ON IMPORTANC E O F T H EMA I N S UB - I T EM S O F “ O THER C A P I TA L”I N F D I S TAT I S T I C S

Other Capital Item Flows StocksInward Outward Inward Outward

Debt securities, preference shares, bonds,debentures, loan stock, etc. for the followingcountries, that have reported data:

flows (IT, UK) 2,879 8,324stocks (UK) 17,300 38,200

Total other capital for the countries included above:flows 31,572 17,132stocks 146,400 51,400

Share of the item in total other capital (%) - - 11.8 74.3

Trade credits for the following countries,that have reported data:

flows (DE, ES, FI, IT, PT, SE) -621 2,742stocks (DE, ES, FI, IT, PT, SE) 28,755 31,145

Total other capital for the countries included above:flows 34,789 36,811stocks 118,383 113,981


Financial leasing / leasing credits not reported or included elsewhere

Other inter-company assets/liabilities(inter-company accounts, short-term loans,long-term loans, transferable deposits,permanent debt, other) for the following countries,that have reported data:

flows (AT, BE, DE, ES, FI, FR, IT, PT, SE, UK) 103,431 -43,306stocks (AT, DE, ES, FI, FR, IT, PT, SE, UK) 550,575 270,829

Total other capital for the countries included above:flows 105,688 -32,240stocks 596,629 340,172


Table 1 Questionnaire on importance of the main sub-items of “Other Capital” in FDIstat ist ics(FDI data for 2001 in EUR millions)

that the figures are so insignificant that they arenot even included at all in the published stocks.

519. Regarding trade credits, six countries outof the ten that have returned the questionnaireshave provided figures both for flows and forstocks, even though not all of them make aseparate entry for this item in their publications.This item represents around 25% of the total ofother capital for inward and outward stocks aswell.

520. As regards, financial leasing/leasingcredits, not a single country was in a position toprovide any figures, either because this item isnot reported at all or it is includedindistinguishably in inter-company loans.


521. The answers provided to the secondquestionnaire, showing a very detaileddisaggregation of other inter-company assetsand liabilities, were so different andinconsistent that it has not been possible tokeep the disaggregated layout in the synthetictable. Instead, they have been grouped undera single heading “other inter-company assets/liabilities“. For the nine countries that havereported stocks under one or several headings,the total position represents 92% of the total ofother capital for inward stocks and 80% foroutward stocks.

133c ECB


Annexes

UBO IN AUSTRIAN FDI-STATISTICS

522. In the Austrian FDI-statistics inward stockscan be allocated according to both, the First-Shotand the Ultimate-Beneficiary Ownershipprinciple (UBO). The OeNB started compiling ofFDI statistics around 1970 following the first-shot principle. Austrian residents were asked forthe “immediate owner or owners” only. But assoon as in 1980 it was felt that the allocationaccording to FS might be misleading. Thereforethe OeNB introduced the so called“Stammhausbereinigung”, which in fact is UBO.The need for UBO was explained with thegrowing popularity of holding companies, whichresulted in a clear bias for certain countries (withfavorable tax and regulatory framework).

523. An additional question (3A) had beenintroduced in the annual survey, which shouldbe filled in, “if the immediate owner is not theultimate beneficiary” (see simplified question;original available only in German).

524. The example is a graphical representationof a four level chain of ownership, with aguideline for the calculation of one or moreUBOs. It is stated that one FS-owner maybelong to several UBOs. On the other handseveral FS-owners may have only one identicalUBO. Respondents have to supply %-shares,names, and domicile of UBOs. The answer tothis question is subject to the condition that theforeign influence measured at the FS-level is ofthe same size as the foreign influence at UBO-level. The number of owners on FS and UBO-level may of course be different. It turned outthat in some cases the UBO behind a FS-investor was in fact an Austrian. Respondent

6 NAT I ONA L D E S C R I P T I ON S O F U BO - B A S EDFD I S TAT I S T I C S

Complete name of company according to Identification number in % with one decimal figureCommercial Register, domicile, country (do not fill out)

a) .........b) .........Total /1070

3A. Parent company form (Information about ultimate beneficial owner)Share of the ult imate benef ic ia l owner

Example: explains how to calculate the share of the ultimate beneficial owner

are encouraged to supply ownership trees toexplain the relation between companies. IfUBOs are identical with FS-investors or if theyare unknown, the respondent leaves thequestion empty.

525. In the most recent survey for 515 out of 3312investment relations an UBO has been mentionedby the respondent. This, of course, does not implythat the country of origin of UBO and FS aredifferent in every case. In 2000 34 countriesare mentioned as ultimate countries of origin,the most important ones being DE (98 cases), theUS (88), GB (51), JP (43), CH (41), FR (41) andIT (34). AT is mentioned in 24 cases. Theseinvest-ment relations are eliminated from the FDIstatistics, when UBO is used.

526. A comparison of the inward FDI-statistic according to FS and UBO-principlerespectively produce the expected pattern.Switching from FS to UBO typically leads to areduction of the share of the Netherlands,Switzerland and Luxemburg, while the US andJapan, but also a number of EU-member statesare the “winners” of the reallocation. DE loosesinvestment relations, but wins invested capitaland employees. Table 1 gives the comparisonof the two concepts measured in numberof investment relations, total capital andemployment for the year 2000.

527. For outward FDI-stocks we do notapply the UBO-principle with one exception.According to the requirements of outward FATS-statistics one should only take into considerationoutward investments (of more than 50%), if theresident investor is a “domestic enterprise”,meaning that it is not “foreign controlled”. At


least in cases where the resident investor himselfis at the same time an investee, it is possible tocheck whether the foreign influence is below orabove 50%. For the time being it is not possible tofollow the chain of ownership within AT.Therefore this procedure overestimates the role of“domestic” enterprises. At least some of the“domestic” enterprises may be indirectly foreigncontrolled.

528. The foreign influence on Austrian investorsis in fact important. Table 25 in our annualpublication on outward stocks shows, that 199Austrian FDI-investors (out of a total of 917) areat the same time affiliates of foreign investors.161 are actually foreign controlled. The totaloutward stock of equity capital in 2000 was23.9 bio €. Almost one third of it (7.5 bio €)belonged to foreign controlled resident investors.

COMPILATION OF FDI STOCK FIGURES BASED ONTHE UBO PRINCIPLE IN GERMANY

529. We have a yearly FDI stock survey in DE.For inward FDI all enterprises in DE have toreport, if a foreign investor has a participating

interest in this enterprise of 10 % or moreand the balance sheet total of the enterprise inDE is above 3 Mio € (for the figures from 2002on; up to the year-end 2001 the threshold forthe balance sheet total for minority-foreignownership – between 10% and less than50% – was 5 Mio € and for majority-foreignownership – 50 % and more – was 0,5 Mio €).In the reporting form the following questionis included: “If the non-resident holding theparticipating interest is a dependent enterpriseitself: country in which the registered officeof the controlling enterprise is located_________”. With this information we areable to compile the FDI inward stock figureseither according to the first shot or according tothe UBO. If there is no answer to this questionin the reporting form, we assume, that the firstforeign owner is also the UBO.

530. We are only able to provide inward FDIfigures in a country breakdown according tothe first owner or the UBO. We do neither haveany information about the chain of ownershipfrom the UBO to the first foreign owner ofthe enterprise in DE nor do we have anyinformation about the name of the UBO.

AUSTRIA Inward FDI-Stock 2000First shot Ultimate owners Difference UBO-FS

Intra-EU15 25,816 25,389 -427

Netherlands 6,368 2,053 -4,315Germany 15,051 15,295 244Luxembourg 1,145 859 -285Sweden 252 636 384France 983 1,399 416United Kingdom 1,138 1,881 743Italy 378 1,692 1,314Austria 1) 0 79 79oth. Intra-EU 502 1,494 992

Extra-EU15 6,967 7,315 348

Switzerland 3,712 2,306 -1,406Japan 118 841 723USA 1,672 2,024 352oth. Extra 1,465 2,144 679

Total 32,783 32,783 0

Table 1 Inward FDI-stocks according to FS and UBO-concepts

(EUR millions)

Source: OeNB.1) Eliminated in publications using UBO.

135c ECB


Annexes

Country/country group First shot UBO Difference

Total 447,347 443,532 3,815EU-countries 341,337 303,832 37,505of which:Austria 5,370 4,802 568Belgium 12,714 6,092 6,622Denmark 4,446 2,056 2,390Finland 3,946 5,887 -1,941France 31,334 33,059 -1,725Ireland 1,180 669 511Italy 5,615 7,105 -1,490Luxembourg 145,839 3,571 142,268Netherlands 88,398 59,218 29,180Spain 1,929 1,340 589Sweden 7,735 6,610 1,125United Kingdom 32,703 172,931 -140,228

Other industrial countries 99,899 131,944 -32,045of which:Australia 142 1,120 -978Canada 1,397 3,142 -1,745Japan 8,414 11,500 -3,086Switzerland 21,689 27,166 -5,477United States 66,602 87,273 -20,671

Countries in transition 1,006 979 27

Developing countries 5,105 6,777 -1,672of which:developing countries in Africa 126 1,925 -1,799developing countries in America 1,729 1,485 244developing countries in Asia and Oceania 3,249 3,367 -118

Table 2 Primary FDI stock in Germany at year-end 2000

(EUR millions)

531. Attached please find a table 2 with acomparison of the primary inward FDI stockfigures in DE at the year-end 2000 in a countrybreakdown according to the first foreign ownerand the UBO. In that table you can see, that thetotal for first shot is higher than for UBO. Thereason for this is, that the UBO of someenterprises in DE with direct capital links toforeign investors is located in DE (see Table 2).

THE UBO PRINCIPLE IN DANISH FDI-STATISTICS

532. Our annual FDI-survey covers the UBO atthe inward side. All inward FDI-data can beallocated to first shot or to the UBO if any. Allenterprises in DK with a foreign investorparticipating in an enterprise with 10 percent ormore have to report to the inward FDI-

statistics. All enterprises have to report firstshot and the UBO if any.

533. We only compile the country code relatedto the first shot and the UBO and not the nameof the investor (first shot or the UBO). We donot cover the chain of ownership between firstshot and the UBO. We allocate equity capital tothe UBO if any. Other capital is alwaysallocated to the first shot.

534. We are able to provide figures allocated tofirst shot or to the UBO at a geographicalbreakdown. For dissemination-purposes we usethe UBO for the geographical breakdown in thenational statistics as well as the internationalreporting, including the inward FDI-data forEurostat.


Country First shot UBO Difference

Total 51,823 51,823 0EU-countries 41,682 22,552 19,129of which:Austria 786 446 340Belgium 893 1,001 -108Finland 497 465 33Germany 647 620 26France 1,676 1,403 273Ireland 87 40 47Italy 641 724 -84Luxembourg 16,951 606 16,345Netherlands 5,929 2,704 3,225Spain 779 273 506Sweden 10,106 9,192 914United Kingdom 2,689 5,078 -2,389

Other essential countriesCanada 48 58 -9Japan 96 141 -45Norway 2,432 2,430 1Switzerland/Liechtenstein 749 776 -27United States 4,175 19,829 -15,654

Table 3 Comparison of the geographical breakdown using f irst shot and the UBOInward FDI Stock in Denmark, 2000 – Equity(EUR millions)

535. From Table 3 you will notice that thereis considerable deviation between thegeographical breakdown using first shot and theUBO as a basis for the geographicalbreakdown. The deviation is mainly due toelimination of the passing through effect fromholding companies - SPEs.

137c ECB


Annexes

INTRODUCTION

536. Reinvested earnings on direct foreigninvestment (FDI) was introduced in the nationalaccounts (NA) with the SNA 93/ESA 95, as oneof the items in the transition from grossdomestic product (GDP) to gross nationalincome (GNI). It is also an item in the transitionfrom GDP according to ESA 95 to grossnational product (GNP) according to ESA 79,which is currently used as a base for calculatingpayments from Member States to the EU. InEurostat, Unit B1 is responsible for assessingthe quality of the GDP, GNI and GNP estimatesin the Member States. The conceptual andpractical aspects of the compilation ofreinvested earnings data is part of thisassessment.

537. Reinvested earnings on direct foreigninvestment is also part of statistics on foreigndirect investment (FDI) flows which is a part ofthe Balance of Payments (BoP) statistics. Inmost Member States, FDI statistics arecompiled by Central Banks. In Eurostat, UnitB5 is responsible for FDI statistics. FDI data,including reinvested earnings are available inNewCronos and in the publication “EuropeanUnion foreign direct investment yearbook 2000,Data 1992-1999”. Eurostat recommends thatcountries follow guidelines issued by IMF andOECD: IMF Balance of Payments Manual,Fifth edition and OECD Benchmark Definitionof Foreign Direct Investment, Third edition.These guidelines are both consistent with SNA93/ESA 95.

538. This annex presents the relevantdefinitions in ESA 95 and compares availabledata from NA and BoP.

ESA 95 DEFINITIONS

539. In the ESA 95, reinvested earnings ondirect foreign investment are discussed inChapter 4 Distributive transactions:

7 R E I N V E S T ED E A RN I NG S I N N AT I ONA LA C COUNT S AND IMPA C T ON GDP

Reinvested earnings on direct foreigninvestment (D.43)

4.64. Definition:

Reinvested earnings on direct foreigninvestment (D.43) are equal to:

the operating surplus of the direct foreigninvestment enterprise

plus any property incomes or currenttransfers receivable

minus any property incomes or currenttransfers payable, including actual remittancesto foreign direct investors and any current taxespayable on the income, wealth, etc., of the directforeign investment enterprise.

4.65. A direct foreign investment enterprise isan incorporated or unincorporated enterprise inwhich an investor resident in another economyowns 10 per cent or more of the ordinary sharesor voting power (for an incorporated enterprise)or the equivalent (for an unincorporatedenterprise). Direct foreign investmententerprises comprise those entities that areidentified as subsidiaries (investor owns morethan 50 per cent), associates (investor owns 50per cent or less) and branches (wholly or jointlyowned unincorporated enterprises), eitherdirectly or indirectly owned by the investor.Consequently, ‘direct foreign investmententerprises’ is a broader concept than ‘foreigncontrolled corporations’.

4.66. Actual distributions may be made out ofthe entrepreneurial income of direct foreigninvestment enterprises in the form of dividendsor withdrawals of income from quasi-corporations.

In addition, retained earnings are treated as ifthey were distributed and remitted to foreigndirect investors in proportion to theirownership of the equity of the enterprise andthen reinvested by them.


Reinvested earnings on direct foreigninvestment can be either positive or negative.

4.67. Time of recording: Reinvested earningson direct foreign investment are recorded whenthey are earned.

In the system of accounts, reinvested earningson direct foreign investment appear:

a) among uses and resources in the allocationof primary income account of the sectors;

b) among uses and resources in the externalaccount of primary incomes and currenttransfers.

REINVESTED EARNINGS IN PERCENT OF GDP

540. 14 EU countries reported data forreinvested earnings on FDI to Eurostat in theGNP Questionnaire 2002. Greece did not reportany data on this item, and Italy reported dataonly from 1998 onwards. As can be seen fromthe Table 1 below, net flows of reinvestedearnings are below 0.5% of GDP in mostMember States. In Sweden and the UK there arenet inflows of around 1.5% of GDP, while

Table 1 Reinvested earnings on direct foreign investment

(net, in percent of GDP)

1995 1996 1997 1998 1999 2000 2001 Ave 95-01

Austria -0.5 -0.1 -0.3 -0.3 -0.4 -0.4 -0.4 -0.3Belgium 0.1 -0.1 -0.4 -0.4 -0.5 -0.4 -0.2 -0.3Denmark 0.0 0.0 0.2 0.6 0.2 0.0 0.2 0.2Finland -0.4 -0.2 0.3 -0.3 0.4 -0.1 -2.4 -0.4France -0.2 0.1 0.1 0.2 0.7 0.4 0.3 0.2Germany 0.2 0.4 0.2 0.3 0.5 0.3 0.2 0.3Greece : : : : : : : :Ireland -2.6 -3.3 -2.8 -3.7 -8.1 -9.6 -7.8 -5.4Italy : : : 0.1 0.0 -0.2 -0.4 -0.1Luxembourg -8.7 -6.4 -11.2 -7.7 -12.3 -11.8 -5.0 -9.0Netherlands 1.2 0.1 0.6 -1.9 0.2 0.2 -1.9 -0.2Portugal -0.4 -0.6 -0.5 -0.4 -0.6 -0.1 -0.1 -0.4Spain 0.0 0.0 -0.1 -0.2 0.1 0.2 0.2 0.0Sweden 1.1 1.4 1.3 1.3 1.8 2.5 1.4 1.5United Kingdom 1.3 1.2 1.2 1.5 1.9 1.5 1.4 1.4

Source: GNP Questionnaire 2002.

Ireland and Luxembourg have large netoutflows.

COMPARISON OF REINVESTED EARNINGS DATA INFDI STATISTICS AND GNP QUESTIONNAIRE 2002

541. The Table 2 below shows that reinvestedearnings data in FDI statistics and in the GNPQuestionnaire 2002 are quite similar in Austria,Finland (the difference in 1998 is due todifferent treatment of one very large and ratherspecial dividend transaction), Germany,Ireland, Italy, Portugal and the UnitedKingdom.

542. Because of missing data, no comparisonsare possible for Belgium and Luxembourg(reported together in FDI statistics), Greece andSpain.

543. For the other EU countries, in particularFrance and the Netherlands, the differences canbe quite large. Possible reasons includerevisions made at different times (in particularfor the latest years, 2000 and 2001) andadjustments to the BoP data made by thenational accounts.

139c ECB


Annexes

1995 1996 1997 1998 1999 2000 2001

Austria FDI Stat -959 -229 -542 -529 -765 -815 -958GNPQ 2002 -918 -223 -544 -533 -765 -815 -857Share 1.04 1.02 1.00 0.99 1.00 1.00 1.12

Belgium FDI Stat : : : : :GNPQ 2002 218 -267 -778 -926 -1,096 -897 -630Share

Denmark FDI Stat : : : 600 500 744GNPQ 2002 -67 33 225 931 370 -5 302Share 0.64 1.35

Finland FDI Stat -370 -180 330 -3,766 526 -193 -3,217GNPQ 2002 -355 -176 327 -384 526 -194 -3,226Share 1.04 1.02 1.01 9.81 1) 1.00 0.99 1.00

France FDI Stat : : 1,739 2,939 3,540 3,412 3,142GNPQ 2002 -1,792 1,251 1,753 2,954 9,794 5,170 4,953Share 0.99 1.00 0.36 0.66 0.63

Germany FDI Stat 3,529 8,364 3,563 5,637 9,800 5,900GNPQ 2002 3,370 8,160 3,560 5,680 9,800 5,900 3,600Share 1.05 1.03 1.00 0.99 1.00 1.00

Greece FDI Stat : : : : :GNPQ 2002 : : : : : :Share

Ireland FDI Stat : : : -2,876 -7,284 -9,898 -8,884GNPQ 2002 -1,383 -1,895 -1,858 -2,871 -7,283 -9,898 -8,883Share 1.00 1.00 1.00 1.00

Italy FDI Stat : : : : 258 -2,911 -5,474GNPQ 2002 : : : 801 256 -2,904 -5,396Share 1.01 1.00 1.01

Luxembourg FDI Stat : : : : :GNPQ 2002 -1,150 -889 -1,736 -1,309 -2,294 -2,450 -1,065Share

Netherlands FDI Stat 3,855 524 3,710 -4,218 -2,695 -9,804 -8,486GNPQ 2002 3,671 304 1,975 -6,585 571 847 -7,986Share 1.05 1.72 1.88 0.64 -4.72 -11.57 1.06

Portugal FDI Stat : -509 -435 -442 -601 -151 -108GNPQ 2002 -285 -497 -431 -445 -601 -151 -108Share 1.02 1.01 0.99 1.00 1.00 1.00

Spain FDI Stat : : : : :GNPQ 2002 26 18 -371 -1,160 495 948 1,307Share

Sweden FDI Stat : : : : 4,187 5,711 1,788GNPQ 2002 2,123 2,952 2,751 2,797 4,187 6,368 3,318Share 1.00 0.90 0.54

United Kingdom FDI Stat 11,352 11,943 14,596 19,063 25,490 23,471 20,882GNPQ 2002 11,009 11,548 14,049 18,552 25,481 23,446 22,720Share 1.03 1.03 1.04 1.03 1.00 1.00 0.92

Table 2 Reinvested earnings on foreign direct investment

(net, ECU/EUR millions)

Sources: FDI Stat: FDI statistics in NewCronos, November 2002. GNPQ 2002: GNP Questionnaire 2002, reported in September/October2002.Note: Share: Value in FDI statistics relative to the value in GNP Questionnaire 2002.1) For Finland in 1998, see the explanation in the text above.


Ms René Dell’mour Oesterreichische NationalbankMr Roger De Boeck Banque Nationale de BelgiqueMs Katrien Adams Banque Nationale de BelgiqueMs Birthe Jensen Danmarks NationalbankMs Beatrix Stejskal-Passler Deutsche BundesbankMr Martin Beck Deutsche BundesbankMs Maria Teresa García Cid Banco de EspañaMs Airi Heikkilä Suomen PankkiMr Frédéric Lambert Banque de FranceMs Corinne Devillers Banque de FranceMr Andreas Karapappas Bank of GreeceMr. George Iliopoulos Bank of GreeceMr Kevin O’Donnell Central Statistics Office, IrelandMr Maurizio Iannaccone Ufficio Italiano dei CambiMs Olga Aarsman De Nederlandsche BankMr Hossein Kabki De Nederlandsche BankMs Margarida Brites Ramos Banco de PortugalMs Paula Casimiro Banco de PortugalMr Lars Fors (Chairman) Sveriges RiksbankMr Philip Jenkins Office for National Statistics,

United KingdomMr Carlos Sánchez Muñoz (secretary) European Central Bank

Balance of Payments Statistics andExternal Reserves Division

Mr Jean Galand (secretary) European Central BankBalance of Payments Statistics andExternal Reserves Division

Mr Paolo Passerini Eurostat/B5Mr Irene Lovino Eurostat/B5Mr Jean-François Yattien-Amiguet Eurostat/B5

8 L I S T O F MEMBER S O F T H E T A S K F OR C E ONFOR E I GN D I R E C T I N V E S TMENT

141

Introduction

cECBFore ign d i rec t inves tment – Task force repor t

March 2004

INTRODUCTION 143

1 REASONING FOR THE IDENTIFICATION OF SPES AND PROBLEMS IN DEALINGWITH SPES 1441.1 Why should SPEs be identified separately 1441.2 Problems in dealing with SPEs 1441.3 Previous investigations by the OECD 145

2 DEFINITIONS OF SPES IN INTERNATIONAL GUIDELINES 1472.1 OECD Benchmark definition of Foreign Direct Investment 1472.2 IMF: 5th Manual, BOP Textbook and BOP compilation guide 1472.3 European System of Accounts ESA95 1482.4 International Accounting Standards (IAS) 1492.5 Conclusions 150

3 THE IMPORTANCE OF SPES AND OFFSHORE ENTITIES IN EUROPEAN STATISTICS 1533.1 Overview of the impact of offshore centres on European aggregates 1543.2 Impact of offshore (financial) centres at national level 1573.3 Result of the empirical exercise on the importance of SPEs

in EU Member States 1583.4 Special treatments of SPEs in EU Member States 1593.5 Conclusions 159

4 CONCLUSIONS 160

Annex A Example by NL 161

Annex B Overview of the different definitions of SPEs and offshore centresin international guidelines 163

Annex C Importance of offshore countries in US data 171

Annex D Data on the importance of SPEs in EU Member States 175

Annex E SPEs in the Eurostat classification by sector (NACE rev1) 181

Annex F Special treatment of SPEs in EU Member States 184

S U P P L EMENTARY DOCUMENT:S P E C I A L P URPO S E EN T I T I E SCONTENT S

143

Introduction

cECBFore ign d i rec t inves tment – Task force repor t

March 2004

I N T RODUCT I ON1. The proceeding globalisation and theformation of multinational enterprises and theincreasing integration of the world economy arereflected by a remarkable increase of cross-border financial transactions. The establishmentof complex intra group structures is more andmore determined by financial and fiscal aspects.Against this background Special PurposeEntities (SPEs) are playing an important role inconducting group financing and in providingfinancial services. Moreover, profit shifting inorder to minimise tax burdens is clearly animportant determinant for the establishment ofSPEs in countries with a convenient taxenvironment for foreign companies. Regardingthese activities, not only the growing amountsof intra group cross-border capital flows, butalso the increased complexity of suchtransactions may raise problems in thestatistical treatment of this phenomenon in somecountries.

2. For the time being the recommendations inBPM5 related to SPEs provide no specialtreatment of SPEs except para. 372, which saysthat only transactions in equity capital andpermanent debt between affiliated banks andaffiliated financial intermediaries – includingSPEs with the sole purpose of serving asfinancial intermediaries – should be included inFDI, whereas other capital transactions shouldbe excluded. With the exception of SPEsmentioned in BPM5, para. 372, most memberstates have no special approach for theidentification and treatment of SPEs within FDIstatistics. On the one hand the absence ofnational legal provisions or specificrecommendations in the national statisticalframework prevent a coherent approach indealing with SPEs. On the other hand thenecessity for a separate identification of SPEsseems often not to be obvious. Indeed one couldask about the reasoning of defining andidentifying SPEs. The discussion within theOECD Workshop on Financial InvestmentStatistics has shown that several compilers donot really recognise the identification of SPEswithin the population of FDI enterprises as atop priority issue. Therefore, some arguments

will be presented in the following in order toprovide some rationale for the specificimportance of SPEs.

144ECBTask force on foreign direct investmentFebruary 2004

1.1 WHY SHOULD SPES BE IDENTIFIEDSEPARATELY?

3. First of all, there is a good reason to identifySPEs for analytical purposes. Users of FDIstatistics may be interested in the nature of FDIactivities, e.g. whether they are stemming fromreal operational business or merely from fiscalcalculus. In this context a key issue is thereforeto question the economic incentives related toSPE activities.

4. The term SPE comprises a large variety ofenterprises with more or less specific activities.In general, SPEs are primarily engaged ininternational activities but in few or no localoperations. Their functions frequently dependon the legal environment and taxation regime intheir host country. SPEs acting as financialvehicles can be economically more accuratelycharacterised as intermediaries rather than asdirect investors or direct investment targets.The impact of SPEs on the economy they areoperating in are often negligible. In particular,holding companies, shell companies, etc.typically have no real economic activity, noturnover and no significant employment.

5. The so-called green field investments areregarded as an important indicator for theeconomic attractiveness and competitiveness ofregions and countries competing for directinvestment. Policy makers and economists doregularly refer to foreign direct investmentfigures as an indication for economic conditionsand the quality of economic policy. Thesediscussions are mainly focused on inward flowsand stocks. The increasing amounts of cross-border financial transactions arising from theformation of SPEs – together with theaccelerated M&A activities – have extendedgross financial flows significantly. However,these transactions have hardly any real impacton the domestic economy and do commonly notaffect the creation of new production capacitiesand employment opportunities. Therefore, theinterpretation of gross flows and theconclusions drawn from this might bemisleading, if comprehensive gross direct

1 R E A SON I NG F OR TH E I D EN T I F I C AT I ON O F S P E S

AND P ROB L EM S I N D E A L I NG W I TH S P E Sinvestment transactions arise from fundschannelled through third countries rather thatdirectly from the investor to the final target.

6. There can also be deducted a connection tothe calculation of direct investment income.International standards congruently recommendthe application of the current operatingperformance concept including only profits andlosses arising from operational business in FDIincome. In this context one might argue that allforeign direct investment transactions andpositions should be identified according to their“economic background”. In the case of holdingor shell companies, the assignment of FDIflows and stocks according to the ultimateinvestor and the ultimate target, respectively,would facilitate a geographical allocation inagreement with real economic incentives.

1.2 PROBLEMS IN DEALING WITH SPES

7. In the context of providing meaningful FDIstatistics, severe problems related to SPEs arisefrom:

– the identification and classification of SPEs,in particular holding companies, shellcompanies, financial intermediaries andfinancial auxiliaries. Detailed informationabout the business segment are essential foran assessment of the economic rationalebehind the investment activities. In thisrespect an industry breakdown might oftennot be sufficient for an unambiguousdistinction between SPEs and other FDIenterprises. Moreover, information about thebusiness segment or industry are oftenlimited to domestic entities. In the case ofoffshore centres one might assume that SPEsplay a major role in these countries, simplyclassifying enterprises located in offshorecentres as SPEs. However, there are evidentdifferences between the group of offshorecountries. Regarding offshore centrespossessing a broad economic basis, e.g.Hong Kong or Singapore, inward andoutward direct investment might also be

145c ECB

Task force on foreign direct investmentFebruary 2004

1

Reasoningfor the

identificationof SPES andproblems in

dealingwith SPES

motivated by real economic activities ratherthan merely by fiscal incentives.

– the geographical allocation: When funds arechannelled through a SPE, the real economicactivity might often not be located in thecountry where the SPE is resident, but inthird countries where the final target of FDIis located. When only direct FDIrelationships are taken into account, FDItransactions and positions might be biasedtowards host countries of SPEs. In thiscontext the identification and coverage ofindirect relationships or the possibility of“looking through” holding and shellcompanies is a critical issue. Severalcountries have already developed approachesto tackle this problem. For example, in theU.S. FDI statistics, foreign shell companiesincorporated abroad having all their physicalassets or operations in a second foreigncountry are treated as incorporated foreignaffiliates in the second country where theirphysical assets or operations are located.However, for the implementation of such atreatment, comprehensive informationprovided by reporting companies arerequired; and for the time being such data areapparently not available in most countries.

– asymmetries between national statistics:according to international standards, theactivities of SPEs integrated in amultinational group structure should betreated as direct investment. However,various national practices in treating SPEs, inparticular financial intermediaries andfinancial auxiliaries, are causingasymmetries. Moreover, the sectorclassification might be differing amongmember states.

1.3 PREVIOUS INVESTIGATIONS BY THE OECD

8. The analysis of the subgroup on SPEs shouldalso shed some light on the different approachesin dealing with SPEs and off-shore enterprisesamong the member states. A categorisation of

country specific approaches in dealing withSPEs was already attempted within the scope ofthe OECD Working Party on FinancialStatistics in the recent past.1 Since we do notintend to duplicate this exercise we focused onan empirical investigation, which was designedto analyse the present stage of possibilities inidentifying SPEs in national statistics and toevaluate the order of magnitude (of parts) ofSPE activities.

9. There is still no common approach in order totackle these open issues among Europeancountries. Though the results of the OECDstudy are still preliminary and will be reviewedin a second stage, the OECD paper represents auseful reference providing an overview of thevarious treatments of SPE activities in differentcountries. The results indicated that only a fewcountries possess a legal provision regardingthe definition of SPEs and that there rarelyexists any specific recommendation for thetreatment of SPEs in FDI statistics. As thediscussion within the OECD Workshop onInternational Investment Statistics showed, itproved hard to categorise the treatment of SPEsamong countries when most countries haveeven not established recommendations for aspecific treatment of FDI enterprises.

10. What can be learnt from the OECD studies?The large variety of answers and commentsprovided by compilers indicates the mostimportant obstacles in identifying and analysingSPEs and their activities:

– there is hardly any definition,recommendation or treatment of SPEs appliedby at least two countries in the same mannerexcept for those countries not distinguishingbetween ordinary FDI enterprises and SPEsat all.

1 See “Report on Special Purpose Entities and Off-ShoreEnterprises”, note by the OECD secretariat prepared for theWorkshop on International Investment Statistics, 5-6 March 2003in Paris, DAFFE/MC/STAT(2003)4.


– there is no country of the euro area collectingdata from resident enterprises on their SPEsabroad.

– information about tasks and activities ofSPEs are not collected systematically by mostcountries. In this context, also the use ofshort-term instruments is still not analysedcomprehensively. However, in particularshort-term intra-group financing causessubstantial cross-border capital flows.

11. Against the background of the follow-upwork of the OECD secretariat, the subgroup onSPEs does not intend to replicate these resultsby an own survey, but to focus on an empiricalexercise exploring the extent of SPE activitiesthat can already be identified for the time being.It might be helpful to identify a least commondenominator with respect to the category ofSPEs which can already be separated by someMember States.

147c ECB


2

DefinitionsOF SPES in

internationalguidelines

12. This section presents some selecteddefinitions of SPEs available in the maininternational guidelines. Mainly the IMF 5th

Manual, the IMF BOP Textbook, the IMF BOPcompilation guide, the OECD Benchmark, theESA95 , and the International FinancialReporting Standards (IFRS-2003 Edition) havebeen investigated.

2.1 OECD BENCHMARK DEFINITION OF FOREIGNDIRECT INVESTMENT

13. The Benchmark points out the fact that SPEsare the results of groups’ strategy aiming atfacilitating the financing of their investmentover the world and contributing to profitmaximisation for the whole group (§69). As aresult, most SPEs have a management or anadministrative function, and are acting mainlyas:

– financing subsidiaries, that is in raising fundsor allocating funds from one unit of the groupto another.

– conduits, that is (only) transferring fundsamong different entities of the group.

– holding companies, managing either financialor non-financial assets. Annex 3 states thatholding companies should be considered asfinancial corporations even though theinvestment that they hold may be in otherindustries.

– base or regional headquarters companies.

– tax haven corporations, “sheltering” incomegenerated by the whole activity of the groupand “minimizing” the overall tax cost of theGroup’s business.

14. According to the OECD, SPEs are usuallylocated in tax havens (Annex 3), and may bestructured in a number of ways, includingseparate legal entities and branches of legallyconstituted entities.2 Because of this referenceto tax havens, it should be interesting to note the

2 D E F I N I T I ON S O F S P E S I N I N T E RNAT I ONA LGU I D E L I N E S

statistical FDI treatment devoted to offshoreentities (§68): where a company Z incorporatedin country A has its management office inanother country B, country A in its outwarddirect investment statistics regard the foreignmanagement office as direct investment bycountry A in a branch in country B etc.

15. Furthermore, SPEs could be found also inthe manufacturing sector or could deal with awide variety of activities such asmerchandising, insurance or shipping activities(Annex 3, p. 46).

16. As regard to direct investment transactions,the Benchmark also stipulates that, since SPEsare an integral part of the organizationalstructure of a multinational enterprise, theirtransactions that arise from direct investmentrelationships should be reflected in the statisticsand, if possible, shown as a sub-component(§69) and on a gross basis (§38). When actingonly as financial intermediaries, the FDItreatment has to be similar to the one defined forthe banks (§60 and 61).

2.2 IMF: 5TH MANUAL, BOP TEXTBOOK AND BOPCOMPILATION GUIDE

17. The IMF deals with SPEs in chapter IV(Resident Units of an Economy) and chapterXVIII (Direct investment) of the 5th Manual, inChapter IX of the Textbook and, to some extent,under chapter XVI of the Compilation guide.

18. Regarding the residency criteria, the IMFstates that offshore enterprises engaged inmanufacturing processes or non-manufacturingoperations, including so-called special purposeentities (SPEs), are residents of the economy inwhich the offshore enterprises is located (§79and 381). The Textbook adds that SPEs are(§542):

2 In recent years not only the population of SPEs but also themagnitude of their activities in countries other than tax havens,like DK, IE and NL, has been increasing.



– engaged primarily in internationaltransactions but in few or no local operations.Engaged in few or no local operations meansthat the company does not carry outproduction, have no employee and do not payincome tax. Books or accounts may bemaintained elsewhere and, thus, beunavailable for the host country compiler(§706 of the Compilation guide).

19. The Textbook also provides two exampleson the recording of typical SPEs’ transactions,one SPE acting as a “treasury (or assets)management company” (§543), another oneacting as a pure financial intermediary (§544).

20. IMF Manuals give some recommendationson the statute of SPEs and the treatment ofSPEs’ transactions in the field of directinvestment. The most important ones are shownbelow:

– it is recommended to consider SPEs asresident direct investment enterprises (§365),if they meet the general criteria that defineFDI relationships. Transactions throughSPEs have to be also included in directinvestments (§372).

– for SPEs with the sole purpose of serving asin a financial intermediary capacity, theManual recommends to restrict their FDItransactions only to those associated withpermanent debt and equity, or fixed assets inthe case of branches (§365 and 372).

– whatever the treatment used by the reportingcountry, the Manual recommends toseparately identify the value of SPEtransactions as a group, to permit consistentinternational comparison (§365).

– always to improve international comparison,compilers should record and send to the IMFthe complete b.o.p. entries of SPEs (i.e. on a

gross basis), even if it is recognised thatsome compilers may prefer to excludetransactions considered to be of no relevancefor the domestic economy (Compilationguide, §708).

21. SPEs can have the structure of a holdingcompany, a base company, or a regionalheadquarters, could serve for administration,management of foreign exchange risk purposes,or facilitate the financing of investments (§365-brackets).

2.3 EUROPEAN SYSTEM OF ACCOUNTS ESA95

22. Chapter 2 of ESA95 (Units and groupings ofUnits) might help the TF-FDI in theclassification of SPEs by National Accountsectors. The TF-FDI proposal for the new IMFmatrix chart is to exclude transactions otherthan equity and permanent debt between allentities belonging to S122, S123 and S124.Unlike other international guides, ESA95provides a clear definition of holdingcompanies, which could be helpful given thatboth the OECD benchmark and IMF Manualsconsider that some SPEs could have the statuteof holding companies.

23. Paragraph 2.14 of ESA95 states thatholding corporations are institutional unitswhose main function is to control and direct agroup of subsidiaries.

24. For SPEs identified as holdingcorporations, ESA95 suggests to classifyholding corporations in the sector betterreflecting the activity of the group ofsubsidiaries as a whole:

– paragraph 2.23-e) states that holdingcorporations controlling a group ofcorporations which are market producers (i.e.a group of subsidiaries) are ‘non-financialcorporations’ (S11) if the preponderant typeof activity of the group of corporations/subsidiaries (measured on the basis of value

149c ECB


2



added) is the production of goods and non-financial services.

– paragraph 2.40-e) states that holdingcorporations controlling a group ofsubsidiaries are financial corporation (S12) ifthe group of subsidiaries within the economicterritory as a whole is principally engaged infinancial intermediation and/or in auxiliaryfinancial activities.

– paragraph 2.43 (and footnote 9) states thatholding corporations which only control anddirect a group of subsidiaries principallyengaged in financial intermediation and/or inauxiliary financial activities are classified insub-sector S123 “other financialintermediaries except insurance corporationand pension funds”. However, holdingcorporations which are financial corporationsthemselves (holdings of holdings?) are to beallocated to the sub-sectors according to themain type of financial activities.

– with a more precise wording, paragraph2.63-b) says that sub-sector S125 (insurancecorporations and pension funds) does notinclude holding corporations which onlycontrol and direct a group consistingpredominantly of insurance corporations andpension funds, but which are not insurancecorporation and pension funds themselves.They are classified in sub-sector S123.

25. For “non-holding” SPEs (mainly SPEs withno subsidiaries) further guidance is provided byESA95 depending if the SPE is acting as afinancial intermediary or as a treasury servicesprovider.

– paragraphs 2.32-2.34 states that financialintermediation is the activity in which aninstitutional unit acquires financial assets andat the same time incurs liabilities on its ownaccount by engaging financial transactions onthe market. Funds are channelled betweenthird parties with a surplus on one side andthose with a lack of fund on the other.Financial intermediaries may invest their

funds in non-financial assets including realestate. However, in order to be considered asa financial intermediary, a corporationshould, in addition incur liabilities on themarket and transfer funds. For this reason,real estate corporations are not considered asfinancial intermediaries.

– paragraphs 2.37 states that financialintermediation does not include institutionalunits providing treasury services to acompany group, unless subject to financialsupervision. When they are not subject tofinancial supervision, units acting as treasuryservices are allocated to the sector betterreflecting the predominant function of thecompany group within the economic territory.

– paragraphs 2.55-f) states that financialvehicle corporations, created to be holders ofsecuritized assets have to be classified underS123, unless they are MFIs (S122).

2.4 INTERNATIONAL ACCOUNTING STANDARDS(IAS)

26. Most of the information available in theInternational Financial Reporting Standards(IFRS-2003 edition) refers to SPEs’consolidation issues (IAS 27 and SIC-12Interpretation). However, the StandingInterpretation Committee-SIC-12 presents anSPE definition, which is more or less presentedin a similar way than the OECD one. The mainkey points of the definition can be summarisedas follows:

– an SPE is an entity created to accomplish anarrow and well-defined objective (e.g., toeffect a lease, research and developmentactivities or a securitisation of financialassets).

– an SPE can be a corporation, a trust, apartnership or an unincorporated entity.

– SPEs are often created with legalarrangements that impose strict and


sometimes permanent limits on the decision-making powers of their governing board,trustee or management over the operations ofthe SPE.

– the creator of the SPE frequently transfersassets to the SPE, obtains the right to useassets held by the SPE or performs servicesfor the SPE, while other parties may providethe funding to the SPE.

27. But the main issue raised in SIC-12 relatesto how SPEs should be consolidated, when itshould be done and which kind of SPE has to beconsolidated. In very simple terms, SIC-12states that:

– an enterprise should consolidate an SPEwhen, in substance, the enterprise controlsthe SPE.

– the concept of control used in IAS requireshaving the ability to direct or dominatedecision-making, regardless of whether thispower is actually exercised. The ability togovern decision-making must beaccompanied by the objective of obtainingbenefits from the SPEs’ activities.

– SPEs frequently operate in a predeterminedway so that no enterprise has explicitdecision-making authority over the SPE’songoing activities after this formation.Because virtually all activities arepredetermined, control may exist for thesponsoring party even though it may beparticularly difficult to assess.

– control may exist even in cases where anenterprise owns little or none of the SPE’sequity.

2.5 CONCLUSIONS

28. The exploration of various internationalguidelines has put forward the difficulty forcompilers to define SPEs in simple words.However, some common and internationally

agreed characteristics that define an SPE couldbe helpful for the identification and theharmonisation of the treatment of SPEs. Thefollowing conclusions rely mainly on the fact-finding exercise done in IMF or OECDManuals. Information relating to SPEs in theIFRS manual deals mainly with the issue linkedto the consolidation of SPEs and the concept ofcontrol in accounting standards.

29. SPEs are defined in different ways ininternational guidelines, the OECD benchmarkfocusing more on the purpose or the structure ofSPEs whilst the IMF puts forward the conceptof their engagement in few or no localoperations. But the essential point is that allthese definitions do not contradict each other.The OECD, in its report on Special PurposeEntities and Off-shore enterprises (p.9),merged both its own definition with the IMFone to present a common understanding ofSPEs according to international guidelines,which can be summarised as follows 3. SPEsare:


– engaged primarily in internationaltransactions but in few or no local operations.

– SPEs can be a financing subsidiary, a holdingcompany, a base company or a regionalheadquarter.

– SPEs can act as a sale and regionaladministration, a management of foreignexchange risk, or have the purpose offacilitating the financing of investment for thewhole group.

30. In practice, the concept of “few or no localoperation” seems to be the one which is more inline with national definitions currently available

3 Report on Special Purpose Entities and Off-shore enterprises –Workshop on International Investment Statistics, 5-6 March2003, paragraph 15, p. 9.

151c ECB


2



in few Member States (mainly DK, NL and, tosome extent, IE).

31. The definitions expressed above, mergingboth IMF and OECD definitions and notcontradicting the current national definitionswill constitute the basis for any future work thatcould be enhanced by Eurostat or the ECB in thefield of SPEs.

4) Regarding residence criteria, both the OECDand the IMF consider that offshore enterprises,including SPEs, have to be considered asresidents of the economy in which they areestablished.

5) Regarding the statute of SPEs and thetreatment of their FDI transactions, both theOECD and the IMF recommend to considerSPEs as resident direct investment enterprisesif they meet the general criteria defining FDIrelationships, therefore to treat their FDItransactions accordingly.

6) For SPEs with the sole purpose of serving ina financial intermediary capacity, allinternational guidelines recommend to restricttheir FDI transactions only to equity capital andpermanent debt (or fixed assets for branches).

7) The IMF Textbook (but not the 5th Manual) isthe only guide defining SPEs according to theiroperational presence in the host country (theOECD only raises the question in the “Taxauthorities’ perception” paragraph):

– SPEs are generally established in economiesother than those in which the parentcompanies are resident.

– SPEs are engaged primarily in internationaltransactions but in few or no local operations.

But nothing is said regarding the evolution ofthe SPEs’ population of enterprises (whatshould be done in case of an expansion of theactivities on the local market).

8) Both the OECD and the IMF recognise thatsome country may net out transactions throughSPEs, but they all recommend that nationalcompilers should provide information on agross basis, for international comparison.

9) All international guidelines agreed on thefact that SPEs can take the form of holdingcorporations. ESA95 has a precise definition ofholding corporations (§2.14), – institutionalunits whose main function is to control anddirect a group of subsidiaries. Theidentification of holding companies is neededfor the sectoral classification of SPEs in S11,S122 to S125. Such an identification isnecessary to indicate which FDI treatmentshould be adopted (the general FDI one or theone defined for SPE acting only as a financialintermediary).

10) Holding SPEs have to be classified in thesector which better reflects the activity of thegroup of subsidiaries as a whole (measuredon the basis of value added, ESA95 §2.23-e,2.40-e, 2.43).

11) ESA95 states that Holding corporationcontrolling a group of subsidiaries consistingpredominantly of insurance corporations andpension funds has to be classified in thefinancial intermediation sector (S123), insteadof the insurance one (ESA95 §2.43 and 2.63-b).However, on this specific topic, paragraph 70of the IMF “Monetary and Financial StatisticsManual” does not mention this exception. ThisManual states that (extract):

“A holding corporation is classified as financialif the preponderant activity of the group ofcorporations as a whole is financial.”

“Similarly, financial holding corporationsshould be allocated to subsectors according tothe type of financial activities mainly carried outby the group they control.”

12) For an SPE not having the statute ofholding, it is necessary to check if the SPE isacting as a financial intermediary or simply as a


treasury provider (ESA95 §2.34, 2.37, 2.55-f,IMF Textbook examples §543, 544, OECDBenchmark §69).

13) ESA95 states that financial intermediationdoes not include institutional units providingtreasury services, unless subject to financialsupervision.

– Units acting as treasury services, and notsubject to financial supervision, are allocatedto the sector better reflecting the predominantfunction of the group.

– Finance vehicle corporations holdingsecuritized assets have to be classified underS123, or S122 if they are MFIs.

153c ECB


3

Importanceof SPEs and

offshore countriesin European

statist ics

32. The TF-FDI highlighted the fact that not allthe MS identify SPEs in their regular statistics.Mainly NL and, to some extent, DK and IEreserve a specific treatment/definition fortransactions involving them. The CentralStatistical Bureau (CSB) of NL does not includethe gross transactions of SPEs in the nationalFDI statistics, nor in the data provided toEurostat. The CSB argues that transactions withSPEs do not affect the Dutch economy, andtherefore are not relevant for National Accountspurposes (no effect on the net worth). Despitethe fact that Eurostat and the ECB compileEuropean aggregates following the samemethod, the results obtained obviously diverge,which is not satisfactory from an analyticalpoint of view, and from the users’ one4.Investigations done within the TF-FDIconfirmed that the other Member States includetransactions with SPEs both in their datatransmission to ECB and Eurostat.

33. Eurostat is not able to evaluate directly theglobal contribution of SPEs using the currentEurostat/OECD questionnaire. Within thisquestionnaire, information relating to eitherFDI capital invested abroad by resident directinvestors acting as financial/managementholding companies, or invested in residentfinancial/management holding companies, canbe found. But it cannot be assumed that allholding companies could be considered asSPEs. Furthermore, non-holding SPEs couldexist if we refer to all the internationaldefinitions expressed in the previous section.For this reason, Member States have been askedby the TF-FDI to provide an estimation of thedirect contribution of SPEs in FDI statistics.

34. The Eurostat/OECD questionnaire asksinformation about FDI relationships with theso-called offshore centres, a group constitutedby “small” economic countries (in term ofGDP). FDI transactions are recorded accordingto the first shot criteria. It is questionablewhether large amounts (if not all) investeddirectly in the EU by offshore centres wouldneed to be reallocated if the UBO criteria had tobe applied. Reversibly, EU direct investment

3 IMPORTANC E O F S P E S A ND O F F S HOR ECOUNTR I E S I N E UROP E AN S TAT I S T I C S

towards offshore centres could be seen asinvestments for which we do not know exactlythe final destination.

35. Given the relative “disconnection” betweentheir GDP size and the amount of FDI capitalthey generate, it seems suitable to go on havinga close follow up of this specific population.Another reason rests on a possible correlationbetween resident SPEs and offshore countries,an assumption clearly stated in Annex 3 of theOECD Benchmark definition (“Location ofSPEs”, p. 45, 1st sentence). But thisassumption will be difficult to check, at least onEuropean data.

36. One of the purposes of the following tablesis to try to convince Member States toinvestigate whether it is necessary to have anin-depth analysis about SPEs and offshorecentres. All the following tables should beanalysed, bearing in mind that:

i) the Eurostat current list of the offshore(financial) centres5 cover in fact “non-European” offshore (financial) centres.

ii) this list has been updated recently (inOctober 2002, therefore mainly for the 2001and 2000 FDI data), which might altercomparisons over time. But the formerEurostat list already contained the mostsignificant offshore countries. Therefore, itcould be expected that these (recent)changes do not influence significantly thelevel of FDI transactions with offshoreentities.

4 Even though, from the purely national viewpoint, the exclusionfrom national statistics of SPE’s transactions may make sense,since, in some cases, national statistics could otherwise beblurred by the volume of financial transactions between non-resident entities channelled through domestic SPE’s.

5 Eurostat current list of offshore financial centres (31 countries):Antigua and Barbuda, Anguilla, Netherlands Antilles, Barbados,Bahrain, Bermuda, Bahamas, Belize, Cook Islands, Dominica,Grenada, Hong-Kong, Jamaïca, St Kitts and Nevis, CaymanIslands, Lebanon, Saint Lucia, Liberia, Marshall Islands,Montserrat, Maldives, Nauru, Niue, Panama, Singapore,Turks&Caicos Islands, Saint Vincent and the Grenadines, VirginIslands (UK), Virgin Islands (US), Vanuatu and Samoa.


3.1 OVERVIEW OF THE IMPACT OF OFFSHORECENTRES ON EUROPEAN AGGREGATES

37. Table 1 presents the evolution of EU FDItransactions with offshore centres since 1995,both in absolute value and with regard to thetotal extra EU transactions. The non-symmetrybetween outward and inward flows is quiteobvious: whilst the evolution is rather volatileon the outward side, EU inflows received fromoffshore centres have regularly increased since1997, from €4.9 bn to €13.4 bn. Furthermorethe issue linked to offshore centres seems to bemore relevant on the inward side: on averageover the 1995-2001 period, EU inflowsreceived from offshore companies accountedfor 9.1% of the total extra EU inflows. Tworeasons could explain the smaller percentage onthe outward side:

i) the Eurostat current list of offshore entitiesdoes not contain any European countries. It

might be possible that EU direct investorsare mainly dealing with similar entities closeto the EU border (inside or outside the EU),whilst EU direct investment companiesreceive FDI capital from foreign groupslocated in other continents (Caribbean areaor South East Asia zone).

ii) inward and outward flows related to SPEsreported from NL are not available in theEurostat database. May be it could fill inpart of the gap of overall flows to and fromoffshore countries.

38. For direct investors of the Eurozone, therelative interest for investing in offshorecountries is quite similar to the one observed atthe EU level (3.4% of extra Eurozone outflows,see Table 2). But the relative weight of inflowsfrom offshore companies is estimated to 4.1%of the total extra Eurozone inflows, 5percentage points less that the weight observed

Table 1 EU FDI f lows transactions with the rest of the world and offshore centres

(EUR millions)

1995 1996 1997 1998 1999 2000 2001 1995-2001

Outward

Extra EU -62,407 -68,665 -109,802 -218,754 -302,395 -408,925 -234,800 -1,405,749Offshore centres -3,337 -5,561 -6,811 -1,830 -8,971 -7,351 -12,332 -46,194% 5.3 8.1 6.2 0.8 3.0 1.8 5.3 3.3

Inward

Extra EU 42,464 36,509 50,160 96,432 102,118 150,407 118,470 596,559Offshore centres 3,427 2,583 4,932 5,816 10,234 13,628 13,403 54,023% 8.1 7.1 9.8 6.0 10.0 9.1 11.3 9.1

Table 2 Euro area FDI f lows with extra Eurozone and offshore centres

(EUR millions)

1999 2000 2001 1999-2001

Outward

Extra Eurozone -310,409 -411,863 -252,341 -974,613Offshore centres -5,523 -11,205 -16,241 -32,969% 1.8 2.7 6.4 3.4

Inward

Extra Eurozone 174,459 343,118 157,474 675,052Offshore centres 8,409 7,788 11,788 27,985% 4.8 2.3 7.5 4.1

Note: Transactions involving Dutch SPEs are excluded from Euro area calculations.

155c ECB


2



Table 3 FDI posit ions with the rest of the world and offshore centres

(EUR millions)

European Union Euro areaValues at end 2000 Values at end 2000

Assets

Extra EU/Eurozone 1,517,186 1,335,073Offshore centres 77,236 44,493% 5.1 3.3

Liabilities

Extra EU/Eurozone 890,709 1,112,495Offshore centres 65,800 43,790% 7.4 3.9

Note: Transactions involving Dutch SPEs are excluded from EU and Euro area calculations.

on EU aggregates. Despite this gap, it seemsreasonable to consider that percentagesobserved on the inward side are quitesubstantial, which justify the need of furtherinvestigation for analytical purposes.

39. The relative importance of the group ofoffshore countries is also confirmed by the FDIpositions data, however with a better balancingbetween assets and liabilities: 5.1% and 7.4%respectively, for the European Union as awhole, 3.3% and 3.9% for the Eurozone as awhole entity.

40. Table 4 presents the classification (indescending order) of the main extra EU FDIpartners, both on the assets and liabilities sides.At the moment, Eurostat doesn’t have thepossibility to offer a full breakdown of EUdirect investment vis-à-vis all extra EUindividual countries. The current list involves50 individual extra-EU countries, and it hasbeen necessary to introduce some groupingzones (eight in all) to better approximate a fullcoverage of the extra EU area (see footnotesunder Table 4). None of these additional groupscontains any of individual listed countries and,apart from the “Gulf Arabian countries” (seefootnote 5), this list also avoids the possibilityof having one non-listed country being includedin more than one economic group.

41. When looking at the regional distribution,the results clearly illustrate the dominant

positions of offshore companies, in the field ofFDI:

i) assets side: offshore financial centres,excluding Singapore and Hong-Kong, hostedaltogether €41 billion of EU external FDIassets, ranking in the top 6 countries’ list justbehind Argentina and before Australia.

ii) liabilities side: offshore financial centres,excluding Singapore and Hong-Kong, werealtogether responsible of €52 bn of EUexternal FDI liabilities, ranking in the 3rdposition just behind Switzerland and beforeJapan.

iii) if we add Singapore and Hong-Kong, then itcould be seen that the group formed byoffshore financial centres, as defined in thecurrent Eurostat list, is the 3rd EU partnerboth on the assets and liabilities side (€77bn on the assets side, €66 bn on theliabilities side).

42. From this table it could also be possible toportray also the impact linked to a possibleextension of the offshore list. For the moment,no official “offshore” list has been established/approved by the International Institutions.Nevertheless, the “Other European countries”group6 defined by Eurostat involves countries

6 Andorra, Faroe Islands, Guernsey, Gibraltar, Isle of Man, Jersey,Moldova, Macedonia, San Marino and Vatican City State.


EU external FDI ASSETS (end 2000) EU external FDI LIABILITIES (end 2000)

US 762,842 US 546,014CH 106,323 CH 128,962BR 69,325 Offshore centres 1) 52,034CA 64,574 JP 47,326AR 44,684 CA 36,719Offshore centres 1) 41,006 AU 19,001Other European c. 2) 38,426 NO 18,478AU 30,042 Other European c. 2) 18,438JP 29,271 HK 1) 7,101NO 26,660 SG 1) 6,664PL 23,940 Gulf Arabian countries 5) 5,606SG 1) 21,533 LI 5,451Other Africa 3) 18,701 Other Africa 3) 2,979MX 18,305 RU 2,739HU 16,597 BR 2,601HK 1) 14,696 NZ 2,101CL 14,425 CY 1,951CZ 14,306 IR 1,669CN 13,788 IL 1,666ZA 13,389 ZA 1,599KR 8,194 MX 1,534Other South America 4) 7,646 TR 1,321VE 6,341 KR 1,273TR 6,228 MY 1,056Gulf Arabian countries 5) 5,931 Other North African c. 6) 1,000MY 5,878 AR 979TW 5,566 TW 688IN 5,303 PL 615RU 5,003 VE 595CO 4,817 Other Near & Middle East 586Other North African c. 6) 4,255 CN 584TH 4,057 ID 583NZ 3,944 MA 539EG 3,697 UY 533PH 3,652 IN 511MA 3,285 MT 465SK 3,244 EG 245ID 2,561 HU 239CY 2,520 IS 234RO 2,227 Other Oceanian c. 7) 183Other Near & Middle East 1,840 CZ 170HR 1,697 TH 152LI 1,567 CO 150IL 1,541 SI 149SI 1,387 BG 116EE 1,122 RO 103LT 1,058 YU 90UY 1,022 PH 86BG 1,012 Other South America 4) 85MT 871 UA 42LV 815 HR 37Other Oceanian c. 7) 791 SK 33IR 677 BY 21UA 526 AL 11YU 467 LT 9IS 143 LV 2AL 117 EE -1BY 34 CL -15

Table 4 Main extra EU partners countries, according to (end 2000) FDI positions

1) Eurostat current list, excluding Singapore and Hong Kong (shown separately).2) Andorra, Faroe Islands, Guernsey, Gibraltar, Isle of Man, Jersey, Moldova, Macedonia, San Marino and Vatican City State.3) Other African countries (as defined by Eurostat) minus South Africa.4) South America, excluding AR, BR, CL, CO, UY and VE (shown separately).5) The Gulf Arabian countries’ list include Bahrain, whis is also an offshore (financial) centres.6) Algeria, Tunisia and Libya.7) Oceanian countries – Australia – New-Zealand.

157c ECB


2



having (more or less) similar characteristics tothose from the Eurostat offshore list. If we hadto include them in the Eurostat list, then the totalamount of EU external FDI assets located inoffshore countries would have jumped from€77 to €115 billion, in the 2nd position behindthe United States and far ahead of Switzerland.

43. These results put forward the powerful roleplayed by offshore centres in the field of FDI, ifwe consider them as a unique entity. With theperspective of improving internationalcomparison it is questionable whether thepresence of offshore entities alter significantlythe regional distribution picture.

44. The extra EU FDI capital structure couldalso point out the specificity of these entities, asit is shown in Table 5: On the liabilities side, thecapital distribution between “equity capital andRIE” and “other capital” is around two third/onethird respectively. For extra EU FDI liabilitiesvis-à-vis foreign offshore companies, the FDIcapital distribution profile divergessignificantly from the extra EU average: 43% ofthe total assets held in the Union by offshorecountries were constituted by inter-companydebt stocks.

3.2 IMPACT OF OFFSHORE (FINANCIAL) CENTRESAT NATIONAL LEVEL

3.2. 1 OVERVIEW45. Data presented in this chapter refer to FDIpositions observed at end 2000, mainly becauseof high volatility observed on the data flows. In

Table 5 Extra EU FDI posit ions capital structure at end 2000

(%)

Equity cap. & RIE Other Capital Total

Assets

Extra EU assets, exc. offshore centres 79 21 100Offshore centres 92 8 100

Liabilities

Extra EU liabilities, exc. offshore centres 65 35 100Offshore centres 57 43 100

addition, Eurostat do not have any FDI datawith offshore centres, reported by Belgium/Luxembourg, Greece, Spain and Ireland, whichconstitute another restriction of the analyticalframework. Data for Switzerland and the UnitedStates have been added to enlarge the overviewat national level. For the United States,information has been extracted from regulartables available in the Bureau of EconomicAnalysis (BEA) website (www.bea.doc.gov).Offshore aggregates reported by the USA havebeen calculated by Eurostat (according to theEurostat current list), by summing up allavailable information in their detailedgeographical breakdown list. Given the fewnumber of missing (confidential) information, itcan be assumed that the estimated resultsobtained are reliable.

46. At first glance, Table 6 seems to point out arather limited impact of offshore centres, inMember State FDI positions statistics7, with amaximum around 6% (Austria) for the assetsside, slightly below 10% for the liabilities side(Denmark).

47. Amazingly, significant percentages shares –close to 14% – can be observed on Swiss andUS assets side. For the US data, a possibleexplanation could be linked to the fact that theEurostat current offshore list favours mainlyUS companies, as most of offshore countriesrecorded in this list are concentrated in the

7 Here the relative importance is measured on MS total externalFDI positions, i.e. with the world total. It explains partially thelower percentage level observed on MS figures, in comparisonwith those shown on European aggregates.


Caribbean Islands. On the liabilities side,offshore entities have more or less the sameimportance on the US figures as those observedon European aggregates (4.6%). On thecontrary, FDI activities of offshore companies,in Switzerland, are negligible.

3.2.2 FOCUS ON US DATA48. Would the application of UBO criteria alterthe importance of offshore entities? A firstattempt to this answer is given by the analysisof US figures, shown in Annex C.

The main comments are synthesised in the threefollowing points:

1) The application of the UBO criteria does notalter the large influence of offshorecountries8 in US statistics. Surprisingly, USfigures indicate even a reinforcement of thisgroup, when going from first shot to UBO.

2) However, if offshore countries are analysedindividually, the role played by each of themis drastically affected by the UBO countryreallocation.

3) There are apparently large differencesbetween offshore centres subject to their

Table 6 Importance of offshore centres inMS FDI statistics(%)

FDI positions at end 2000

Reporter Assets Liabilities

Belgium/Luxembourg : :Denmark 0.4 9.7Germany 2.2 0.4Greece : :Spain : :France 1.5 0.9Ireland : :Italy 3.6 1.9Netherlands 1) 1.5 5.8Austria 6.1 0.3Portugal 5.2 5.2Finland 0.7 0.0Sweden 1.1 :United Kingdom 3.2 3.1

Switzerland 13.6 0.9United States 13.8 4.6

Data with offshore centres are not available.1) Excluding SPEs

economic background. It seems that themove from first shot to UBO does notalways enable the identification of the UBOowners’ nationality. In fact, the constraintsencountered by the BEA for a fullapplication of their UBO definition are notknown.

3.3 RESULTS OF THE EMPIRICAL EXERCISE ONTHE IMPORTANCE OF SPES IN EU MEMBERSTATES

An empirical investigation concerning the roleof SPEs was conducted by the TF-FDI. Theoutcome of the empirical exercise on the impactof SPEs in b.o.p. and i.i.p. data illustrates thedifficulties of such an exercise in the absence ofa specific definition of SPEs in a majority ofcountries: four countries replied that they couldnot provide any information and five (BE, DE,FI, FR and GR) provided data, essentially byapproximating the definition of SPEs with oneor more of the following: co-ordination centres,financial holding companies, managementholding companies and “other services forenterprises”.

The answers received are attached in Annex D.No meaningful aggregation has been deemedfeasible, given the disparity of the proxies usedto compile the data.

The examination of the figures provided showthat the amounts involved both for flows andfor stocks are quite significant, especially inDE. As the four countries which have provideddata for the exercise are not known to beparticularly attractive for SPEs, (contrary toNL), this would indicate that the validity of theapproximation SPEs = holding companies isquestionable.

As already said, SPEs cannot be identified assuch within the NACE classification of

8 Offshore centres estimated with Bahamas +Bermuda+Netherlands Antilles +Panama +UK Islands, Caribbean +OtherOWH +Liberia +Lebanon +Hong Kong +Singapore.

159c ECB


2



activities. In international guidelines, SPEs areidentified not only according to their maineconomic activity, but also through theirstructure’s analysis (holding or non-holdingcorporation).

As some Member States proceed, Eurostatapproximated the importance of some type ofSPEs using a correspondence table to link theOECD definition (Holding company, regionalheadquarters, Management of foreign exchangerisk etc…) with the main concerned classes/groups (of the NACE rev1) for hosting SPEs.Annex E gives an overview of this linkagetogether with an estimation of the impact ofthese groups, at the European level.

The fact that, apart from NL, DK and, to a lesserextent, IE, there exists no specific definition ofSPEs in EU countries may also indicate thatthere is no real need for such a definitionbecause the issue is not important in thesecountries.

3.4 SPECIAL TREATMENTS OF SPES IN EUMEMBER STATES

As already mentioned SPEs play an importantrole in a few countries and therefore specialtreatments were established in these countries.As the net flows of SPE transactions throughNL are close to zero and hardly affect thenational economy, the gross flows are notincluded in the national FDI statistics of NL.DK has still not decided if certain SPE activitiesshould be excluded for national reporting andIreland identifies separately a part of its SPEpopulation. The treatment of SPEs in DK, IEand NL is described in Annex F.

3.5 CONCLUSIONS

1. Offshore financial centres are important FDIpartners, having a significant impact onEuropean statistics. For internationalcomparison, it would be necessary to have an

agreed list of offshore financial countries,which is also stable over time.

2. The US figures show that a moving from firstshot to UBO criteria does not “globally” alterthe importance of offshore financial centres.But it seems necessary to distinguish betweenoffshore centres with a real economicbackground and offshore centres whichattractiveness is merely based on fiscalincentives. Such a distinction could also help tofurther elaborate the final offshore listexpressed in 1. However, a high sensitivenessto change in criteria classifications has beenobserved for each offshore country takenindividually. The use of UBO implies ageographical breakdown of FDI data betterreflecting economic reality. But results obtainedon US figures leave some “open questions” inthe sense that it is not sure whether they haveidentified the real UBO company: between thefirst shot identified partner and the real UBOone, FDI capital could be in the hand of severalintermediary offshore entities, and we do notknow the real reasons for assuming direct FDIrelations between EU SPEs and “offshoreSPEs”.

3. The restricted analysis on holding activities,using the Eurostat/OECD questionnaire byactivity, has put forward the “potential” impactof SPEs in FDI statistics, both on Europeanaggregates and at the Member States level. Atthe moment, a more accurate measure of SPEs’transactions is not feasible: None of currentlyavailable definitions – international guidelinesor the national one – considers all holdingcompanies as SPEs entities.


The analytical interest of a separateidentification of transactions/positions relatedto SPEs was derived in chapter 1. However,various problems in dealing with SPEs restrictthe separate identification of SPEs. The mainproblem in identifying SPEs is simply theabsence of an appropriate data basis. Theempirical investigations on the importance ofSPEs among member states has revealed thateven the identification of domestic SPEs is atpresent time obviously not possible for mostmember states. At the best SPEs could roughlybe approximated by certain sectors potentiallyincluding SPEs. Moreover, it seems that a fullanalysis would also require the identification ofnon-resident SPEs when they are counterpart toa transaction/position of a resident company,which seems an impossible goal to achieve.

The importance of SPEs varies significantlyamong member states. Consequently, practicesand efforts in identifying SPEs and theapplication of special treatments differ betweencountries. Potential asymmetries betweennational statistics respectively differences indata dissemination to internationalorganisations should be avoided or properly beexposed.

The TF-FDI recommends the inclusion oftransactions/positions of/with SPEs or SPE-likecompanies in b.o.p./i.i.p. reporting concerningthe contributions to the euro area/EUaggregates.

Notwithstanding all the practical andconceptual difficulties previously stated, theTF-FDI recommends that the possibility tocollect separate statistics for SPEs continuebeing assessed by both working groups and inthe framework of ad hoc workshops in thefuture. To this aim, coordination should beensured with the related work currently beingdeveloped in the OECD.

Additionally, chapter 59 recalled the decision ofthe IMF, in co-ordination with the ECB’s WG-BP&ER and the OECD’s Working Party onFinancial Statistics (WPFS), concerning inter-

4 CONC LU S I ON Scompany loans between affiliated MFIs. Inparticular, there is an explicit reference in theIMF resolution to SPEs principally engaged infinancial intermediation for a group of relatedenterprises.

According to the IMF decision, SPEsprincipally engaged in financial intermediationfor a group of related enterprises should beincluded in the category of affiliated financialintermediaries and, therefore, inter-companyloans with any other institution included in thiscategory should be excluded from directinvestment and should be recorded in otherinvestment10.

9 See annex 1.10 Permanent debt should still be recorded in direct investment.

161ECB

Foreign direct investment – Task force reportMarch 2004

ANNEX A

Examplesof SPE transactions

throughthe Netherlands

Holding companies are one of the possible typesof SPEs in the Netherlands. By order of theirforeign parent company these SPEs not onlymanage participations but also distribute dividendsgained from different participations. In general,the inward and outward equity capital transactionsshould be registered at their market value. Basedon the reporting of an SPE, two examples of afictive SPE are presented below. In the firstexample the purchase of the participation by DutchSPE is booked at market value while in the secondexample the purchasing price is equal to the bookvalue. The holding gains realised in the secondcase are extra ordinary high and should not beconsidered as Direct Investment income.

CASE A, REALISING GAIN ACCORDING TO COPC:1. Holding company B is a Dutch SPE and for

100% owned by German multinational

ANNEX A : E X AMP L E S O F S P E T R AN S A C T I ON STHROUGH THE N E THER L AND S

enterprise A. Company A has transferred itsGerman subsidiary X (also Germanresident) to company B in for an amount ofEUR 100 (market value). As a result of thistransfer the equity capital of A in B hasincreased by the same amount.

2. After a year company B has sold the sharesof the German subsidiary X via the NewYork stock exchange for EUR 110.

3. Company B has distributed the receivedamount including the realised earning to itsparent company A. The EUR 10 realisedgain can be considered as earning accordingto COPC.

These transactions are illustrated in Chart 1.

A

B

NEW YORK

STOCK EXCHANGE

(2) X-shares EUR 100 (2) Cash EUR 110

(1) B-shares EUR 100 (1) X-shares EUR 100

(3) EUR 110repayment ofinvestmentand distributionof dividend

US

NL

DEX

Impact of these transactions to the balance of payments of The Netherlands:(1) – Inward equity capital investment by Germany in the Netherlands by EUR 100 (investment

of A in B)– Outward equity capital investment by the Netherlands in Germany by EUR 100 (X-sharesto B)

(2) – (Increase in cash by EUR 110)– Outward equity capital disinvestment by EUR 110

(3) – (Decrease in cash by EUR 110)– Inward equity capital disinvestment by EUR 100 (Disinvestment of A in B)– Distributed dividend by B to A by EUR 10

In this example the outward DI income of the Netherlands to Germany is equal to EUR 10.

Chart 1 Real is ing gain according to the Current Operating Performance Concept

162ECBForeign direct investment – Task force reportMarch 2004

CASE B, REALISING EXTRA ORDINARY HOLDINGGAINS:1. Holding company B is a Dutch SPE and for

100% owned by German multinationalenterprise A. Company A has transferred itsGerman subsidiary X (also Germanresident) to company B in for an amount ofEUR 100 (book value). As a result of thistransfer the equity capital of A in B hasincreased by the same amount.

2. After a year company B has sold the shares ofthe German subsidiary X via the New York

stock exchange for EUR 200. This results inan extra ordinary holding gain of EUR 100.

3. Company B has distributed the receivedamount including the realised holding gainto its parent company A. The EUR 100holding gain from the sale of X can not beconsidered as earning according to COPCand consequently not as FDI income.

These transactions are illustrated in Chart 2.

Impact of these transactions to the balance of payments of The Netherlands:(1) – Inward equity capital investment by Germany in the Netherlands by EUR 100 (investment

of A in B)– Outward equity capital investment by the Netherlands in Germany by EUR 100 (X-shares

to B)(2) – (Increase in cash by EUR 200)

– Outward equity capital disinvestment by the Netherlands in Germany by EUR 200(X-shares to US)

– Adjustment (difference between market value and book value) of the inward equity capitalinvestment by Germany in the Netherlands by EUR 100

– Adjustment (difference between market value and book value) of the outward equitycapital investment by the Netherlands in Germany by EUR 100

(3) – (Decrease in cash by EUR 200)– Inward equity capital disinvestment by Germany in the Netherlands by EUR 200

(Disinvestment of A in B)

A

B

NEW YORK

STOCK EXCHANGE

(2) X-shares EUR 100 (2) Cash EUR 200

(1) B-shares EUR 100 (1) X-shares EUR 100

(3) EUR 100holding gain

US

NL

DEX

Chart 2 Real is ing extraordinary holding gains

163cECB


ANNEX B

Overview of thedifferent definitions

of SPEs andoffshore centresin international

guidelines

This annex presents the different definitions ofSPEs and offshore financial centres, as they areliterally expressed in international guidelines.Main international guidelines explored are theIMF 5th Manual, the IMF BOP textbook, theIMF Balance of Payment compilation guide, theOECD Benchmark definition of Foreign DirectInvestment, The ESA95, and the InternationalFinancial Reporting Standards-2003 Edition(incorporating IAS and interpretations).

IMF 5TH MANUALThe Manual deals with SPEs issues, but noexplicit definition can be found. SPEs issuesare expressed in paragraphs 79, 365, 372 and373 and 381.

§ 7 9“Offshore enterprises engaged in manufacturingprocesses (including assembly of componentsmanufactured elsewhere) are residents of theeconomies in which the offshore enterprises arelocated. This statement applies regardless oflocation in special zones of exemption fromcustoms or other regulation concessions. Thestatement also applies to non-manufacturingoperations (i.e. trading and financial enterprises),including so-called special purpose enterprises.(See paragraphs 365 and 381.)”

§ 3 8 1“The residency of offshore enterprises, includingthose engaged in the assembly of componentsmanufactured elsewhere and in trade and financialoperations and those located in special zones, isattributed to the economy in which the enterprisesare located. (See paragraph 79)”

§ 3 6 5“This Manual recommends that so-called specialpurpose entities (SPEs) be included as directinvestment enterprises if they meet the criteriastated in previous paragraphs. Whatever thestructure (e.g., holding companies, base company,regional headquarters) or purpose (e.g.,administration, management of foreign exchangerisk, facilitation of financing of investments),SPEs are an integral part of the structure of thedirect investment network as are, for the most

ANNEX B : O V ERV I EW O F TH E D I F F E R ENTDE F I N I T I ON S O F S P E S AND O F F SHORE C EN TR E SI N I N T E RNAT I ONA L GU I D E L I N E S

part, SPE transactions with other members of thegroup. However, for SPEs with a sole purpose ofserving in a financial intermediary capacity (as isthe case for banks and other financialintermediaries such as security dealers),transactions recorded under direct investment arelimited to those associated with permanent debtand equity. (See paragraph 372.) For bothcountries employing other treatments of SPEs andcountries employing the recommended treatment(if it is feasible to do so), the value of SPEtransactions as a group should be separatelyidentified in terms of standards components topermit consistent international comparisons.”

§ 3 7 2“Intercompany transactions between affiliatedbanks (depository institutions) and affiliatedfinancial intermediaries (e.g., security dealers)– including SPEs with the sole purpose ofserving as financial intermediaries – recordedunder direct investment capital are limited thosetransactions associated with permanent debt(loan capital representing a capital interest) andequity (share capital) investment or, in the caseof branches, fixed assets. Deposits and otherclaims and liabilities related to usual bankingtransactions of depository institutions andclaims and liabilities of other financialintermediaries are classified, as appropriate,under portfolio investment or other investment.The stock of foreign assets and liabilities ofbanks and other financial intermediaries(international investment position) should betreated in a parallel manner.”

§ 3 7 3“Transactions through SPEs (with theexceptions noted in paragraphs 365 and 372)are included in direct investment capitaltransactions, and the related stocks of assetsand liabilities are covered in the directinvestment position.”

IMF BALANCE OF PAYMENTS TEXTBOOK

§ 5 4 2“Special Purpose Entities (SPEs) are (1) generallyorganised or established in economies other than


those in which the parent companies are residentand (2) engaged primarily in internationaltransactions but in few or no local operations.SPEs meeting the criteria presented in paragraphs514-518 are included, with one exception, asdirect investment enterprises. Excepted are SPEswith the sole purpose of serving as financialintermediaries; for these, investments recordedunder direct investment are limited to equitycapital and permanent debt.”

§543 gives an example of typical SPEs’transactionsTwo examples illustrate the BOP treatment ofSPEs. An Australian enterprise sets up anenterprise in Bermuda with share capital of $2.The enterprise is

(1) To purchase and hold $2 million of portfolioequity investment in the United States;

(2) To purchase and hold $1 million of bondsissued by a German company;

(3) To purchase, for $5 million, and hold a 50%interest in a United Kingdom company.

Half of the $8 million required for the investmentis provided by the Australian direct investor andhalf is provided by a bank in the NetherlandsAntilles. Bermuda’s balance of Payments wouldshow the following transactions:

Credit Debit

1. Direct investmentDirect investmentin Bermuda

Equity capital $2 (AU)Other capital $4,000,000 (AU)

Direct investmentabroad

Equity capital $5,000,000 (UK)Other capital

2. Portfolio InvestmentAssets

Equity securities $2,000,000 (US)Debt securities

Bonds and notes $1,000,000 (DE)3. Other investment

LiabilitiesLoans

Banks $4,000,000(NL Antilles)

As the enterprise in Bermuda is not purely afinancial intermediary, BOP transactions withrelated enterprises are recorded on the samebasis as other direct investment transactionsare-although the enterprise has no operations inBermuda.

§544 gives an example of typical SPEs’acting as purely financial intermediary“A New Zealand company wishes to borrowfunds on the US capital market by issuingbonds valued at $3 million. Under USregulations, only resident companies areallowed to issue such securities on the USmarket. So the New Zealand companyestablishes “a $2 subsidiary” in Delaware (a USstate) and the subsidiary issues bonds and lendsthe proceeds to its parent. As this SPE actspurely as a financial intermediary, only equitycapital and any permanent debt provided by thedirect investor are classified as directinvestment. The following transactions wouldbe recorded in New Zealand’s balance ofpayments (all with the USA):

Credit Debit

1. Direct investmentDirect investment abroad

Equity capital $22. Portfolio Investment

LiabilitiesDebt securities

Bonds and notes $3,000,0003. Other investment4. Reserve assets $2,999,998

(or other appropriatefinancial account item)

IMF BALANCE OF PAYMENTS COMPILATION GUIDEThe IMF BOP compilation guide doesn’t haveexplicit reference to SPEs, but deals with “Non-operating direct investment enterprises”, aterminology that is likely to be close to SPEs.These entities are covered along paragraphs705-711 of the guide. In addition, it seems thatexamples shown relate only to the ownership ofshipping vessels.

§ 7 0 5“While compilers should identify and collectinformation from all legal entities that fall

165cECB


ANNEX B



guidelines

within the definition of direct investmententerprises, adequate data may be unavailable.Of particular concern are brass plate companies,such as those established to register ownershipof shipping vessels or to raise capital throughthe issuance of securities.”

§ 7 0 6“To take advantage of various legislation, certaincompanies may register in a country but, for allpractical purposes, have no operational presencein that country. (Some security markets – forexample, those in the United States – permitsecurities to be issued only by locally registeredcompanies.) That is, the company do not carry outproduction, have no employees, and do not payincome tax. Many companies established for thepurpose of issuing securities may have no otherpresence in a host country. Brass plate companiesmay pay a fee to register in a host country and mayshare an office or directors with similarenterprises. However, books or accounts may bemaintained elsewhere and, thus, be unavailable tothe host country compiler.”

§ 7 0 7“Despite the difficulties caused by thesearrangements, compilers should make every effortto compile complete sets of accounts for theseenterprises. Countries that permit registration ofthese enterprises may also exempt them fromsupplying information that compilers require.However, some suitable data may be availablefrom tax or other authorities. Alternatively,compilers may approach partner countrycompilers for information. In the country of thedirect investor, the collection of data should besomewhat easier, and it is desirable thatinformation on certain categories of enterprises becompiled separately so that relevant data can beprovided (subject, of course, to anyconfidentiality constraints) to partner countries toassist them in compiling complete accounts.”

§ 7 0 8“According to the BPM, compilers shouldrecord the complete BOP entries of theseenterprises. However, some compilers mayprefer not to record transactions considered to

Credit Debit

1. Current AccountServices (incidental expenses) 5Transfer (registration fee) 25

2. Financial AccountOther investment

Bank assets 30

Credit Debit

1. Current AccountGoods 1,000Services

Leasing of vesselwithout crew 110

IncomeDirect investment income

Distributed income 52. Financial Account

Direct investment inthe reporting

Equity capital 1,030 105Other investment

Bank assets 30

be of no relevance to the domestic economy.Nevertheless, compilers should, for purposesof reporting to the IMF, prepare the gross entryas supplementary data.”

§ 7 0 9“The following example illustrates alternativemethods of recording. A brass plate company isestablished in one country for the purpose ofowning a shipping vessel operated by a non-resident parent enterprise located elsewhere. Inthe relevant period, the cost of registering thevessel is 25, and incidental expenses in thecountry are 5. The operator pay 110 to lease thevessel, and this amount is immediately remitted bythe enterprise to the non-resident owner. If thisbrass plate enterprise is essentially ignored, BOPentries for the country of registrations would be:

§ 7 1 0“However, the treatment required in the BPMrequires the gathering of additionalinformation. The value of the vessel at the timeof acquisition by the brass plate company is1,000, and the vessel depreciates by 75 duringthe period. According to recommendations ofthe BPM, BOP accounts for the country ofregistration should be:


§ 7 1 1“In this case, the direct investment inflow is1,030, which is equal to the value of the vesselplus the cost of registration and incidentalexpenses. The operating profit of the enterpriseis the difference between the revenue of 110 andexpenses (registration, incidental, anddepreciation) of 105. All of this is remitted tothe direct investor. A further 105 is remittedand, as the enterprise has no retained profitsfrom previous periods, this remittance

OECD BENCHMARK DEFINITION OF FOREIGNDIRECT INVESTMENT – 3RD EDITION

The OECD Benchmark definition deals withoffshore centres in its paragraphs 67 and 68,with SPEs in paragraph 69. Annex 3 is alsopresented, given that this annex gives verydetailed explanations or recommendations toidentify SPEs.

Offshore Companies

§ 6 7 .“There are a number of companies incorporatedin one country with their management office inanother country which do not trade in theircountry of incorporation. The managementoffice holds all the assets of the company andthe only transaction with the country ofincorporation is that the management officepays dividends on behalf of the company to anyresident shareholder of the company in thecountry of incorporation. These companies mayalso have direct investment in third countries,the dividend and capital flows then beingbetween the third countries and the country ofthe management office. The third country willprobably assume in its statistics that thesetransactions are with the country ofincorporation, while the management office’scountry will probably show the transactions asbeing between it and the third countries. Anadded complication is that the company owningthe management office may itself be asubsidiary of an enterprise in another country.”

§ 6 8 .“OECD recommends that where a company Zincorporated in Country A has its managementoffice in another Country B, Country A in itsoutward direct investment statistics regard theforeign management office as direct investmentby Country A in a branch in Country B. Ifcompany Z has any subsidiary and associatecompanies, these should be regarded as beingdirectly owned by the foreign branch in CountryB and thus only indirectly owned by companyZ. The host countries of the subsidiary andassociate companies should in their inwarddirect investment statistics regard the immediateinvesting country as being that of themanagement office, that is Country B, andregard the ultimate investing country either asCountry A, the country of incorporation ofcompany Z, or as Country C if company Z isitself a subsidiary with its ultimate parent inanother Country C. Country B, the country ofresidency of the management office, should inits inward direct investment statistics regard themanagement office as an inward branch ownedby Country A, with the ultimate investingcountry being Country C if company Z is asubsidiary with its ultimate parent in CountryC. Country B in its outward direct investmentstatistics should regard the subsidiaries andassociates of company Z that are not resident inCountry B as part of Country B’s outwarddirect investment.”

Special Purpose Entities

§ 6 9 .“As multinational enterprises mature, theydiversify their investments geographically,through adequate organizational structures.These include certain Special Purpose Entities(SPEs) which facilitate financing ofinvestments for the group from sources bothinternal and external to the multinationalenterprises. Additionally, such SPEs also serveother functions such as sale and regionaladministration including management of foreignexchange risks and other activities aimed atprofit maximization. Special Purpose Entity is ageneric label applicable to such organizational

167cECB


ANNEX B



guidelines

structures which are also variously referred toas financing subsidiaries, conduits, holdingcompanies, base companies and regionalheadquarters. In some instances, multinationalenterprises use existing operational companiesto perform functions usually associated withSPEs. Since these SPEs are an integral part ofthe organizational structure of a multinationalenterprise, their transactions that arise fromdirect investment relationships (except as notedin paras [38] and [39 c]) should be reflected inthe statistics and, if possible, shown as a sub-component. In some instances, thesetransactions may give rise to negative directinvestment positions. For a more detaileddescription of the SPEs or of transactions thatexhibit the characteristics of functions carriedout through these entities, refer to Annex 3.”

INTERNATIONAL ACCOUNTING STANDARDS (IAS)

Information about SPEs can be found in the2003 edition of the International FinancialReporting Standards (IFRS) manual: IAS27deals with the issue of Consolidated FinancialStatements and Accounting for Investment insubsidiaries, and interpretations-SIC12 ofIFRS relates to how/when/which SPE should beconsolidated.

INTERPRETATIONS: SIC 12-1: CONSOLIDATION –SPECIAL PURPOSE ENTITIES

ISSUE1. An entity may be created to accomplish anarrow and well-defined objective (e.g., toeffect a lease, research & development activitiesor a securitisation of financial assets). Such aspecial purpose entity (“SPE”) may take theform of a corporation, trust, partnership orunincorporated entity. SPEs often are createdwith legal arrangements that impose strict andsometimes permanent limits on the decision-making powers of their governing board,trustee or management over the operations onthe SPE. Frequently, these provisions specifythat the policy guiding the ongoing activity ofthe SPE cannot be modified, other than perhaps

by its creator or sponsor (i.e., they operate onso-called “autopilot”).

2. The sponsor (or enterprise on whose behalfthe SPE was created) frequently transfers assetsto the SPE, obtains the right to use assets heldby the SPE or performs services for the SPE,while other parties (“capital providers”) mayprovide the funding to the SPE. An enterprisethat engages in transactions with an SPE(frequently the creator or sponsor) may insubstance control the SPE.

3. A beneficial interest in an SPE may, forexample, take the form of a debt instrument, anequity instrument, a participation right, aresidual interest or a lease. Some beneficialinterests may simply provide the holder with afixed or stated rate of return, while others givethe holder rights or access to other futureeconomic benefits of the SPE’s activities. Inmost cases, the creator or sponsor (or theenterprise on whose behalf the SPE wascreated) retains a significant beneficial interestin the SPE’s activities, even though it may ownlittle or none of the SPE’s equity.

4. IAS 27 requires the consolidation of entitiesthat are controlled by the reporting enterprise.However, the Standard does not provideexplicit guidance on the consolidation of SPEs.

5. The issue is under what circumstances anenterprise should consolidate an SPE.

6. This interpretation does not apply to post-employment benefit plans or equitycompensation plans.

7. A transfer of assets from an enterprise to anSPE may qualify as a sale by that enterprise.Even if the transfer does qualify as a sale, theprovision of IAS 27 and this Interpretation maymean that the enterprise should consolidate anSPE. This Interpretation does not address thecircumstances in which sale treatment shouldapply for the enterprise or the elimination ofthe consequences of such a sale uponconsolidation.


CONSENSUS

8. An SPE should be consolidated when thesubstance of the relationship between anenterprise and the SPE indicates that the SPE iscontrolled by that enterprise.

9. In the context of an SPE, control may arisethrough the predetermination of the activities ofthe SPE (operating on “autopilot”) orotherwise. IAS 27.12 indicates severalcircumstances which result in control even incases where an enterprise owns one half or lessof the voting power of another enterprise.Similarly, control may exist even in cases wherean enterprise owns little or none of the SPE’sequity. The application of the control conceptrequires, in each case, judgement in the contextof all relevant factors.

10. In addition to the situations described inIAS 27.12, the following circumstances, forexample, may indicate, a relationship in whichan enterprise controls an SPE and consequentlyshould consolidate the SPE (additionalguidance is provided in the Appendix to thisInterpretation).

(a) in substance, the activities of the SPE arebeing conducted on behalf of the enterpriseaccording to its specific business needs so thatthe enterprise obtains benefits from the SPE’soperation;

(b) in substance, the enterprise has thedecision-making powers to obtain the majorityof the benefits of the activities of the SPE or, bysetting up an “autopilot” mechanism, theenterprise has delegated these decision-makingpowers;

(c) in substance, the enterprise has rights toobtain the majority of the benefits of the SPEand therefore may be exposed to risks incidentto the activities of the SPE; or

(d) in substance, the enterprise retains themajority of the residual or ownership risks

related to the SPE or its assets in order to obtainbenefits from its activities.

11. Predetermination of the ongoing activitiesof an SPE by an enterprise (the sponsor or otherparty with a beneficial interest) would notrepresent the type of restrictions referred to inIAS 27.13(b).

BASIS FOR CONCLUSIONS12. IAS 27.11 states that “a parent which issuesconsolidated financial statements shouldconsolidate all subsidiaries”. IAS 27.06 definesa parent as “an enterprise that has one or moresubsidiaries”, a subsidiary as “ an enterprisethat is controlled by another enterprise (knownas the parent)”, and control as “the power togovern the financial and operating policies of anenterprise so as to obtain benefits from itsactivities”. Paragraph 35 of the Framework andIAS 1.20(b)(ii) (revised 1997) require thattransactions and other events are accounted forin accordance with their substance andeconomic reality, and not merely their legalform.

13. Control over another entity requires havingthe ability to direct or dominate its decision-making, regardless of whether this power isactually exercised. Under the definitions of IAS27.06, the ability to govern decision-makingalone, however, is not sufficient to establishcontrol. The ability to govern decision-makingmust be accompanied by the objective ofobtaining benefits from the entity’s activities.

14. SPEs frequently operate in a predeterminedway so that no enterprise has explicit decision-making authority over the SPE’s ongoingactivities after its formation (i.e., they operateon “autopilot”). Virtually all rights, obligations,and aspects of activities that could be controlledare predefined and limited by contractualprovisions specified or scheduled at inception.In these circumstances, control may exist forthe sponsoring party or others with a beneficialinterest, even though it may be particularlydifficult to assess, because virtually all

169cECB


ANNEX B



guidelines

activities are predetermined. However, thepredetermination of the activities of the SPEthrough an “autopilot” mechanism oftenprovides evidence that the ability to control hasbeen exercised by the party making thepredetermination for its own benefit at theformation of the SPE and is being perpetuated.

15. IAS 27.13(b) indicates that a subsidiaryshould be excluded from consolidation when it“operates under severe long-term restrictionswhich significantly impair its ability to transferfunds to the parent.” Predetermination of theactivities of an SPE by an enterprise (thesponsor or other party with a beneficialinterest) is often a demonstration of controlover ongoing activities as determined by thatenterprise and would not represent the type ofrestrictions referred to in IAS 27.13(b).

Date of Consensus: June 1998

Effective Date: This interpretation becomeseffective for annual financial periods beginningon or after 1 July 1999; earlier application isencouraged. Changes in accounting policiesshould be accounted for according to thetransition requirements of IAS 8.46.

APPENDIXThe purpose of the appendix is to illustrate theapplication of the Interpretation to assist inclarifying its meaning.

Indicators of control over an SPE

The examples in paragraph 10 of thisInterpretation are intended to indicate types ofcircumstances that should be considered inevaluating a particular arrangement in light ofthe substance-over-form principle. Theguidance provided in the Interpretation and inthis Appendix is not intended to be used as “acomprehensive checklist” of conditions thatmust be met cumulatively in order to requireconsolidation of an SPE.

(a) Activities

The activities of the SPE, in substance, arebeing conducted on behalf of the enterprise,which directly or indirectly created the SPEaccording to its specific business needs.

Examples are:

– the SPE is principally engaged in providing asource of long-term capital to an enterprise orfunding to support an enterprise’s ongoingmajor or central operations; or

– the SPE provides a supply of goods or servicesthat is consistent with an enterprise’s ongoingmajor or central operations which, without theexistence of the SPE, would have to beprovided by the enterprise itself.

Economic dependence of an entity on thereporting enterprise (such as relations ofsuppliers to a significant customer) does not, byitself, lead to control.

(b) Decision-making

The reporting enterprise, in substance, has thedecision-making powers to control or to obtaincontrol of the SPE or its assets, includingcertain decision-making powers coming intoexistence after the formation of the SPE. Suchdecision-making powers may have beendelegated by establishing an “autopilot”mechanism.

Examples are:

– power to unilaterally dissolve an SPE;– power to change the SPE’s charter or bylaws;

or– power to veto proposed changes of the SPE’s

charter or bylaws.

(c) Benefits

The reported enterprise, in substance, has rightsto obtain the majority of the benefits of theSPE’s activities through a statute, contract,


agreement, or trust deed, or any other scheme,arrangement or device. Such rights to benefitsin the SPE may be indicators of control whenthey are specified in favour of an enterprise thatis engaged in transactions with an SPE and thatenterprise stands to gain those benefits from thefinancial performance of the SPE.

Examples are:

– rights to a majority of any economic benefitsdistributed by an entity in the form of futurenet cash flows, earnings, net assets, or othereconomic benefits; or

– rights to majority residual interests inscheduled residual distributions or in aliquidation of the SPE.

(d) Risks

An indication of control may be obtained byevaluating the risks of each party engaging intransactions with an SPE. Frequently, thereporting enterprise guarantees a return orcredit protection directly or indirectly throughthe SPE to outside investors who providesubstantially all of the capital to the SPE. As aresult of the guarantee, the enterprise retainsresidual or ownership risks and the investorsare, in substance, only lenders because theirexposure to gain and losses is limited.

Examples are:

– the capital providers do not have a significantinterest in the underlying net assets of theSPE;

– the capital providers do not have rights to thefuture economic benefits of the SPE;

– the capital providers are not substantivelyexposed to the inherent risks of theunderlying net assets or operations of theSPE; or

– in substance, the capital providers receivemainly consideration equivalent to a lender’sreturn through a debt or equity interest.

171cECB


Importanceof offshorecountries

in US data

ANNEX C

From Table C-1, few geographic zonespublished by the BEA to obtain independentregional groups of countries and a “closed”regional distribution of US external FDIliabilities (i.e. the sum of individual countriesand the new defined groups matches the totalUS FDI liabilities) have been selected/modified.

In Table C-2, the presence of foreign countriesin the US economy, measured in term of FDIcapital invested, is evaluated according to theapplicatione of both first shot and UBO11

criterion (column 1 and 2 respectively). Foreach partner country, a relative change in FDIliabilities is calculated, resulting from the movetowards UBO criteria. Column 3 presents aranking of US partners according topercentages changes. Nine offshore centrecountries and one group of offshore countries(Other OWH12) have been selected from the USdistribution list, with available data on bothfirst shot and UBO criteria.

With US$ 55 bn (€59 bn13), offshore centresranked altogether at the 8th position (fifthposition if we merge EU countries), in the “firstshot” classification list. With US$ 66 bn (€71bn), they ranked all together at the 7th position(or fourth position behind EU, Japan andCanada), in the UBO distribution list.

ANNEX C : I MPORTANC E O F O F F SHORE COUNTR I E SI N U S DATA

11 US definition of UBO: “The ultimate beneficial owner is thatperson, proceeding up a US aff iliate’s ownership chain,beginning with and including the foreign parent, that is not ownedmore than 50% by another person.”.

12 Other OWH = Other Western Hemisphere = Anguilla, Antigua &Barbuda, Aruba, Barbados, Cuba, Dominica, Dominican Rep.,French Islands-Caribbean, Grenada, Haiti, Jamaica, St Lucia, StVincent & Grenadines, Trinidad&Tobago, UK Islands.

13 1€ = 0.9305 US$, at end 2000.


Partner country First shot UBO Partner country First shot UBO

World Total 1,214,254 1,214,254 Dominican Republic 79Canada 114,599 127,864 French Islands, Caribbean 3Europe 835,137 781,462 Grenada 5

Austria 3,174 2,439 Haiti -1Belgium 14,585 10,881 Jamaica -5Luxembourg 53,794 1,832 St. Lucia 0Finland 9,107 9,281 St. Vincent and the Grenadines 3France 131,484 130,346 Trinidad and Tobago 40Germany 124,839 135,693 UK Islands, Atlantic (OWH) 0Ireland 23,528 6,744 Africa 2,756 2,971Italy 5,994 13,401 South Africa 1,218 2,075Netherlands 146,493 121,635 Other Africa 1,538 896Spain 5,459 6,754 Liberia 1,549Greece 952 Middle East 6,189 11,722Portugal -68 Israel 2,690 2,773Sweden 22,427 23,791 Kuwait 908 1,155Denmark 4,428 3,014 Lebanon 1 635United Kingdom 213,820 252,397 Saudi Arabia 4,721Norway 2,241 2,627 United Arab Emirates 64 1,592Liechtenstein 202 268 Other Middle East 845Switzerland 69,240 59,108 BahrainIceland Iran 1Other Europe 1) 3,325 254 Jordan -3

South and Central America 13,682 17,649 Oman -11Brazil 886 1,616 Qatar 37Mexico 7,832 10,271 Syria 1Panama 3,726 410 Yemen -9Venezuela 802 4,042 Asia and Pacific 201,110 218,791Other S. & C. America 435 1,310 Australia 20,701 20,437Argentina 362 Hong Kong 1,544 12,825Chile 24 Japan 163,577 165,812Colombia 2 Korea, Republic of 3,287 3,401Uruguay 40 Malaysia 92 731

Other Western Hemisphere (OWH) 40,782 44,282 New Zealand 385 430Bahamas 1,268 78 Philippines 50 103Bermuda 18,502 38,378 Singapore 7,751 7,846Netherlands Antilles 3,940 1,193 Taiwan 3,131 5,250United Kingdom Islands, Caribbean 15,353 4,591 Other Asia and Pacific 593 1,956Other OWH 1,718 42 China 296

Anguilla 1 India 96Antigua and Barbuda 20 Indonesia 39Aruba 14 Thailand 116Barbados 1,560 Back to US 0 9,512Cuba 0 Offshore centres 1) 55,353 65,998Dominica 0

Table C-1 Foreign Direct investment posit ions in the United States, end

(US$ million)

Source: Bureau of Economic Analysis – US1) Eurostat estimates.

However, if we analyse individually the tenidentified offshore countries, the role played byeach of them is drastically affected by the UBOcountry reallocation:

– This reallocation lead to a doubling of FDIassets held by Bermuda in the USA, fromUS$19 to US$38 bn, boosting this country

within the top 8 major partners. The situationof Hong Kong is even more spectacular, withan increase of its positions in the USA bymore than 700%. Lebanon, who has almost nopresence on the US market, appears as non-negligible in the “UBO” distribution’s list.

– Apart from Singapore, UBO reallocation has

173cECB


Importanceof offshorecountries

in US data

ANNEX C

Percentage changeFirst shot UBO (from First shot to UBO)

1 United Kingdom United Kingdom Back to US (by definition) nc2 Japan Japan Lebanon 634003 Netherlands Germany United Arab Emirates 23884 France France Hong Kong 7315 Germany Canada Malaysia 6956 Canada Netherlands Venezuela 4047 Switzerland Switzerland Other Asia and Pacific 2308 Luxembourg Bermuda Other S. & C. America 2) 2019 Ireland Sweden Italy 124

10 Sweden Australia O. Middle East & Saoudi Arabia 1) 12011 Australia Italy Bermuda 10712 Bermuda Hong Kong Philippines 10613 UK Islands, Caribbean Belgium Brazil 8214 Belgium Mexico South Africa 7015 Finland Back to US Taiwan 6816 Mexico Finland Liechtenstein 3317 Singapore Singapore Mexico 3118 Italy Spain Kuwait 2719 Spain Ireland Spain 2420 Denmark O. Middle East & Saoudi Arabia 1) United Kingdom 1821 Netherlands Antilles Taiwan Norway 1722 Panama UK Islands, Caribbean New Zealand 1223 Other Europe 1) Venezuela Canada 1224 Korea, Republic of Korea, Republic of Germany 925 Austria Denmark Sweden 626 Taiwan Israel Korea, Republic of 327 Israel Norway Israel 328 O. Middle East & Saoudi Arabia 1) Austria Finland 229 Norway South Africa Japan 130 Other OWH 3) Other Asia and Pacific Singapore 131 Hong Kong Luxembourg France -132 Other Africa 4) Brazil Australia -133 Bahamas United Arab Emirates Switzerland -1534 South Africa Other S. & C. America 2) Netherlands -1735 Greece Netherlands Antilles Austria -2336 Kuwait Kuwait Belgium -2537 Brazil Greece Denmark -3238 Venezuela Other Africa 4) Other Africa 4) -4239 Other Asia and Pacific Malaysia Netherlands Antilles -7040 Other S. & C. America 2) Lebanon UK Islands, Caribbean -7041 New Zealand New Zealand Ireland -7142 Liechtenstein Panama Panama -8943 Malaysia Liechtenstein Other Europe 1) -9244 United Arab Emirates Other Europe 1) Bahamas -9445 Philippines Philippines Luxembourg -9746 Lebanon Bahamas Other OWH 3) -9847 Back to US Other OWH 3) Greece nc48 Portugal Portugal Portugal nc

Table C-2: Classi f ication of US FDI partners, according to f irst shot, UBO and %

1) Eurostat estimates.2) Mainly Argentina.3) Mainly Caribbean offshore centres of Eurostat list.4) Mainly Liberia (also member of the offshore list).

sharply reduced the importance of the sixother offshore countries. As an example, theBahamas and “Other OWH” islands havealmost disappeared from the UBO list (less

than US$ 80 Mio). Offshore countries aremore “sensitive” to criteria change. In column3, offshore countries can be found either inthe top or in the bottom countries’ list


(nothing in between, except Singapore). Atthe bottom, they all have a percentage changebelow “-70%”, like Ireland, Luxembourg and“Other European” country group.

The ranking of Hong Kong at the top UBO listmakes sense, since Hong Kong enterprisesmight also be “real” investors/investedregarding the economic position of Hong Kong.The ranking of Bermuda is more questionable.

175cECB


ANNEX D

Data on theimportance

of SPEs in EUMember States

ANNEX D : D ATA ON TH E IMPOR TANC E O F S P E S I NEU MEMBER S TAT E S

Table D 1 – Impact of SPEs in FDI statist ics

(EUR millions)


Euro area European Union National


FDI abroad 31,229 81,199 54,239 58,189 112,428Equity capital 19,758 50,126 40,927 28,957 69,884of which SPEs or assimilated cies 961 2,483 1,129 2,315 3,444Other capital 11,471 31,073 13,312 29,232 42,544of which SPEs or assimilated cies 34,031 -18,966 24,839 -9,775 15,065

FDI in the reporting economy 33,835 60,820 53,929 40,726 94,655Equity capital 39,329 48,269 58,791 28,807 87,598of which SPEs 5,759 2,478 7,214 1,023 8,237Other capital -5,494 12,551 -4,862 11,919 7,057of which SPEs or assimilated cies 1,722 4,486 1,498 4,709 6,208

Net 65,064 142,019 108,168 98,915 207,083




FDI abroad 115,233 49,288 128,977 35,545 164,521Equity capital 57,984 15,891 60,984 12,892 73,875of which SPEs or assimilated cies 0 0 0 0 0Other capital 57,249 33,397 67,993 22,653 90,646of which SPEs or assimilated cies 52,862 30,918 63,258 20,522 83,780

FDI in the reporting economy 134,251 75,419 160,016 49,655 209,670Equity capital 102,233 55,328 124,324 33,238 157,561of which SPEs or assimilated cies 24,271 17,909 25,358 16,822 42,180Other capital 32,018 20,091 35,692 16,417 52,109of which SPEs or assimilated cies 14,251 17,293 16,665 14,879 31,544

Net -19,018 -26,131 -31,039 -14,110 -45,149

COUNTRY: BELGIUM

TF-FDI – SUBGROUP ON SPES

Remarks on the SPE data of Belgium

Coordination Centers in Belgium are included under the category SPE’s. The same list of enterprises is used to preparethe figure in both tables but there are some other differences mentioned below.

The BOP data in the first table contains the figures of Belgium and Luxemburg together. Until end 2001, Belgiumcollected and published figures for the BLEU (Belgian-Luxembourg Economic Union). Only since 2002 there arefigures of Belgium seperately available.

In the second table, stock data of the annual survey are used and not the figures of IIP because that are stocks calculatedon accumulated flows. The data of the survey are only covering Belgium.

In other capital, in the first table short term payments are included whereas in the second table only the payments > 14days are included in the figures.


COUNTRY: FINLAND



(EUR millions)




FDI abroad -3,345 -6,009 -7,745 -1,609 -9,354Equity capital -259 -3,239 -1,656 -1,842 -3,498of which SPEs or assimilated cies NA NA NA NA NAOther capital -3,086 -2,770 -6,089 233 -5,856of which SPEs or assimilated cies NA NA NA NA NA

FDI in the reporting economy 144 4,026 4,556 -386 4,170Equity capital 258 4,056 4,036 278 4,314of which SPEs NA NA NA NA NAOther capital -114 -30 520 -664 -144of which SPEs or assimilated cies NA NA NA NA NA

Net -3,201 -1,983 -3,189 -1,995 -5,184




FDI abroad 20,271 35,730 36,013 19,988 56,001Equity capital 16,923 22,245 30,381 8,787 39,168of which SPEs or assimilated cies NA NA NA NA NAOther capital 3,348 13,485 5,632 11,201 16,833of which SPEs or assimilated cies NA NA NA NA NA

FDI in the reporting economy 6,438 19,647 22,842 3,243 26,085Equity capital 4,036 14,917 17,018 1,935 18,953of which SPEs or assimilated cies NA NA NA NA NAOther capital 2,402 4,730 5,824 1,308 7,132of which SPEs or assimilated cies NA NA NA NA NA

Net 13,833 16,083 13,171 16,745 29,916

In the case of Finland it was not possible to identify SPEs among the resident direct investors/directinvestment enterprises and thus no separate FDI data related to resident SPEs, can be ??ovided.

In the attached table, the equity figures include reinvested earnings.

177cECB


ANNEX D



COUNTRY: FRANCE



(EUR millions)




FDI abroad -47,572 -44,973 -59,336 -33,209 -92,545Equity capital -31,175 -20,990 -34,585 -17,580 -52,165of which SPEs or assimilated cies -11,366 -12,665 -14,545 -9,486 -24,031Other capital -12,125 -22,005 -20,203 -13,927 -34,130of which SPEs or assimilated cies -2,731 -7,110 -5,520 -4,321 -9,841

FDI in the reporting economy 36,501 22,305 51,278 7,528 58,806Equity capital 13,551 6,653 15,548 4,656 20,204of which SPEs 5,069 2,851 7,374 546 7,920Other capital 22,218 13,276 34,290 1,204 35,494of which SPEs or assimilated cies 6,877 3,248 9,874 251 10,125

Net -11,071 -22,668 -8,058 -25,681 -33,739




FDI abroad 169,080 295,897 232,162 232,815 464,977Equity capital 122,810 216,980 166,058 173,732 339,790of which SPEs or assimilated cies 20,466 39,136 29,098 30,504 59,602Other capital 46,271 78,917 66,105 59,083 125,188of which SPEs or assimilated cies 12,514 21,342 17,878 15,978 33,856

FDI in the reporting economy 150,394 126,668 197,309 79,753 277,062Equity capital 102,568 90,565 135,751 57,382 193,133of which SPEs or assimilated cies 35,964 32,797 47,691 21,070 68,761Other capital 47,826 36,103 61,559 22,370 83,929of which SPEs or assimilated cies 11,997 9,057 15,442 5,612 21,054

Net 18,686 169,229 34,853 153,062 187,915

Since SPEs are not identified as such in French statistics (no official definition), a category has been created forthe purpose of this assessment including all companies with NACE 74.15 (“Management of companies”).


COUNTRY: GERMANY


Table D 4 – Impact on SPEs in FDI statist ics

(EUR millions)




FDI abroad(resident SPEs = Investors) 28,507 -73,765 15,273 -60,531 -45,258

Equity capital -5,175 -56,443 -16,246 -45,372 -61,618of which SPEs or assimilated cies 7,058 -8,769 -2,123 412 -1,711Other capital 33,682 -17,322 31,519 -15,159 16,360of which SPEs or assimilated cies 12,006 321 14,477 -2,150 12,327

FDI in the reporting economy 29,996 11,650 30,972 10,674 41,646(resident SPEs = Investees)

Equity capital 18,325 11,764 27,620 2,469 30,089of which SPEs 17,692 212 22,954 -5,050 17,904Other capital 11,671 -114 3,352 8,205 11,557of which SPEs or assimilated cies 7,585 5,179 2,683 10,081 12,764

Net 58,503 -62,115 46,245 -49,857 -3,612




FDI abroad(resident SPEs=Investors) 160,003 299,592 220,494 239,101 459,595Equity capital 159,845 243,034 208,918 193,961 402,879of which SPEs or assimilated cies 86,737 115,651 120,086 82,302 202,388Other capital 158 56,558 11,576 45,140 56,716of which SPEs or assimilated cies 23 14,114 3,988 10,149 14,137

FDI in the reporting economy 301,476 167,953 347,208 122,221 469,429(resident SPEs=Investees)Equity capital 167,944 77,061 193,916 51,089 245,005of which SPEs or assimilated cies 140,493 54,333 163,447 31,379 194,826Other capital 133,532 90,892 153,292 71,132 224,424of which SPEs or assimilated cies 107,845 61,336 121,861 47,320 169,181

Net -141,473 131,639 -126,714 116,880 -9,834

APPROXIMATION OF SPE ACTIVITIES IN GERMANFDI STATISTICSIn accordance with the international standards,activities of SPEs established in Germany bynon-residents are generally included in thedirect investment data. Contrary to theinternational standards, in the case of SPEsestablished in Germany by non-residents thathave the sole purpose of financial

intermediation, transactions with affiliatedbanks and affiliated financial intermediaries,except transactions in equity capital andpermanent debt, are not excluded from the data.Furthermore, there is no special definition,identification and treatment of SPEs in GermanFDI statistics.

179cECB


ANNEX D



However, a sector breakdown allows theidentification of “holding companies” (NACEcode 6523), management holding companies(NACE code 7415) and “other services forenterprises”. These sectors are regarded as abest approximation for domestic SPEs.Certainly not all companies within these sectorscan be regarded as SPEs, but more specificinformation are not available for the time being.Financial and management holding companiescover about 50% of German direct investmentassets abroad and more than 75% of Germandirect investment liabilities held abroad. Thisillustrates the dominant role holding and shellcompanies are playing in conducting FDIactivities. A case by case study showed thatabout 60% of German holding companiesowned by foreign direct investors haveparticipations mainly in domestic enterprises,whereas about 40% hold exclusively foreignsubsidiaries. In the case of German directinvestment abroad about 39% of Germanholding companies do also hold Germansubsidiaries. However, the increasingcomplexity of transactions related to financialvehicles challenges the compilers’ possibilitiesin analysing the whole extent of SPE activities.

The sector “other services for enterprises”comprises among other business assetmanagement and fiscal advice. SPEs involved infinancial leasing activities are not classified in“other services for enterprises” but in variousother sectors related to the business in whichthey are actively mitigating financial leases.Financial auxiliaries dealing with asset backedsecurities are not at all included in FDIstatistics.


COUNTRY: GREECE



(EUR millions)




FDI abroad 0 0 0 0 0Equity capitalof which SPEs or assimilated ciesOther capitalof which SPEs or assimilated cies

FDI in the reporting economy 0 0 0 0 0Equity capitalof which SPEsOther capitalof which SPEs or assimilated cies

Net 0 0 0 0 0




FDI abroad 635 5,654 1,750 4,539 6,289Equity capital 565 5,458 1,669 4,354 6,023of which SPEs or assimilated cies 337 2,814 1,046 2,105 3,151Other capital 70 196 81 185 266of which SPEs or assimilated cies 22 25 24 23 47

FDI in the reporting economy 8,977 4,435 10,178 3,234 13,412Equity capital 8,168 4,176 9,312 3,032 12,344of which SPEs or assimilated cies 383 478 383 478 861Other capital 809 259 866 202 1,068of which SPEs or assimilated cies 0 0 0 0 0

Net -8,342 1,219 -8,428 1,305 -7,123

The data refer only to Holding Companies asthey were reported by the respondents to theannual questionnaire on FDI positions data.

181cECB


ANNEX E

SPEs in theEurostat

classificationby sector

(NACE rev1)

Because of their involvement in variouseconomic activities of services sector, SPEscannot be identified along the Eurostatclassification by sector. Furthermore, not all theMember States are able to identify SPEs’ intheir surveyed population of enterprises.

The OECD benchmark and the IMF fifthManual (§365 and 372) give some examples of

ANNEX E : S P E S I N T H E E URO S TATC L A S S I F I C AT I ON B Y S E C TOR ( N A C E R E V 1 )

SPEs defined according to Nace rev1 code Included in Eurostat code

The company’s structure:Holding companies 74.15 7415 -> 7410 ->7400 -> 7395Regional headquarters 74.15 7415 -> 7410 ->7400 -> 7395

The purpose of the activityManagement of foreign exchange risk 67.11 or 67.13 6890 -> 6895Facilitation of financing of investment 65.23 6520 -> 6895Dealing for own account by security dealers 65.23 6520 -> 6895Security dealing on behalf of the others 67.12 6890 -> 6895

possible SPEs, defined either according to theirstructure or the purpose of their activity. Fromthese examples, Eurostat has tried to identifythe localisation of these activities in the officialNACE rev 1 nomenclature, either in thefinancial intermediation (6895) or the “Realestate and business activity” (7395) sectors.

Eurostat does not process estimates at classlevel of the NACE (65.23, 67.11, 67.12, 67.13,74.15), as few Member States are able toprovide such details, at the moment.

Eurostat makes regular estimates for NACEgroup 65.2 (Eurostat code 6520), NACE

Assets (2000) Liabilities (2000)

Total (All sectors) 1,517,186 890,709

Total services 792,494 547,207

Financial intermediation (6895) 295,180 189,831Monetary intermediation (6510) 97,077 45,417Other financial intermediation (6520) 2) 102,347 110,283

Financial holding companies (6524) : :Insurance & activities auxiliary to insurance (6730) 83,348 30,679Miscellaneous Financial intermediation (6890) 2) 12,408 3,452Real estate & business activities (7395) 273,718 218,793Real estate (7000) 23,396 20,612Computer activities (7200) 18,857 7,714Research and development (7300) 1,385 1,453Other business activities (7400) 2) 217,057 182,162

Management holding companies (7415) : :Miscellaneous Real estate & business activities 13,012 6,849

Total “6520, 6890 and 7400” 2), in % of Total services 42 54

Total “6520, 6890 and 7400” 2), in % of All sectors 22 33

1) Excluding Dutch SPEs.2) Sectors “potentially” hosting SPEs.

Table E-1 Importance of some “potential ly” hosting SPEs sectors, in Extra EU FDI positionsdata1)

(EUR millions)

division 74 (Eurostat code 7400), and for themiscellaneous activities of the financialintermediation sector (Eurostat code 6890), asindicated below.


On the basis of EU available data, the relativeimportance of assumed “SPEs sectors” isappreciated relatively to the services sector andthe total sectors.

The analysis of Table E-1 indicates that:

1. 42% of EU FDI assets held by the servicessector could be under the effective control ofresident SPEs, that is E332 bn or 22% oftotal EU external FDI assets.

2. 54% of FDI assets invested by the rest ofthe world in resident enterprises of theservices sector could have been, in fact,addressed to resident SPEs: around E296 bnor one third of Extra EU liabilities.

Apart from the Extra EU zone, EU FDIestimates by activity are processed by Eurostatwith partners “total EFTA”, the USA, Canadaand Japan. Therefore, an rough idea about thepotentiality of the link between the main hostingSPEs sectors” and offshore countries could begiven in the narrowed geographic space “ExtraEU – EFTA – USA – Canada – Japan”.

The analysis of Table E-2 shows that:

1. At end 2000, resident SPEs might have heldabroad up to 15% (€80 bn) of EU external

Assets (2000) Liabilities (2000)

Total (All sectors) 525,805 107,528

Total services 275,987 75,621

Financial intermediation (6895) 106,994 29,295Other financial & misc. intermediation (6520+6890)2) 39,170 13,882

In % of Total Financial intermediation 37 47In % of Total services 14 18

Real estate & business activities (7395) 58,913 20,434Other business activities (7400)2) 41,358 13,508

In % of Total Real estate & business act. 70 66In % of Total services 15 18

Total “6520, 6890 and 7400”2), in % of Total services 2 9 3 6

Total “6520, 6890 and 7400”2), in % of All sectors 1 5 2 5

Table E-2 Importance of some “potential ly” hosting SPEs1) sectors, in the restricted space“Extra EU – EFTA c. – USA – Canada – Japan”(EUR millions)

1) Excluding Dutch SPEs.2) Sectors “potentially” hosting SPEs.

FDI assets located in the “Extra EU-EFTA-USA-CA-JP” area. This amount is of thesame magnitude than the total EU FDI assetslocated in offshore centres (€77 bn, section3.1-table 3 of the report).

2. 25% of FDI capital (€28 bn) held in theEU by foreign companies located in the“Extra EU-EFTA-USA-CA-JP” zone, wereprobably routed to EU based SPEs. A largepart of these €28 bn could have beengenerated in the offshore countries area. Butif we compare this amount with the €66 bnhold by offshore countries in the EU market(see section3.1-table 3 of the report), it couldbe an indication that offshore centres mightnot have invested essentially in EU basedSPEs. In principle, there are no fiscal/taxincentives for establishing an EU based SPEin the chain between an offshore SPE and thetarget company in the EU. At the Europeanlevel, the existence (on a large scale) of adirect relationship between the so-called“EU based SPE” and the offshore countriescannot be established in the absence of FDIinformation with offshore countries brokendown by activity.

183cECB


ANNEX E

SPEs in theEurostat

classificationby sector

(NACE rev1)Activity code in

EurostatEconomic activity Questionnaires NACE Rev. 1

Agriculture and fishing 0595 sec A, B

Mining and quarrying 1495 sec C

Manufacturing 3995 sec D

Electricity, gas and water 4195 sec E

Construction 4500 sec F

Total services 5095 sec G, H, I, J, K, L, M, N, O, P, Q

Trade and repairs 5295 sec GHotels and restaurants 5500 sec HTransports, communication 6495 sec IFinancial intermediation 6895 sec JMonetary intermediation 6510 Group 65.1Other financial intermediation 6520 Group 65.2Financial holding companies 6524 part of class 65.23Insurance & activities auxiliary 6730 div 66 & group 67.2Total Other fin. intermed & insurance 6795Real estate & business act 7395 sec KReal estate 7000 div 70Computer activities 7200 div 72Research and development 7300 div 73Other business activities 7400 div 74Business & manag. Consultancy 7410 Group 74.1Manag. holding companies 7415 class 74.15Advertising 7440 Group 74.4Total Computer, Research & Other bus. 7495Misc. real estate & business activities 7390 div 71Other services 9995 sec L, M, N, O, P, QNot allocated economic activity 9996

Sub-total 9997 sec A,B,C,D,E,F,G,H,I,J,K,L,M,N,O,P,Q

Priv. purchases & sales of real estate 9998

Total 9999

Table E-3 Correspondence table between EUROSTAT classi f ication of services activit ies andthe NACE (summary)

Financial holding companies-Eurostat definition of class 65.23

This class includes other financial intermediation primarily concerned with distributing funds other than making loans:

– Investment in securities, e.g. shares, bonds, bills, unit trust units, etc.

– Dealing for own account by security dealers

– Investment in property where this is carried out primarily for other financial intermediaries (e.g. property unit trusts)

– Writing of swaps, options and other hedging arrangements

This class excludes:

– Financial leasing

– Security dealing on behalf of others

– Trade, leasing and renting of property

– Operational leasing

Management activities of holding companies-Eurostat definition of class 74.15

No additional description is made for this class.


F.1 TREATMENT OF SPES AND OFF-SHOREENTERPRISES IN DK

BACKGROUNDThe number and importance of offshoreenterprises and specifically of SPEs establishedin Denmark have grown rapidly during the lastyears. Especially, the importance of SPEs hasgrown rapidly since 1999, where the Danish taxlegislation was changed in favour of this type ofenterprises.

SPEs influence statistical data very drasticallyon a gross basis, especially data for in- andoutgoing transfers of equity capital and othercapital, and for dividends, reinvested earningsand interest income, even though they do notcontribute to the local economy. On a net basisthe effect on the current account is insignificant.Most of these enterprises are established asholding companies for tax reasons or legislationreasons. There is not legal obligation to registeras an SPE, and because the transactions of thecompanies are typically without “real”payments it is very difficult to identify thecompanies for statistical purposes.

Definition and identification of SPEs andoffshore enterprises

In Denmark we define an SPE as a holdingcompany, which has no interact with the localeconomy and hardly any employees. In manycases, these enterprises do not have anyemployees at all, but are managed by lawyers orauditors. The company is not establishedfor a kind of physical activity in Denmark.The inward FDI transactions and positionscorrespond to the outward FDI transactionsand positions and are typically madesimultaneously. The compound of the FDI-elements may differ in the inward and outwardsite, but the total assets and liabilities in eachindividual enterprise do not. We have so faridentified about fifty enterprises in the statisticsas SPEs according to our own definition.

When it comes to offshore enterprises thingsare more complicated. We have no special

ANNEX F : S P E C I A L T R E ATMENT S O F S P E S I NMEMBER S TAT E S

definition on offshore enterprises. One of thereasons being that the offshore enterprises aredifficult to identify. Our temporary definitionon offshore enterprises is that it is not an SPE,but the enterprise does not contribute to thelocal economy. The enterprise may beestablished as an inter-company bank for agroup of companies. None of these enterprisesare situated in Denmark and the group ofcompanies have no logical relation to Denmark.

We have only identified one enterprise as anoffshore enterprise.

Treatment of SPEs and offshore enterprises inDK

Flow – stock

We identify the SPEs in the statistics accordingto the mentioned above criteria. In one or twocases the enterprises have fulfilled criteriaexcept that the inward FDI and the outwardFDI were not balanced. In this case we havemade an individual judgement of thetransactions as a whole and have decided tobalance the figures, so the transactions/positions do not change the net position. Weinclude the SPEs in our data for both stock andflow.

Profit/loss and reinvested earnings

When it comes to calculation of profit/loss, welook at interest of inter-company loans, profit/loss, and dividend as a whole. Also in this casethe assets and the liabilities must balance.

Offshore companies are treated individually.

Future treatment of SPEs and offshoreenterprises in DK

It is still under consideration whether to includeor the exclude the SPEs from our FDI-data. Onthe one hand, the international manuals (and theinternational reporting requirements) defineSPEs as an ordinary FDI transaction on a parwith other FDI transactions, to be included in

185c ECB


ANNEX F

Specialtreatmentsof SPEs in

Member States

the statistics. On the other hand, the SPEs havea significant influence on the national data, evenif the SPEs do not contribute to the localeconomy. SPEs disturbs the analytical use ofdata and for this reason alone it could beconsidered to exclude them from the statistics.

One solution considered is to compile FDI fromthe SPEs and the offshore enterprises and createtwo separate datasets, one including the SPEsused for international reporting and oneexcluding the SPEs used for national reporting.

No decision has been taken yet.

F.2 BRIEF DESCRIPTION OF THE DUBLININTERNATIONAL FINANCIAL SERVICESCENTRE IN THE CONTEXT OF BOP/IIPSTATISTICS FOR IRELAND

The purpose of this note is to clarify the role ofthe International Financial Services Centre(IFSC) in Dublin and its significance in theBOP statistics for Ireland, in particular, as theyrelate to the CPIS and to the BOP concept ofresidency.

At the meeting of CPIS to discuss the 1997results some countries seemed to misunderstandthe nature of activity at the IFSC considering itto be primarily Direct Investment, this note willshow the range of activities at the IFSC whichinclude Direct Investment, Portfolioinvestment, Other Investment activities andFinancial Derivatives.

The note is structured as follows, it begins witha general overview of the IFSC, which isfollowed by a description of the variousactivities at IFSC and then concludes with aprofile of BOP financial account positions andtransactions that are relevant to each activity.

OVERVIEW

The IFSC (International Financial ServicesCentre, Dublin) was set up in 1987 offering a

range of tax and other incentives approved bythe European Union. In constructing aregulatory regime for financial services entities,the Irish Government adopted an innovativestance, both in devising new legislationdesigned to facilitate operators in internationalfinancial services and implementing EU singlemarket Directives.

The establishment of the IFSC was against abackdrop of high unemployment, massemigration and a very difficult economicsituation. Its objective was to help remedy thissituation by providing sustainable employmentin an active and substantive internationalfinancial services industry for well the educatedand highly motivated workforce which wasreadily available in this country at that time. Thelegal and regulatory environment underpinningthe IFSC, while being rigorous, in line with EUand International standards, is regarded as‘user-friendly’ by the operations setting up.

The IFSC is regarded as an example of an‘onshore’ financial services centre offering theopportunity to the global financial servicesindustry to conduct and expand its internationalbusiness. Apart from insurance and certainother financial services, many of the operationsset up within IFSC and their activities aresupervised by the Central Bank of Ireland. TheDepartment of Enterprise and Employmentexercises the supervisory role in respect ofinsurance operations. The IFSC is subject to allEU legislation applicable to financial services.A central attraction for companies locating inthe IFSC has been the reduced 10 per centcorporation tax rate that applies to suchcompanies.

However, in July 1998 the Irish Governmentannounced that it had reached agreement withthe European Commission on the phasing out ofthe existing preferential 10 per cent corporatetax rate regime for certain sectors of theeconomy, including the IFSC and itsreplacement with a general across-the-board12.5 per cent single corporation tax rate. Thenew framework will see the application of a new


corporation tax rate of 12.5 per cent on tradingincome from January 1st 2003 to financialservices activities which previously would havebeen established in the Dublin IFSC. Thiscountry-wide rate will apply to trading incomeof all corporations in Ireland, no matter whattheir business activity or where they arelocated.

LICENSING

The IFSC licensing process starts with contactbetween the prospective Applicant Companyand the IDA (Industrial DevelopmentAuthority) who market the Centre on behalf ofthe Government and who facilitate theapplication process. All license applications arechannelled through the IDA and then reviewedby the Certification Advisory Committee(CAC) which includes interests representingthe official regulatory bodies i.e. Departmentsof Finance and Trade, Enterprise &Employment, the Central Bank and the IDA.The licence lists the activities the company canundertake and specifies the employment criteriato be met to qualify for the incentives offered.As regards employment, this can be directlywithin the company licensed or designated tothe Management Company servicing thelicensed company or collective investmentundertaking.

Subsequent ongoing supervision is carried outby the Central Bank, which is responsible forthe prudential and other regulatory functions ofIFSC operations under its remit. TheDepartment of Trade, Enterprise & Employmentsupervise insurance operations. Corporatetreasury and asset financing activities byentities that are not subsidiaries of overseasbanks are monitored by the Central Bank.

IFSC ACTIVITIES

The activities undertaken in the IFSC whichconsist principally of Mutual Funds andBanking activities are described below in moredetail under the relevant headings shown. Inthis context an implicit activity within IFSC is

the establishment by ‘management’ companiesof various types of special purpose companiesor SPCs through which investments are made orservices provided (e.g. captive insurance orreinsurance companies, agency reinsurancecompanies, captive finance companies, agencytreasury companies).

Initially all activity of the IFSC entities waswith non-residents, thus all transactions wereBOP relevant. However, since its inceptionthere has been a trend towards greaterintegration with the domestic economy whichwill be complete once changes in the taxationregime come into being on 1 January 2003.

Banking & Asset Finance – lending, deposits,corporate finance

This refers to conventional banking activitiesincluding lending and deposit taking and alsostructured finance. Corporate finance and otherfinancial engineering services are also includedunder this activity, either on or off balance sheetto the IFSC Company or its parent. Assetfinance refers to the financing of operationssecured on particular assets. Aviation financeis an example, but other types of assets areincluded e.g. ships, computer hardware, railwaystock etc.

Collective investment schemes (CISs) andSpecial purpose investment companies (SPICs)

CISs cover activities such as unit trusts, mutualfunds, UCITS (units for collective investmentin transferable securities) entities. Investorsare non-resident and the bulk of the investmentsare into non-resident portfolio securities.SPICs are inward direct investment entitieswhose assets consist of non-resident issuedportfolio securities.

Insurance/re-insurance

Conventional insurance and re-insurancebusiness is covered here including the activitiesof captive companies.

187c ECB


ANNEX F


Member States

Corporate treasury management

This includes the corporate treasury activity ofbanks providing risk-management servicesdirectly to customers. Individual corporateshave established stand-alone and agencycompanies for activities such as internationalnetting, cash management etc. carried out onbehalf of affiliates.

Financial Services ProvidersThis principally covers administrative/trusteeand custodial management as well as servicingfor international fund managers and promotersof funds. Management in administrativeservices provided to insurance or treasurycaptive companies is included here.

Money markets/securities dealing/brokingThis refers to trading in equities, bonds, andother instruments either on own account or onbehalf of clients.

Other activitiesThere are a number of specialist activities,which are not categorised under a genericheading. Some examples are: factoring/invoicediscounting, bond/note issuance facilities,receivables securitisation, purchasing,managing and collecting trade accountreceivables, balance sheet structuring, grantingof financial rights in the entertainment industry(e.g. world-wide tours, concert performances,merchandising etc.).

BALANCE OF PAYMENTS TREATMENT OF IFSCENTITIES

All entities operating at IFSC are deemed to beResidents for Balance of Payments purposesapart from non-resident funds (registered inother jurisdictions both off shore and on shore).These resident entities are registered in Irelandand pay tax to the Irish Government on tradingincome. The funds are listed on the Irish StockExchange. Therefore, comprehensive reportingof Balance of Payments statistics is required (anexample of the data required for a collectiveinvestment scheme is attached).

A review of the relevant transactions of IFSCentities already discussed above with anindication of the appropriate BOP FinancialAccount statistics required is set out below:

Banking & Asset Finance – lending, deposits,corporate financeThese are direct investment entities with non-resident owners.

Financial Account Transactions/Positions:Most if not all of these companies are financialintermediaries or MFIs therefore onlytransactions in permanent debt and equity areconsidered as direct investment transactions.

They may also be involved in PortfolioInvestment – buying and selling of Bonds &Notes and Money Market Instruments.

Most activity will occur in the Other Investmentaccount with transactions in loans and deposits,and trade credits and other assets/liabilities.

There are also some transactions in allderivative types i.e. swaps, options, futures andforward contracts.

Insurance/re-insuranceIn general, these companies are inward DirectInvestment enterprises and some aresubsidiaries of Irish resident parents.

Financial Account Transactions/Positions:Equity investment into these companies isdirect investment.

However the bulk of the activity is theinvestment of policyholders funds into tradedsecurities, which is classified as PortfolioInvestment.

Changes in Technical /Actuarial reserves arerecorded under Other investment – otherliabilities


Corporate treasury managementIn general, these companies are inward DirectInvestment enterprises and some aresubsidiaries of Irish resident parents.

Financial Account Transactions/Positions:The majority of transactions in this sector areincluded in Direct Investment – Other capitalwhich captures inter-company loans/depositsand other transactions with related companies.

Where third parties are involved we recordtransactions in loans/deposits and other assets/liabilities under Other Investment. If Bonds orother traded securities are involved thesetransactions are recorded in PortfolioInvestment.

Collective Investment Schemes and SPICSCIS enterprises are primarily involved inportfolio investment both inwards andoutwards, while SPICs are inward DirectInvestment enterprises engaged in outwardPortfolio investment.

Financial Account Transactions/Positions:The main activity is in Portfolio investment,transactions by investors into a fund beingrecorded under Portfolio investment liabilities –equities

The investment activity of the fund is recordedunder portfolio investment assets – Equities,Bonds and MMIs.

Any non-resident transactions in cash arerecorded under Other Investment, and financialderivatives are recorded under Financialderivatives.

Financial Account EUR millions

Direct Investment net -34,947

Portfolio Investment Assets 120,859Portfolio Investment Liabilities 97,380

Other Investment Assets 99,886Other Investment Liabilities 73,863

Total net 14,555

Table F-1 UBO Summary statistics on IFSCactivit ies (end 1998)

189c ECB


ANNEX F


Member States

1) Custodial fees are payable to the Custodial Company.2) Administration fees are payable to the Administration Company.3) Management fees are payable to the Management Company.

F.3 SPECIAL PURPOSE ENTITIES – THE CASE OFTHE NETHERLANDS

1 BACKGROUNDAccording to the OECD Report on SpecialPurpose Entities and Offshore Enterprises{DAFFE/MC/STAT(2003)4}, the Netherlandsseems to be the only OECD country with a legaldefinition of Special Purpose Entities (SPEs)or, as they are called in the Netherlands, SpecialFinancial Institutions (SFIs). Conversely, inDutch statistical regulations the offshoreenterprises are not defined. This means that theconcept of offshore is unknown and that cross-border transactions of these enterprises are nottreated differently in the compiled statistics. AllSFIs established in the Netherlands areconsidered as residents.

From April 2003 the Statistical Information andReporting Department of DNB will implement aDirect Reporting system (DRA). In the newreporting system SFIs will have differentreporting obligations than non-SFIs. Non-SFIsare obligated to send in on a monthly basis, afully reconciled statement of stocks and flowsfor each BoP-item separately. SFIs only have toreport transactions on a monthly basis. A fullyreconciled statement including stocks isreported by the SFIs only on an annual basis.For the monthly and annual reportingobligations about 1,200 SFIs, with a coverageof about 95% of the figures of all SFIs, havebeen selected as direct reporters. In addition,DNB foresees an annual benchmark reporting,in which registered SFIs not being selected asdirect reporters, will report once a year their

SHARES REDEEMEDOutward portfolio investmentFinancial account 2.1.1.4.Liabilities Dr

SHARES ISSUEDInward portfolio investmentFinancial account 2.2.1.2.Liabilities Cr

DIVIDEND PAYMENTS TO UCITSInvestment Income, Portfolio InvestmentCurrent account 2.2.1/2.

INVESTMENTS BOUGHTOutward portfolio investmentFinancial account 2.1.1/2/3.Assets Dr

INVESTMENTS SOLDOutward portfolio investmentFinancial account 2.1.1./2/3.Assets Cr

AUDIT FEESNo BoP

DIVIDENDDISTRIBUTIONSInvestment IncomePortfolio InvestmentCurrent Account2.2.1 Cr

MANAGEMENTFEES 3)

30% No BoP70% Current AccountFinancial Services 1.6Cr

CUSTODIALFEES 1)

No BoPas custodiansresidents

ADMINISTRATIONFEES 2)

No BoPadministratorsresidents

OPERATINGEXPNSListing in FTCurrent AccountServices 1.9Cr

UCITS

Table F-2 Units for col lect ive investment in transferable securit ies (UCITS)


annual positions only. This benchmark willserve firstly as a basis for updating thepopulation of direct reporters and secondly as asource for levelling up the SFI-statisticscompiled on the basis of the figures reported bydirect reporters.

2 DEFINITION AND IDENTIFICATION

2.1 DEFINITION

Dutch SFIs are defined as:in the Netherlands established companies orinstitutions, regardless their legal form,whose shares are directly or indirectly heldby non-residents and are mainly dealingwith receiving funds from non-residents andchannelling them to non-residents.

2.2 IDENTIFICATION AND REGISTRATIONAccording to article 9 paragraph 1 of the“Reporting Instructions for External Payments2003”14, which are founded on the “ExternalFinancial Relations Act 1994”, SFIs areobliged to register at DNB within three weeksafter their establishment. On the basis of theinformation provided through the registrationform DNB will take a decision about whether ornot to appoint the SFI as a direct reporter.

There are three conditions that should be met byan entity to be considered an SFI: firstly, thecompany should be a resident; secondly, theshares should be directly or indirectly in handsof non-residents and, finally, the money flowsshould be mainly raised from non-residents andhanded over to non-residents.

Beside the legal obligation of SFIs to register atDNB, trust companies, as the representativesof over 75% of the total population of the SFIs,play an important role in identification andregistration of these entities. Maintenance of anup to date SFI-register is a necessary conditionto produce reliable statistics. In the case of SFIsthis is a challenging task, because a significantpart of these registered SFIs are not operatingregularly. In some cases they are established to

carry out just a limited number of transactionsin a short period of time. By the end of 2002over 12,500 SFIs were registered at DNB.

3 TREATMENT

3.1 INCLUSION IN STATISTICSWith regard to SFIs the Netherlands can beconsidered as a transit country for their moneyflows. The net outcome of the incoming andoutgoing flows should be close to zero. Thisshould be the case for each individual SFI aswell as for the total of the money flows of thewhole population of SFIs. In practice, however,a small net outcome may result for limitedperiods of time, mainly due to time differences.

The net effect of the SFI-transactions isincluded in the published data of the DutchBalance of Payments within Other Investment-Liabilities- Other Liabilities- Other Sectors .There are two reasons to include only the neteffect of these transactions in the BoP and toexclude them from the International InvestmentPosition (IIP) statistics. Firstly, the grossmoney flows are irrelevant for the Dutcheconomy; secondly, including the gross moneyflows would blow up the statistics, therebyhampering the analysis of the development ofthe external sector. For the same reasons SFI-transactions are also not included in theNational Accounts compiled by StatisticsNetherlands.

The exclusion of gross flows results in the so-called cleaning of the Dutch BoP from SFI-transactions. As other countries do include SFImoney flows from and to the Netherlands intheir BoPs as, respectively, inward and outwardDirect Investment (or Portfolio Investment), theregularly published data by DNB of bilateralinvestment flows from and to the Netherlandscan not be compared with the investmentstatistics compiled in other countries.

14 Rapportagevoorschriften Buitenlandse Betalingsverkeer(RV) 2003.

191c ECB


ANNEX F


Member States

For the compilation of the BoP for the Euro-area, DNB reports BoP data including grosstransactions of SFIs to the ECB. These BoP-data contain a geographical split between intra-EMU and extra-EMU countries. If the ECBwould not have the gross figures for theNetherlands (divided by intra-EMU and extra-EMU countries) in order to eliminate the intra-EMU transactions it would, ceteris paribus,result in errors & omissions for the BoP of theEuro-area. DNB also reports its IIP includingthe positions of SFIs to the ECB, therebyenabling the ECB to compile internationallyconsistent IIP statistics for the Euro-area.

3.2 TYPESDepending on their activities, four types ofSFIs can be distinguished. Considering themagnitude of their cross-border transactions thefinancing companies followed by holdingcompanies are the largest types of SFIs.

Financing companies

This type of SFIs are engaged in taking up andon-lending funds within their own groupcompany outside the Netherlands. Thesecompanies can also raise funds from non-residents to lend on within their own group orraise funds within the group to lend on outsidetheir own group to non-residents. The relativeshare of financing companies in total turnoverof SFIs in 2002 was about 71%.

Holding companies

These SFIs invest within the group company,manage the participations and distributedividends gained from the participations to theparent company. About 2% of total transactionsof SFIs in 2002 were realised by SFI holdingcompanies. The relative share of theseenterprises in the IIP-statistics in much higherthan in the gross transactions. In 2001 theparticipations of SFIs outside the Netherlandswas about 43% of all foreign assets of DutchSFIs.

Royalty and Film right companies

This group concerns a limited number of SFIswith a small share in the total turnover of SFIs.They exploit the licences, patents and filmrights for their parent companies. The measuredrelative share of these companies in 2002 cameto less than 1% of the total SFI transactions.

Rebilling companies

These SFIs mainly deal with rebilling activitiesof their parent company on behalf of the cross-border sales in goods. Rebilling companies areset up not only to gain from the tax regulationsfor these companies but also to minimiseexchange rate risks.

Beside these four types, different varieties ofSFIs can arise from a combination of two ormore of the above mentioned types.

3.3 PROBLEMSAlthough there are some practical difficulties inthe treatment of SFIs and their figures in theBoP, these problems can be solved. A shortdescription of the problems faced and thepossible solution for each difficulty ispresented below.

Non-responseIn spite of their obligations SFIs do notalways report their establishment andconsequently their activities to DNB. Thisproblem may well occur in the case of newlyestablished SFIs, which may not have beeninformed about their registration obligationsto DNB. There are some methods to identifythose unknown, i.e. newly established SFIs,and to point out to them their obligationstowards DNB. One method is the awarenessof trust companies about the registration andreporting obligations of their clients to DNB.Regular and intensive contacts with thesetrust companies has shown to be an effectiveinstrument in covering the SFI activities instatistics.

Some of the reporting delays of SFIs stemfrom the fact that their administration islocated outside the Netherlands. To deal with


the non-response and delay problem, DNBhas the disposal of the instrument of fines orceases and desist orders under penalty,which is based on the External FinancialRelations Act. Increasing the awareness ofSFIs about their reporting obligations, alsoby periodically sending reminders, combinedwith the legal instrument of fines and ceaseshas proven to be very successful in the datacollection process.

AsymmetryA large (short-term) difference between theincoming and outgoing money flows iscalled an asymmetry. As mentioned beforeall incoming money flows through SFIsshould be transferred to other countriesagain. In the majority of cases thetransactions will be carried out at the sametime or within a short period of time. Toenable SFIs to pay their expenditures in theNetherlands, however, the incoming moneyflows could be higher than the outgoingflows. Compared to the gross flows the totalexpenditures of SFIs or the net effect of themoney flows, however, is negligible. In thenew reporting system DRA the asymmetriescaused by each individual reporter can beeasier identified and solved.

Multistage SFIs

Some multinationals own more than one SFI,each used for different purposes. These SFIsare considered multistage SFIs if they arefinancially related to each other, for examplethrough loans or participations. In this casemoney flows may enter the Netherlands throughone of the related SFIs and leave the countrythrough another. Multistage SFIs can usually berecognised by their (almost) identical names,which usually include the name of themultinational parent company. The reports ofthe interrelated SFIs are managed by the sameaccount manager at DNB in order to be alert onthe interrelationship between the individualSFIs of a multistage SFI.

Change of activities of SFIs

A company can also perform a mixture of SFIand non-SFI activities. At the moment ofregistration a company not fully fulfilling thecriteria of an SFI will not be considered an SFI.It sometimes becomes more difficult, if SFIstake on board domestic activities. As soon asthis is recognised the status as an SFI will bereconsidered and, if necessary, withdrawn.

4 SUMMARY

Although the process of identifying SFIs andprocessing their data in some aspects is morecomplicated than that of non-SFIs, there arepractical solutions for the problems being facedwithin this process. This allows DNB tocompile reliable statistics about the activities ofSFIs. In the new reporting system, that is to beimplemented in April 2003, SFIs will continueto be treated separately from non-SFIs in thenational BoP.

FORE IGN D IRECT INVESTMENTTASK FORCE REPORT

MARCH 2004

FORE

IGN

DIR

ECT

INVE

STM

ENT

– TA

SK F

ORC

E RE

PORT

MAR

CH 2

004

Foreign direct investment task force report, March …...6 ECB Foreign direct investment – Task...

Documents

Transcript of Foreign direct investment task force report, March …...6 ECB Foreign direct investment – Task...