The euro bond market study, December 2004

THE EURO BOND MARKET STUDYDECEMBER 2004

In 2004 all ECB publications will feature

a motif taken from the

€100 banknote.

THE EURO BOND MARKET S TUDYDECEMBER 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 2004.

ISBN 92-9181-525-X (print)ISBN 92-9181-526-8 (online)

3c ECB

The euro bond marke t s tudyDecember 2004

CONTENT S1 INTRODUCTION 4

2 EXECUTIVE SUMMARY 5

3 MARKET OVERVIEW 8

3.1 Developments in theeuro-denominated governmentbond market 11

Box 1 Developments in euro arealong-term government bondyield spreads and fiscalpositions 14

Box 2 Recent developments ineuro area governments’debt duration 18

3.2 Developments in theeuro-denominated non-governmentbond market 22

Box 3 Recent developments incorporate bond spreads inthe euro area 26

4 MARKET INFRASTRUCTURE 30

5 MARKET TRENDS 38

5.1 Covered bond market 38

5.2 Securitisation market 41

5.3 Corporate and high-yield bondmarket 44

5.4 Credit derivatives market 47

5.5 Inflation-protected financial instruments 52

Box 4 Examples of inflationswaps 56

5.6 Real-time indices andexchange-traded funds 57

5.7 Rating agencies 63

5.8 Impact of the FSAP on the bond market and progress towardsintegration 67

Box 5 Degree of integration in theeuro area markets forgovernment bonds 70

ANNEXES

A Quantitative analysis of corporate bond spread dynamics 71

B Empirical evidence of pricediscovery in the CDS and bondmarkets 72

GLOSSARY 74

4ECB cThe euro bond marke t s tudyDecember 2004

1 I N TRODUCT I ONIn July 2001 the Market Operations Committee(MOC) of the European System of CentralBanks (ESCB) issued a report on the euro bondmarket. The study focused on major trends inthis market segment after the introduction of theeuro. Three years later, the MOC prepared anew report on the euro bond market that isintended to be an analytical enhancement of thefirst study. In addition to taking a moreanalytical approach, the report also covers allsignificant developments in the euro bondmarket in the period from 2001 to 2003. Theeuro bond market study 2004 draws on datafrom both public and private sources as well asqualitative information from selectedintermediaries and final investors. The reporthas also benefited from data support of theESCB Statistics Committee. Furthermore, aquestionnaire that was circulated among banksin the European System of Central Banks duringthe preliminary stages of drafting the reportprovided valuable information about specialcharacteristics of the euro area bond market.1

1 The study was prepared by staff members from DeutscheBundesbank (lead), ECB, Banque de France, and Banca d’Italia,and subsequently discussed by the ESCB Market OperationsCommittee.

5ECB


EXECUTIVE SUMMARY2 E X E CU T I V E S UMMARYThe euro bond market developed quite wellsince 2001. The growing importance of the euroas an international investment currency hasmade the market for euro-denominated issuesmore attractive for both investors and issuers.A key element behind these developments of theEuropean bond market in this period was theimpetus for a better integrated and more liquidmarket and the increasing diversity ofinnovative products, such as index-linkedbonds, real-time bond indices, fixed incomeexchange traded funds, credit derivatives andstructured products. The euro bond marketstudy includes most of these developments,including, as appropriate, in some empiricaltesting of special market features. The executivesummary is hence only a shortcut to finding thechapter that is of special individual interest.

The attractiveness of investments denominatedin euro is associated with greater competitionbetween market segments and different issuertypes in the euro area. In continental Europe,where financing structures are still stronglybank-based, government bonds and bank debtsecurities have dominated the market fordecades and still do. However, of particularimportance is the rapidly growing marketsegment of private issuers. At the beginning ofEuropean monetary union, corporate bonds hada share of only 9% in the stock of outstandingbonds. This share went up to 14% towards theend of 2003 as access was gained to a largerpotential pool of investors than existed beforethe introduction of the euro.

Improved access to financial markets withinthe EU allows investors to diversify theirportfolios and to invest more easily in marketsof countries other than their own. Since manyinvestors prefer assets denominated in localcurrency, the introduction of the euro hasreduced the home bias of euro area investorsand further promoted the diversification ofinvestments within the euro area. Furthermore,the development of a relatively broad andhomogenous financial market in the euro areaattracts international investors. Efforts toreduce information asymmetry and to improve

transparency (as enforced by the FinancialServices Action Plan) together with increasedliquidity and declining transaction costs furtherfoster the attractiveness of European bondmarkets for European and internationalinvestors.

In recent years the more intense competition,also due to the introduction of the euro, hasaccelerated the process of reshaping marketinfrastructure and has involved trading,clearing and settlement stages. The differentcomponents of the financial marketplaces havedeveloped new services and slimmer ownershipstructures. Strong synergies, the need to lowercosts and the drive to strengthen the position ofthe main management companies have spurredintegration between the trading circuits and thesettlement systems.

The euro covered bond market, an example foron-balance sheet securitisation, has witnessedinteresting developments over recent years.While the issuance of covered bonds declineduntil 2001, mainly due to the sharp reduction inissuance of German Pfandbriefe, a recoverystarted in 2001. However, apart from the risingvolumes since 2001 and continuous productinnovation, the interesting feature in this marketsegment is the growing share of issuance fromEuropean countries other than Germany, whosecovered bonds nonetheless still dominate themarket to a large extent. While issuance hasincreased in the existing covered bond markets,new markets have also developed or are aboutto be born. This is an outcome of themodernisation of existing covered bondslegislation in several countries, while othercountries have already adopted coveredbonds legislation or will soon do so. Thesedevelopments show the current dynamism in thecovered bond market in a pan-European context.

Off-balance-sheet securitisation has seenimpressive growth since the late 1990s and hasbecome an established asset class in theEuropean fixed income market. Total issuancevolumes rose to €268 billion in 2003 andexpectations are that securitisation issuance


will continue to grow and even outstripcorporate bond issuance in 2004. Differentkinds of securitisation in terms of asset classeshave been introduced, among them residentialmortgage-backed securities (RMBS) andcollateralised debt obligations (CDOs). Withrespect to the degree of development of thesecuritisation market there are still largedifferences between European countries whichcan be explained mainly by differing legal,regulatory, tax and accounting rules applicableto securitisation transactions. In some countriesthe legal and tax conditions for securitisationhave been improved recently. Whether the highgrowth rates seen over recent years can bemaintained, however, remains to be seen.

Considering the performance of the Europeanbond markets, spreads of corporate bond yieldsover government bond yields were atexceptionally low levels by the end of 2003after having peaked in the autumn 2001 andin 2002. Quantitative assessment of thisphenomenon suggests that much of corporatespread depression is due to historically lowinterest rate levels, encouraging investors tosearch for yield. In addition, spreads, taken aspremia for default risk, have been depressed bydeclining corporate leverage, a possibleindicator of companies’ solvency. Finally, theincreasing market liquidity associated with thematuring corporate bond market has squeezedliquidity premia. The current broadening anddeepening of the European corporate bondmarket is expected to continue in the future.This gives reason to believe that the dampeningimpact of lower liquidity premia on spreadmovements will continue.

Another segment of European credit marketswhich has expanded rapidly in recent yearsis the credit derivatives market. Creditderivatives, which allow the transfer of creditrisk to other sectors that lack direct originationcapabilities, are on the way to becoming one ofthe most successful financial innovations inrecent history. The remarkable development ofcredit derivatives markets especially in Europeand the ongoing integration of European credit

markets is contributing to the evolution ofliquid markets, thus facilitating the efficientpricing and trading of credit risks. Meanwhile,credit default swaps (CDS), which also providethe basis for more complex structuredinstruments, fulfil an important function insecondary credit markets with respect to pricediscovery. On 21 June 2004, the Iboxx/TRAC-X merger led to the launch of DJ iTraxx indices.This set of new rules-based and transparentindices is comprised of the most liquid names inthe European financial and corporate creditdefault swap (CDS) market. Discussions areongoing among market participants for listingfutures on DJ iTraxx indices and for havingthem traded on electronic trading platforms andcleared in a central clearing house. Should theseavenues or similar developments becomeconcrete and successful, it is likely that theywill enhance transparency and liquidity in theoverall credit markets, ultimately expanding thecorporate market, both in terms of instrumentsand market participants.

Inflation-linked bonds is a small but growingsegment of the euro bond market. Most of theEU national treasuries which have alreadyissued some inflation-linked bonds (UK,Sweden, France, Italy and Greece) are tendingto increase their issuance, whereas the GermanTreasury is expected to start to issue in 2005. Inparallel, the market for inflation linkedderivatives has picked up over the last threeyears, expanding the hedging and tradingopportunities of inflation risk.

One of the most recent innovations in theEuropean bond market was the development ofexchange-traded funds (ETFs), which allow adiversified portfolio to be bought or sold morecost-efficiently through one single transactionthan is currently possible with traditionalfunds. Another means by which innovationcould help would be through the development offutures contracts based on portfolios ofcorporate bonds, with delivery taking placethrough either cash or through ETFs. Aprerequisite for the development of the former(cash-settled futures contracts based on

7ECB


EXECUTIVE SUMMARY

corporate bond portfolios) is, however, thedevelopment of indices whose integrity isbeyond doubt and whose computation andpublication is effected in real time. It isinteresting to monitor the development of real-time indices and ETFs in the bond market andthe corporate bond segment in particularbecause it may allow two impediments to thedevelopment of this market to be overcome: theshortage of convenient hedging instruments andthe relatively high transaction costs associatedwith portfolios composed of many small issues.

Rating agencies have been playing a pivotal rolein the development of the euro bond market asproviders of independent credit assessments onbond issuers’ creditworthiness. One of thefactors underpinning this growth has been theincreasing coverage and use of credit ratingsprovided by rating agencies. However, owing tostill greater reliance on bank intermediation, thecoverage of credit ratings in Europe is stillunder-developed compared with the UnitedStates. Both general structural factors andspecific European drivers explain the role ofrating agencies in the European bond market.The advent of the euro and the integration ofEuropean financial markets conferred an evenmore determinant role to credit ratings. Byeliminating currency risk, the use of the euroallowed bond investors to focus on credit riskwhile the enlargement of their investmentuniverse increased their need for simpleindicators of this risk.

The Financial Services Action Plan (FSAP),adopted by the European Commission in 1999and endorsed by the Lisbon European Councilin March 2000, presents the most ambitiousinitiative to date to foster the integration ofcapital markets and to achieve a single marketfor financial services in the EU. The fourstrategic objectives underlying the FSAP relateto the single EU market for wholesale financialservices, open and secure retail markets, state-of-the-art prudential rules and supervision, andwider conditions for an optimal single financialmarket (namely tax and corporate governanceissues). With regard to the euro-denominated

bond market, a relatively high degree ofintegration can be observed. Nonetheless, in thecase of the euro area government bond market,additional integration may be possible.


This chapter reviews the trends in the primaryand secondary market in respect of the bondissuance by the public and the private sectors inthe EU. The general trends observed in theprevious bond market studies are still valid.2

Openness and competition between the euroarea issuers has increased. However,competition is likely to be given fresh impetusby new EU Member States.

The re-denomination of bonds from formernational currencies into euro at the beginning ofEMU paved the way for a European bondmarket. The trends already observed in theprevious studies still hold true. Bond marketconventions were harmonised among theparticipating Member States, which resulted ina relatively homogenous euro-denominatedbond market. That market has since becomemuch larger and more liquid than the nationalmarkets of the participating Member Stateswere in the pre-EMU era. The European bondmarket has also made distinct progress in termsof competition with markets of other developedcountries.

Although the size of the euro area securitiesmarket in terms of the outstanding volume isstill smaller than the market for domestic debt

3 MARKE T O V E RV I EWsecurities in the United States, this market has asustained rate of growth and is characterised bymore individual investment products. Domesticdebt securities in the developed countries havean overall outstanding volume of USD 24,495billion3 (at the end of 2003) and the share of theeuro area countries is now nearly on a par withJapan. Furthermore, the euro gained increasingimportance on the international bond market(outstanding volume USD 11,103 billion at theend of 2003). The growing relevance of the euroas an international investment currency hasmade the market in euro-denominated issuesmore and more attractive for both investors andissuers.

The relative attractiveness of the investmentsdenominated in euro is associated with morecompetition between market segments anddifferent issuer types in the euro area. Ofparticular importance is the rapidly growingmarket segment of private issuers. At thebeginning of EMU, corporate issuance had amarket share of only 9% of the outstanding of

2 See European Central Bank, “The euro bond market”, July 2001and Javier Santillán, Marc Bayle and Christian Thygesen, “Theimpact of the euro on money and bond markets”, OccasionalPaper No 1, European Central Bank, July 2000.

3 Bank for International Settlements, Quaterly Review, June 2004.

Chart 1 Domestic debt securit ies

(as a percentage of the outstanding volume in developed countries)

1999

euro area20.2

United States49.3

Japan22.0

UnitedKingdom

3.2

others5.3

2003

euro area22.3

UnitedStates47.4

Japan21.5

UnitedKingdom

3.4

others5.4

Source: BIS, Quarterly Review, June 2004.

9ECB


3 MARKETOVERVIEW

euro-denominated bonds. This share went up to14% in 2003 as a result of the access to a largerpotential pool of investors than existed beforethe introduction of the euro. The fundingrequirements associated with large mergersand acquisitions and the burst of the equitybubble have also played a significant role.Nevertheless, the outstanding amount ofeuro-denominated bonds issued by euro arearesidents is dominated by government issuancewith a share of 51%, albeit with a slightlydecreasing proportion. In relative terms, theoutstanding volume of monetary financialinstitutions (MFIs) has remained more or lessconstant, achieving a share of 35% at the end of2003 (see Table 1). Another trend, which is notsurprising, is that the share of euro arearesidents’ debt denominated in foreigncurrencies has fallen to negligible levels. This isalso due to the fact that debt in legacy currencies(i.e. formerly denominated in nationalcurrencies) was re-denominated in euro andthere is no longer issuance in those currencies.

The characteristics of European bonds arerelatively homogenous. Concerning the coupontype, the fixed rate segment still plays the mostimportant role. Regarding maturities, the wholespectrum is represented. Government bonds

provide nearly the complete range of maturities(from 1 to 30 years), while MFIs and especiallycorporates dominate in the short- and medium-term segments. The size of single issues variesconsiderably. Due to the refinancing needsarising from the high degree of indebtedness,governments issue rather large bonds, whilecorporates generally sell smaller bonds. Eventhough the European bond market is relativelyhomogenous, many niches have emerged. Someissuers have specialised in floating rate notes,index-linked bonds or the covered marketsegment. Furthermore, structured finance andderivative instruments have gained inimportance, as is shown by the growingsegments for asset backed securities, creditderivatives or exchange traded funds. Thistrend is tailored to the needs both of issuerspursuing cost-efficient debt management and ofinvestors optimising their risk management.

Easier access to foreign financial marketswithin the EU helps investors to diversify theirportfolios and to invest in European marketsabroad. Since many investors prefer assetsdenominated in the local currency, theintroduction of the euro has reduced the homebias of euro area investors and further promotedthe diversification of investments within the

Chart 2 Currency of international bonds and notes

(as a percentage of outstanding volume)

Euro29.1

Dollar47.2

Yen10.0

Pound7.8

others6.0

1999

Euro43.5

Dollar40.5

Yen4.4

Pound7.0

others4.6

2003

Source: BIS, Quarterly Review, June 2004.


euro area. Furthermore, the development of arelatively broad and homogenous financialmarket in the euro area is attracting internationalinvestors. Efforts to reduce informationasymmetry and to improve transparency (asenforced by the FSAP) together with increasedliquidity and declining transactions costs fosterthe attractiveness of European bond marketsfor European and international investors. Aquantitative assessment of the questionnairelinked to this study showed that European andinternational investors are increasingly seekingexposure to European bond markets. The shareof non-resident investments (investments fromoutside the single country) in euro-denominatedbonds issued by local entities grew from 30.7%in 2000 to 38.6% in 2003.

International diversification of EU investors’portfolios has significantly increased in recentyears, too, whereas for most EU countries theshare of foreign euro bond holdings rangedfrom 19% to 48% (median 37%) in 2000, thisincreased to 27% to 79% (median 61%) in2003.4

3.1 DEVELOPMENTS IN THE EURO-DENOMINATEDGOVERNMENT BOND MARKET

The sovereign issuance segment is still the mostimportant market segment of the bond market inthe EU. The relative importance of this marketsegment arises from a number of factors thatdistinguish sovereigns from other securities.These include the size of the market, thecreditworthiness of the borrowers, theavailability of a wide range of maturities, thefungibility of issues facilitating trading, thehigh liquidity (particularly of recently issuedsecurities), the fact of being accepted in openmarket operations and lending facilities, theexistence of a well-developed repo- andderivatives market and, as a result of thesefeatures, the coexistence of benchmark yieldcurves. The introduction of the euro created one

Table 1 Euro-denominated bonds by issuertype: historical development

Source: ECB securities database; bonds with a maturity > 1 year;the percentages represent the ratio of the amount of theindicated category to the total amount for the related quarter.Due to rounding, totals may not add up to 100%.1) MFIs include the Eurosystem and central banks.2) Non-MFI Corporations include non-financial corporationsand non-monetary financial corporations.

4 These figures are derived from the results of the questionnairelinked to this study after adjustment for outliers and referring tothe upper and lower quantile of the distribution.

RedemptionTotal in Non-MFIEUR bn Government MFI1) Corporations2)

2001 Q1 288 58% 34% 7%Q2 193 42% 48% 10%Q3 238 53% 39% 8%Q4 252 40% 50% 10%

2002 Q1 295 55% 37% 8%Q2 242 47% 41% 12%Q3 244 45% 45% 9%Q4 336 47% 45% 8%

2003 Q1 325 45% 47% 7%Q2 307 51% 42% 7%Q3 289 51% 40% 9%Q4 298 43% 46% 10%

average 276 48% 43% 9%

Gross issuanceTotal in Non-MFIEUR bn Government MFI1) Corporations2)

2001 Q1 388 45% 41% 13%Q2 336 43% 37% 20%Q3 311 45% 38% 16%Q4 353 31% 41% 27%

2002 Q1 423 53% 37% 10%Q2 370 45% 36% 18%Q3 295 46% 39% 15%Q4 363 40% 37% 23%

2003 Q1 450 48% 37% 14%Q2 463 48% 35% 17%Q3 391 47% 38% 15%Q4 376 32% 46% 22%

average 377 44% 39% 18%

OutstandingTotal in Non-MFIEUR bn Government MFI1) Corporations2)

2001 Q1 6,104 54% 37% 9%Q2 6,245 54% 37% 10%Q3 6,318 53% 37% 10%Q4 6,419 53% 36% 11%

2002 Q1 6,547 52% 36% 11%Q2 6,674 52% 36% 12%Q3 6,715 52% 36% 12%Q4 6,742 52% 36% 13%

2003 Q1 6,868 52% 35% 13%Q2 7,024 52% 35% 13%Q3 7,129 52% 35% 14%Q4 7,205 51% 35% 14%

average 6,666 52% 36% 12%

11ECB


3 MARKETOVERVIEW

of the world’s biggest markets for sovereignissuance. According to BIS data, the Europeangovernment bond market ranks third after theUnited States (USD 6,150 billion) and Japan(USD 5,022 billion), and the three togetheraccount for 84% of all government bondsoutstanding. Although the market is comparablein size to the US or Japanese markets themultiplicity of issuers and differences increditworthiness distinguish the euro areagovernment bond market from ist counterparts.

In the past public issuers in the euro areabenefited from a quasi-monopoly situation, butthey now compete for the same pool of funds.In this more competitive environment thesovereign issuers have had no alternative but toimprove the attractiveness of their securities.However, competition exists not only betweeneuro area countries but also with the top-ratednon-sovereign market segment. In theirfunction as basic investment products forinstitutional investors or central bank reservesEuropean government bonds also compete withUS Treasuries. A fresh impetus will alsoemerge from the 10 new EU Member States.

INCREASED SIZE OF THE GOVERNMENT BONDMARKET

As already mentioned, the government bondmarket is the dominant segment in terms ofmarket size. The amounts outstanding havedecreased relative to other segments, but it stillcontinues to represent nearly half the Euro areabond market. A share of 70% of this market isprovided by only three countries: Italy, Franceand Germany. A further share of 20%comprises the government issues of Spain,Belgium and Netherlands, while Austria,Finland, Portugal, Greece, Luxemburg andIreland account for 10% of the sovereign issuesoutstanding.

Although, in relative terms the amountsoutstanding have decreased over the past threeyears, the issuance activity of euro area MemberStates increased. This trend has occurred notonly in response to a high level of redemptionbut was also the result of an increase in centralgovernment funding needs as budget deficits ofthe largest euro area countries escalated. Fiscalperformance has had a significant impact on theissuance activity of the euro area governments.In the early years of EMU lower budget deficitscombined with revenues from the sales ofUMTS licences led to a reduction in the netborrowing requirements. In the followingyears, due to lower economic growth, there wasa shift from a budget surplus for the euro areaof 0.1% of GDP in 2000 to a deficit of 2.7% ofGDP in 2004. As a consequence, the grossborrowing requirements of euro area countriesincreased. For example, the net borrowing ofthe biggest debt issuers, Italy, Germany andFrance, was as high as the volume outstandingof the seven “small” euro area countriestogether.

However, several factors facilitated theplacement of sovereign issues. The marketenvironment, with historically low yields,provided convenient conditions for issuinglong-term debt. In addition, the high quality ofeuro area sovereign bonds furthers their

Chart 3 Residency of investors ineuro-denominated bonds

(%)

Source: Questionnaire to the Euro Bond Market Study 2004.Note: The term “foreign investors” refers to all investmentsfrom outside the single national economy. The share of foreigninvestors is calculated as the average of foreign investments ineach individual Member State (also including intra-euro areainvestments).

0

10

20

30

40

50

60

70

80

90

100

0

10

20

30

40

50

60

70

80

90

100

2000 2001 2002 2003

foreign investorsnational investors


acceptance in repo transactions or in openmarket operations and lending facilities.

Sovereign issuance in euro denominated bondswas predominantly from euro area sovereigns,but there have been occasional – albeitquantitatively unimportant – issues by othersovereigns from the new Member States5 (seeTable 3).

BENCHMARK GOVERNMENT BOND YIELDSPREADS OF EURO AREA COUNTRIES TIGHTENED

The trend of tightening government yieldspreads continued. Prior to EMU, yielddifferentials within euro area government bonds

had been determined by four main factors:expectations of exchange-rate fluctuations,different tax treatment of bonds issued bydifferent countries, credit risk and liquidity.After the introduction of the euro, currency-related premia were eliminated by theirrevocable fixing of legacy currency pairs.With respect to taxation differences, goodprogress was made in harmonising national taxtreaments. Thus yield differentials are mainlygenerated by the credit premium and theliquidity of the market. It was widely believedthat spreads would tighten.

Outstanding amounts Gross issuance RedemptionsDecember 2003 January-December 2003 January-December 2003

Country short-term long-term short-term long-term short-term long-term

Italy 120 972 217 215 211 211Germany 35 743 64 157 58 127France 109 713 238 130 217 70Spain 39 257 36 37 33 39Belgium 27 220 68 25 69 24The Netherlands 20 181 76 34 72 24Greece 4 130 4 31 2 18Austria 1 112 6 16 6 12Portugal 6 59 22 8 18 6Finland 6 46 11 13 13 11Ireland 1 28 63 7 64 1Luxembourg - - - - - -

Table 2 Euro-denominated central government debt securities in 2003

(EUR billions)

Source: ECB securities database.

5 See also European Central Bank, “Bond markets and long-terminterest rates in European Union accession countries”, October2003.

Debt securities outstanding Market share of debt securities Market share of debt securitiesCountry Dec. 2003 (EUR millions) issued by general government (%) denominated in euro (%)

Cyprus 7,020 79 34Czech Republic 45,146 40 1Estonia 268 43 47Hungary 45,490 79 13Latvia 1,098 90 40Lithuania 2,661 95 62Malta 3,055 87 0Poland 69,462 85 12Slovakia 10,992 88 17Slovenia 11,480 55 53

Table 3 New EU Member States

Source: ECB, statistical survey August 2004, (all maturities).

13ECB


3 MARKETOVERVIEWYields have indeed converged substantially

compared to the pre-euro period. The lowestyielding sovereign securities in the ten-yearsegment are still German government bonds(Bunds), not least because of the liquid EurexBund-Future contract that accepts only GermanBunds as deliverable bonds. The observableyield spreads of euro area sovereigns in 2000were 14 to 41 basis points (bp) over GermanBunds. That premium narrowed steadily andranged from -3 to 15 bp at the end of 2003. Inaddition to the risk and liquidity premium,another (more speculative) reason for thereduction of yield spreads could be the fact thatthe relative credit standing of the GermanBunds shifted a little last year as the budgetdeficit for Germany was above the 3% limit andGermany’s rating was being monitored closelyby rating agencies. A further reason is that theeuro capital market acquired more relevance ininternational bond indices, and smallercountries, in particular, could take advantage ofthis as any international investors seeking toreplicate these indices are also bound to investin the smaller issues.

Over the past few years the yield spreads ofAAA-rated government bonds (S&P rating)have narrowed to such an extent that theten-year yields of Austria, Belgium, France,Germany, Netherlands, Portugal and Spainhave nearly the same level. Finland had apremium of 8 bp, followed by Ireland with apremium of 13 bp, Italy with AA- rating(having recently been downgraded) and apremium of 15 bp and Greece with an A ratingand the highest premium of 16 bp (yield spreadsas of 22 November 2004).

Good progress was made on taxationdifferences within the euro area. As part of the“tax package” aimed at combating harmful taxcompetition, the EU decided to draw up alegislative instrument to overcome existingdistortions in the effective taxation of savingsincome in the form of interest payments. The taxregulations are aimed at an exchange of

information on non-resident EU nationals. Theinstrument in question is the Council Directive2003/48/EC of 3 June 2003 on taxation ofsavings income in the form of interestpayments. The Directive will come into force1 July 2005, provided that agreements are inplace with certain third countries (Switzerland,Andorra, Liechtenstein, Monaco and SanMarino) to ensure that equivalent measures areapplied in those countries. Furthermore, a longtransitional exemption has been granted toAustria, Belgium and Luxembourg, allowingthem to replace the exchange of information bya withholding tax.

In the future, yield spreads will thereforepredominantly result, all other things beingequal, from the different creditworthiness of theEU Member States and the different liquidityof the bonds as exchange rate fluctuationshave disappeared for legacy currencies anddifferences in tax treatment are gradually beingtackled.

Chart 4 Yield di f ferentials in the segmentfor ten-year government bonds

Sources: Bloomberg, Generic Treasury Bonds, DeutscheBundesbank calculations.

-5

10

25

40

-5

10

25

40

2000 2001 2002 2003

France: AAA – Aaa – AAAItaly: AA- – Aa2 – AABelgium: AA+ – Aa1 – AAIreland: AAA – Aaa – AAAThe Netherlands: AAA – Aaa – AAAAustria: AAA – Aaa – AAAFinland: AAA – Aaa – AAAPortugal: AA – Aa2 – AASpain: AA+ – Aaa – AAAGermany: AAA – Aaa – AAA


Box 1

DEVELOPMENTS IN EURO AREA LONG-TERM GOVERNMENT BOND YIELD SPREADS AND FISCALPOSITIONS

The development of different euro area countries’ government bond yields since the start ofEMU is of particular interest, not least for assessing convergence tendencies within the euroarea sovereign bond market. Instead of using the government bond yields themselves, it seemsmore intuitive to define yield spreads of the countries with respect to a “benchmark” country in

1 For example, for the Bloomberg’s EFFAS Government Bond Index, Germany’s share in the composition of the index for the maturitybucket of 7-10 years amounted to 25.1% in December 2003. The next largest share, namely 19.2%, was for France.

2 This argument for calculating the yield spreads relative to German long-term government bond yields was presented previously in theECB Monthly Bulletin, November 2003, pp 22-23.

3 Since the comparability of yields between countries is critical, harmonised long-term interest rates were taken from a Eurostatdatabase which provides these data for convergence assessments among EU Member States. Luxembourg was excluded from thesample, since no harmonised long-term interest rate based on secondary market yields of government bonds with maturities of closeto ten years are included for Luxembourg in the Eurostat database.

4 The cross-sectional max-min range is defined as the difference between the cross-sectional maximum and minimum yield spreadin a given month.

Chart A Long-term government bond yield spreads against Germany

Sources: Eurostat and Deutsche Bundesbank calculations.

0

10

20

30

40

50

0

10

20

30

40

50

1999 2000 2001 2002 2003

AustriaBelgiumSpainFinlandFrance

0

10

20

30

40

50

60

70

0

10

20

30

40

50

60

70

1999 2000 2001 2002 2003

GreeceIrelandItalyThe NetherlandsPortugal

the sample. For the long-term government bond segment investigated here, the best“benchmark” candidate seems to be Germany because of Germany’s large share in the long-termeuro area government bond segment1 and the fact that harmonised long-term Germangovernment bond yields were the lowest at the start of EMU.2 Chart A illustrates the fact that thecross-sectional pattern of the various countries’ harmonised long-term government bond yield3

spreads against Germany has gone through different phases since then. The general trend in thegovernment yield spreads showed an increase over 1999-2000, which is mirrored in thedevelopment of the cross-sectional mean spread and the dispersion as captured by the max-minspread range4 over this period (see Chart B). This trend did not turn around until early 2001.Afterwards, both the mean spread and the max-min range showed a strong decline until the end

15ECB


3 MARKETOVERVIEW

of 2003. The mean spread declined even belowthe levels of early 1999.

These general convergence patterns after early2001 are also valid for the sample includingGreece from 2001 onwards. While Greek yieldspreads at 55 bp in January 2001 were farabove those of the other countries in thesample, by early 2003 this strong “outlier”effect on the cross-sectional mean spread andmax-min spread was no longer visible. Acloser inspection of harmonised long-termgovernment bond yield spreads againstGermany shows that at the end of 2003, apartfrom Greece, spreads above 10 bp wereobserved only for Italy, Portugal and Austria.However, despite the increase in spreads overthe period from January 1999 to January

2001, the net change in spreads between January 1999 and December 2003 was negative for allcountries in the sample.

Empirical results for sovereign risk premia in the European government bond market stress thatfiscal developments (e.g. the debt-to-GDP ratio) play an important role for the yield spreads ofindividual government bonds.5 It therefore seems plausible that (relative) fiscal developmentsmight also have been one of the driving factors6 for the development of the “harmonised” centralgovernment bond yields and their spreads against Germany over time.7 This is also emphasised

Chart B Euro area government bond yields:cross-sectional mean spreads andmaximum-minimum spread ranges


mean spread, including Greece after 2001mean spread, excluding Greecemax-min spread range, including Greece after 2001max-min spread range, excluding Greece

0

10

20

30

40

50

0

10

20

30

40

50

1999 2000 2001 2002 2003

Spreads (in bp) Changes in spreads

Jan. 1999- Jan. 1999- Jan. 2001-Country Jan. 1999 Jan. 2001 Dec. 2003 Dec. 2003 Jan. 2001 Dec. 2003

Austria 14 30 11 -3 16 -19Belgium 20 36 9 -11 16 -27Spain 18 28 5 -13 10 -23Finland 21 22 4 -17 1 -18France 7 14 5 -2 7 -9Ireland 19 23 7 -12 4 -16Italy 22 38 17 -5 16 -21The Netherlands 10 13 4 -6 3 -9Portugal 20 36 11 -9 16 -25

Average 16,78 26,67 8,11 -8,67 9,89 -18,56

Greece (since 2001) 55 16 -39

Average (including Greece) 29,5 8,9 -20,6

Spreads of euro area countries’ long-term government bond yields against Germany


5 See, for example, Bernoth, von Hagen and Schuhknecht, “Sovereign risk premia in the European government bond market”, WorkingPaper Series No 369, European Central Bank, June 2004.

6 Other factors might include, for example, different degrees of liquidity of the underlying government bond issues. Furthermore, thef iscal position factor might not be linear in its influence; in times of high overall liquidity, investors might be searching for yield andmore willing to accept smaller yield spreads in compensation for risks.

7 See European Central Bank, Monthly Bulletin, November 2003, p 23.


by the observation that the turning point in 2001 and the strong decline of the mean spreadcoincided with a period during which Germany experienced a significant worsening of its fiscalposition.8

Simple scatter plots of the changes in yield spreads against the changes in relative debt-to-GDPratio, while no substitute for a thorough econometric analysis, allow – at least visually –potential “correlation patterns” between a pair of variables to be examined, even for smallsample sizes. While no clear picture emerges for the early 1999-2001 period, for the latterperiod from 2001 to 2003 the plot suggests that a positive correlation between an improvementin the countries’ debt-to-GDP ratios relative to Germany and a decline in the country’s long-term government bond yield spread might in fact exist. At first glance, this could be interpretedas suggesting that investors, apart from other factors, may also have rewarded countries forrelatively sound fiscal policies with a decline in spreads.9 While this story would fit economicintuition and empirical results for bond pricing at the disaggregate level, given the scarcity ofdata one has to be careful not to over-interpret such scatter plot “aggregate level” evidence.Furthermore, even if taken at “face value”, the difference in tendencies across the two plotswould suggest that such a relationship need not be stable. Finally, the French exampleunderlines the fact that – even for the trend in yield spreads over 2001-2003 – this factor can atbest give a partial explanation since the change in the French yield spread is negative despite aworsening of France’s relative fiscal position. Other factors that might have been relevant –especially at the end of the observation period – include, for example, the relatively highliquidity in capital markets and the ensuing search for a better yield, which are said to have beencontributing factors for yield spread compression observed elsewhere in bond markets,particularly, for example, in the emerging economies’ bond markets up to the end of 2003.10

8 This is best illustrated by the fact that in 2002 and 2003 Germany found itself unable to keep the budget deficit as a percentage of GDPbelow the critical threshold of 3%. However, it was already close to this threshold in 2001, with an absolute value of 2.8% of GDP.Source: Eurostat.

9 See also European Central Bank, Monthly Bulletin, November 2003, pp 22-23.10 See, for example, IMF, Global Financial Stability Report, April 2004.

Chart C Changes in relative debt-to-GDP ratios and changes in yield spreads

Sources: Eurostat and Deutsche Bundesbank calculations.Notes: The change in the yield spread (relative debt-to-GDP ratio) is calculated as the difference between the value at the end pointversus the starting point of the reference period. The relative debt-to-GDP ratio is defined as the difference between the debt-to-GDPratio (in percentage points) of the individual country and the German ratio.

0

4

8

12

16

20

0

4

8

12

16

20

-12 -10 -8 -6 -4 -2 0 2 4

x-axis: change in relative debt-to-GDP ratio (1999 to 2001)y-axis: change in yield spread (Jan. 1999 to Jan. 2001)

Belgium ItalyAustria

Spain

Portugal

France

IrelandThe Netherlands

Finland

x-axis: change in relative debt-to-GDP ratio (2001 to 2003)y-axis: change in yield spread (Jan. 2001 to Dec. 2003)

-40

-36

-32

-28

-24

-20

-16

-12

-8

-40

-36

-32

-28

-24

-20

-16

-12

-8

-16 -12 -8 -4 0 4

The Netherlands

Greece

France

Finland

Portugal

Austria

Ireland

ItalySpain

Belgium

17ECB


3 MARKETOVERVIEWGREATER DIVERSITY IN THE CHARACTERISTICS

OF GOVERNMENT BONDS

The existence of national legislative andregulatory frameworks together with a largevariety of investment and debt managementstrategies made it difficult to establish a singlemarket. Every country has its own issuancepolicy with individual strategies for creatingtheir debt products. However, from certainelements it is apparent that, because of theirdomestic focus in the EU, the trend is forgovernments not to cooperate but to coordinate.One example of coordination betweengovernments is the adoption of a similar couponcalculation convention. Bilateral andmultilateral contacts have also been developedand one forum for such contacts is the WorkingGroup on EU Government Bonds and Bills(“Brouhns Group”) established by theEconomic and Financial Committee.

One clear trend in the government bond marketis towards larger issue sizes as national debtmanagers have focused on improving liquidityin their instruments by launching benchmarkbonds of €5 billion or more in order to beeligible for trading on the EuroMTS electronictrading platform. Bonds up to a volumeoutstanding of €20 billion have an 80% shareof total bonds outstanding, while small bondsof up to €500 million have all but disappearedand bonds up to €5 billion have a market shareof only 4% of bonds outstanding. Smallercountries in particular are tending to increasethe volume of existing bonds rather than issuenew debt to obtain market liquidity (fungibleissues). Certain countries have also arrangedprogrammes to buy back or exchange bondsprimarily in order to increase the liquidity ofon-the-run issues and to exchange old illiquidbonds.

The maturity spectrum up to ten years isrelatively homogenous in terms of the share ofsovereign issuers. More interesting in terms offragmentation is the long-term spectrum, whereno single maturity is offered by all of the sixbiggest euro area sovereign issuers. These

incomplete maturity structures for each issuerresult from the different sizes of the debtrequirements of the countries concerned. Onlybig countries are able to serve the wholematurity spectrum.

The coupon structure itself did not changesignificantly. It has been possible to observe aslight reduction in the share of fixed coupons,but 65% of government bonds still have fixedrate coupons.

One clear trend in EMU is the search for newtypes of instruments or the replication ofexisting securities by issuing euro-denominatedbonds linked to an inflation index. Othergovernments in the EU, such as the UnitedKingdom and Sweden, have many years ofexperience of issuing index-linked bonds. TheFrench Treasury started issuing bonds linked tothe French inflation index in 1998. Other EUMembers States also placed securities in thismarket segment and Greece and Italy issuedeuro area inflation-linked bonds in 2003.Germany has also announced that it will issueindex-linked bonds. For detailed informationsee Chapter 5.5.

Another niche is the “TEC10” issued by France,with a quarterly coupon that is linked to thecreated TEC-10 index, an average yield of

Chart 5 Euro area government bondmaturity prof i le

(EUR billions)

Sources: Bloomberg, Deutsche Bundesbank calculations.

0

50

100

150

200

250

300

350

400

450

0

50

100

150

200

250

300

350

400

450

2005 2007 2009 2011 2013 2015 2017 2019 2022 2024 2026 2028 2030 2032 2034


Box 2

RECENT DEVELOPMENTS IN EURO AREA GOVERNMENTS’ DEBT DURATION

The average debt duration – i.e. the duration of a composite of bonds, such as a bond portfolioor the government bond market – is derived as a weighted average of the duration of theindividual bonds, taking into account their weights in the respective composite. All factors thatinfluence the duration of bonds therefore also affect the average debt duration. The duration ofa bond is calculated as the weighted average term to maturity1 of its discounted cash flows. Onlyfor a zero coupon bond with no cash flows prior to maturity is the debt duration equal to theremaining maturity; for a fixed coupon bond, the cash flows from remaining coupon paymentsimply that the duration will be smaller than the maturity. Furthermore, the larger the coupon andhence the discounted cash flows from coupon payments relative to the discounted cash flow

1 It is therefore denoted in units of time, such as years.

from the repayment of the principal, the smaller the duration. The duration is also used as arough approximation for the interest sensitivity of a bond’s price. A shorter duration impliessmaller interest rate sensitivity since an investor who receives a larger portion of the discountedstream of cash flows at an early date may take advantage of reinvestment opportunities shouldthe interest rate rise. The investor who expects interest rates to rise (fall) in the future willtherefore choose to hold bonds with a shorter (longer) duration. From the issuer’s point of viewthe opposite is true. Faced with a relatively low (high) interest rate environment in the presentbut an expected rise (fall) in interest rates in the future, an issuer will prefer, all other thingsbeing equal, to issue long-term (short-term) bonds today. However, debt management bygovernments also has to take account of other factors affecting their financing costs, such as theliquidity of bond issues. Since investors demand compensation for risks through higher yields

Chart A Index characterist ics over t ime

Sources: Bloomberg and Deutsche Bundesbank calculations.

average lifeaverage adjusted duration

4.5

5.0

5.5

6.0

6.5

7.0

7.5

8.0

4.5

5.0

5.5

6.0

6.5

7.0

7.5

8.0

1999 2000 2001 2002 20031

2

3

4

5

6

7

1

2

3

4

5

6

7

1999 2000 2001 2002 2003

3-month Euriboraverage couponaverage yield

EFFAS Euro Bloc “long-term” Government Bond Index (maturities above one year):index characteristics and money market interest rates

19ECB


3 MARKETOVERVIEW

2 For further details of the index, see Bloomberg.

and a high liquidity of bonds implies a lowerliquidity risk for the investor, a governmentmay benefit, all other things being equal, frombetter financing terms if it concentrates on asmall number of liquid benchmark bond issuesrather than on a larger number of smallerissues. Depending on market shares andconcentrations of issue activities in certainmaturity ranges, a government’s managementof issues may over time have a significantimpact on the development of the averagematurity structure of government debt in theeuro area. However, the development of debtduration depends not only on new issues, butalso on the duration structure of outstandingdebt.

To illustrate the aggregate development ofeuro area government debt duration over time,Chart A looks at some characteristics of theBloomberg/EFFAS Euro Bloc governmentbond index.2 This representative index isavailable for bonds with maturities above one

year. Also included is the three-month Euribor, to capture trends in money markets. As Chart Ashows, the average life – as a measure of weighted-average maturity of bonds – and the average

Chart B Shares of maturity ranges in theeuro area long-term government bondindex

Sources: Bloomberg and Deutsche Bundesbank calculations.Note: Based on face value data for the Bloomberg/ EFFAS EuroBloc Government Bond Index covering bonds with maturitiesabove one year and for the corresponding maturity range EuroBloc sub-indices. The monthly shares in per cent are given forthe period of January 2001-June 2004.

1-3 years3-5 years5-7 years7-10 years> 10 years

8

12

16

20

24

28

32

8

12

16

20

24

28

32

2001 2002 2003

(Share in per cent for maturity range)

Chart C Government long-term bond indices of euro area member countries: average l i fe

(average life, years)

Sources: Bloomberg and Deutsche Bundesbank calculations.Note: December 2000-June 2004.

AustriaBelgiumFinlandFranceGreeceGermany

4.5

5.0

5.5

6.0

6.5

7.0

7.5

8.0

4.5

5.0

5.5

6.0

6.5

7.0

7.5

8.0

2001 2002 2003

IrelandItalyThe NetherlandsPortugalSpainEuro Bloc

5

6

7

8

9

10

5

6

7

8

9

10

2001 2002 2003


OATs (obligation assimilable du Trésor) with aconstant maturity of ten years. These bonds

adjusted duration3 of the bond index has increased over time. This has coincided with adecreasing trend in the index’s average coupon. These trends in average life and adjustedduration are visible throughout 1999 and until June 2004. However, in the period since 2001,which is characterised by a relative decline in the average yield and short-term interest rates, thedevelopment has been more pronounced. Governments may therefore have taken advantage ofthe relatively favourable interest rate environment by increasing the debt duration. Acorresponding picture emerges in Chart B for the period after 2001. A comparison of the sharesin the face value of Bloomberg/EFFAS sub-indices disaggregated by maturity ranges shows adecreasing trend for the face value share of the one to three-year maturity segment and anincreasing trend for the share of bonds with maturities above ten years. This matches theobservation that the average life and the adjusted duration for the full maturity range increasedover this period. However, as Chart C shows, the trends in maturity distribution evolveddifferently in different countries. The average life of the indices covering bonds with maturitiesabove one year shows, for example, a declining trend for Portugal, while a strong increase from7.45 years in December 2000 to 8.94 years in June 2004 is observed for Italy. As argued above,such different developments in the cross-country indices’ average life trends over time may atleast partly reflect the countries’ individual management strategies for structuring governmentdebt.

3 Bloomberg’s average adjusted duration is a market-value weighted index based on the duration of the bonds. For details seeBloomberg.

Chart 6 Maturity spectrum occupancy bythe s ix biggest euro area sovereignissuers

Sources: Bloomberg, Deutsche Bundesbank calculations.

ITDEFRESBENL

0

20

40

60

80

100

0

20

40

60

80

100

2005 2007 2009 2011 2013 2015 201 7 2019 2022 2024 2026 2028 2030 2032 2034

(%)

were issued for the first time on 9 April 1996.This market segment accounts for 3% of theamount outstanding of French governmentbonds.

The Italian Treasury also operates in nicheswith its floating rate issues, with a share ofoutstanding debt that has decreased from 20%in 2001 to 17% in 2003 as gross issuance didnot reach the value of redemption.

HARMONISATION IN PUBLIC DEBTMANAGEMENT THROUGH TRANSPARENCY ANDSTANDARDISATION

One competitive disadvantage of the euro-denominated sovereign bond market relative toother bond markets has resulted from theperception of the investors as 12 self-containedcountries. Each country’s specific legislationand regulatory framework combined with itsown strategy for managing public debt madeunification of the debt market difficult. Asalready demonstrated, the characteristics of thesovereign bond market are, on the whole,relatively homogenous but with a growing

21ECB


3 MARKETOVERVIEWtendency to move towards niches. Nevertheless,

harmonisation in public debt management isobservable as individual debt managersendeavour to achieve greater transparency intheir strategies and to develop standardisedtools for issuing debt.

One trend is towards greater transparency, withinformation about public debt being publishedon the internet and in periodical bulletinsand annual reports. One example is thestandardisation in the regularly published andpre-announced issuance calendar. Everycountry has a calendar which contains the dateand the planned issue volume. However, thecalendars vary in terms of the time horizon.Some plan for the next three months and othersfor the next whole year. In addition, the bindingcharacter of the issue calendar differs withregard to the time as well as to the issuancevolume.

For the primary market procedure nearly allcountries’ issuing activities generally use thetraditional auction method. However, a methodcombining an auction procedure with the use ofsyndication is also becoming more common.The secondary market activity traditionallytakes place in the wholesale over-the-counter(OTC) market. In the euro area the widelyrecognised need to create the facilitiesnecessary for cross-border trading gave rise toone technological innovation. Designed by theItalian MTS Group, the London-basedEuroMTS was set up as an electronic tradingsystem for euro benchmark bonds with a pool ofinstruments including government bonds ofevery eligible issues within the EU.

Another trend in the trading environment is thevirtually exclusive use of primary dealersystems as competition between issuingMember States has intensified and the nationalgovernments need to broaden their investorbase throughout Europe and beyond. Thereforenearly all euro area countries have implementeda primary dealer system. The primary dealersare obliged not only to buy securities whenissued, but also to set ask and bid prices in the

secondary market. In order, in particular, toincrease the attraction for internationalinvestors, the extension of the workingrelationship to foreign primary dealersincreased. The emergence of an internationalcorps of primary dealer banks present in amajority of EU sovereign issuers is aninteresting phenomenon.

MARKET INTEGRATION ADVANCED

As mentioned above, a certain degree offragmentation still exists in the euro areagovernment bond market as well as in theEuropean bond market. Structural discrepanciessuch as the non-unification of tax structuresalong with different accounting rules,settlement systems, market conventions andissuing procedures do not make it easy tointegrate the government bond market.

The introduction of the euro removed exchangerate risk for euro area investors by reducing thehome bias of investments within the euro area.This contributes to an increase of cross-borderinvestments within the euro area whilemaintaining the ongoing integration of the eurobond market. Furthermore, the Euro areagovernment bond market has gained inimportance as developments in the internationalportfolio composition show a tendency to anincrease in the shares of non-residents’investments in euro denominated bonds. Thisled to the suggestion that the introduction of theeuro has contributed to a geographicalreallocation of portfolios.

3.2 DEVELOPMENTS IN THE EURO-DENOMINATEDNON-GOVERNMENT BOND MARKET

Non-government issuing activity has gainedincreasing importance since the introduction ofthe euro. From end-2000 to end-2003 theoutstanding amount of private (non-government) debt securities issued by euro arearesidents rose from €3,052 billion to €4,558billion, which is an increase of almost 50%. Theamount outstanding issued by non-financial


euro area corporations in the same period greweven more strongly, by 58%. In continentalEurope, which particularly exemplified thetraditionally strongly bank-based nature offinancing structures, government bonds andbank debt securities have dominated the marketfor decades and still do. Nevertheless,Pfandbriefe-like and corporate issuing iscatching up.

There are a variety of reasons why other issuershave increased their recourse to the bondmarket, especially in the period up to 2001.First, the introduction of the euro led to greatermarket integration and, at least in part, to thedevelopment of a homogeneous euro capitalmarket which represented real competition forthe US dollar-denominated capital market.Major M&A activity in the corporate sectorensued as the 1990s came to a close. Inaddition, the hi-tech boom ushered in anenormous demand for investment, especiallyamong IT and telecommunications enterprises.For the first time, many enterprises took majorrecourse to the bond market to cover theirfinancing needs.

In addition, demand-side reasons also played arole. The low-interest rate environment on themoney and capital markets caused investorsto look increasingly for higher-yieldinginvestment opportunities. Conversely, the lowinterest rate level made it more attractive forprospective borrowers to issue bonds too. Theintroduction of the euro led to the abolition oflegal or contractual investment restrictions formany institutional investors which werepreviously not allowed to invest in foreigncurrencies. Thus, as pension funds andinsurance companies were not currency-constrained any longer, the pool of “domestic”investors was enlarged considerably by theintroduction of the euro. The elimination ofcurrency risk within the euro area broadened thedemand for other diversification opportunities,especially by means of credit spreads. Thesefactors have contributed to the share ofcorporate bonds in European bond marketissuing activity rising from just under 10% toover 14% between 1999 and 2001. This processwas promoted by a distinct easing of thebudgetary situation of central government inmany European countries; governmentborrowing was reduced, affordingopportunities to “crowd in” private issuers.Whether there really was a “crowding-in” effectobservable in the markets cannot be saiddefinitively. One indication could be anincrease in private issuance in market segments(concerning, for example, maturity spectrumand risk characteristics) that governments arevacating. It is in fact difficult to detect thoseshifts in issuance activity and attribute thembeyond doubt to the “crowding-in” effect, asdifferent influences on market activities cannotbe separated clearly from each other. The post-2001 economic slump, the unfavourabledevelopments in capital markets and a series ofeconomic scandals (Enron, WorldCom etc.) allled to investors becoming increasingly risk-averse beginning in 2001, in some casesreversing the previous years’ trends. Thepercentage of overall euro-denominated bondmarket issuance accounted for by corporatebonds declined sharply between 2001 and 2002,rising only slightly to a mere 8.5% in 2003.

Chart 7 Breakdown of gross issuingvolumes of long-term euro-denominatedbonds by type of (euro area) issuer 2003(in per cent)

centralgovernment

41

other generalgovernment

3

MFIs39

non-MFIfinancials

11

non-financialcorporations

6

Source: ECB securities database.

23ECB


3 MARKETOVERVIEWDespite its impressive growth the segment of

non-government issuance is still confrontedwith several problems which might deterinvestors from taking positions in these marketsegments.

Institutional investors in particular might stayaway from the private bond market because of arelative lack of liquidity.6 While governmentbonds issued by the large EU Member Statestend to amount to €15 billion to €20 billion perline, the standard size of a corporate bond isonly one-tenth of that amount or even less.Mainly due to issuance volumes on thisscale, secondary market activity remainscomparatively low. Other important factorslimiting the secondary market activity arerelated to the specific nature of the risk which istraded. Trading private credit is much moreinformation sensitive than trading (similar oreven larger volumes of) government credit. Inaddition, the credit market is characterised byirregular and asymmetric flows of informationwhich is another reason why market makershave so far limited their willingness to commitcapital and resources to guaranteeing liquidityin the market and therefore trading tight spreadson a continuous basis.

Nevertheless, some progress has been made:among corporate bonds the share of smallissues below €500 million decreased from about40% of total issuance in 1999 to well below20% in 2003.7

An additional problem is the relatively limitedsupply of hedging instruments for non-government issues. The most commonly usedinstruments for the hedging of bond positionsare bond futures. These are typically developedon the basis of government bonds. When thesefutures are used to hedge financial or corporatebonds, major basis risks are incurred. In case offinancial turmoil the losses on riskier assets canlead to a “flight to quality” with the result thatthe prices of government and non-governmentbonds move in opposite directions. This wasindeed the case after the LTCM crisis in 1998.

There is a key interrelation between theliquidity of a market and the existence ofhedging opportunities. Without the possibilityto hedge positions efficiently, the tendency toinvest in this market remains subdued. Thedevelopment of indices in non-governmentbonds and the introduction of futures andexchange-traded funds based on these indicesmay be a possible answer to these problems.The success of the “Jumbo-Pfandbrief” inGermany shows a possible way to solve theliquidity problem in a formerly arcane,fragmented market through the enlargement ofissue sizes. However, even in the case of theflourishing Jumbo market, the introduction of afutures contract in 1998 did not succeed and thecontract was discontinued some six monthslater. Another possible solution to the liquidityproblem could be to increase the issue size or toissue bonds fungible with previous bonds witha limited set of maturities. The transparency ofthe private bond market can be improved byquotations on electronic trading systems.

The further role of non-government issuance inbond markets will largely depend on the rolethat market-based financing plays in the future.As banks have to adjust to the new supervisoryrules of Basel II, bank loans may become moreexpensive and the disintermediation tendenciesin the financial sector might become stronger.On the demand side the pool of investors will beenlarged as the problems of ageing societiesbecome more immediate and the pensionsystems in many European countries movetowards funded pension plans.

TYPES OF NON-GOVERNMENT BONDS

The following paragraphs offer some informationabout the different types of non-government

6 See F. Fabozzi and M. Choudhry, “The handbook of Europeanfixed income securities”, 2004, according to which no morethan 20% of investment-grade corporate bonds outstanding aretraded regularly: “The rest are locked away in investors’portfolios...trading tends to be concentrated in newer and largerissues.”

7 European Commission, Quarterly bond market note, December2003, “Special Feature: Five years of EMU – Evolution of theeuro-denominated bond market since 1999”.


bonds in Europe. Most of the data presented inthis section are based on statistics published bythe European Commission.8 Unless otherwiseindicated, the data refer to the issuance of bonds,not to amounts outstanding. New marketdevelopments can be observed earlier in issuanceactivity than in amounts outstanding.

BANK DEBT SECURITIESAlthough the corporate bond market hasattracted particular attention following theintroduction of the euro, the non-governmentbond market in all euro area countries is stilldominated by bank debt securities. Thissegment encompasses numerous different typesof bonds, ranging from unsecured bank debtsecurities to covered bonds (mostlyPfandbriefe), which were at first issued only byGerman mortgage banks. Unsecured bank debtsecurities are considered in this paragraphalthough they bear more resemblance tocorporate bonds than to Pfandbriefe or othercovered bonds. “Jumbo Pfandbriefe”, on themarket since the mid-1990s, are regarded bymany investors as an alternative to governmentbonds because of their high liquidity and first-class credit rating. In 1999 German issuersaccounted for around 95% of all Pfandbriefeissued in the market. Although German banksstill accounted for more than 70% of theissuance of euro-denominated Pfandbriefe in2003, in the meantime a number of Europeancountries (e.g. Spain, Luxembourg and Ireland)have passed legislation legalising the issuanceof covered bonds.

Along with government bonds, bank debtsecurities have traditionally played a dominantrole in European bond markets, which is anindicator of the dominant role of banks inEurope’s financing structures. The dominanceof bank debt securities might be explained bythe privileged treatment of bank debt in thecapital adequacy regime of the current BaselAccord.9 Bank debt securities share thecommon traits of being small to medium-sizedissues (the proportion of newly issued financialbonds with an issuance volume below €500million was almost 50% in 2003) and have

credit ratings ranging from good to very good.The dominance of financial bonds also explainswhy the average rating in the investment gradebond market in Europe is considerably higherthan in the United States (around Aa1,compared to A3 in the USA).10 An importantdifference to the government bond sector liesin the large percentage of variable rateinstruments, which in 2003 accounted for over40% of the issue volume.

The agencies can be regarded to a certain extentas a subset of the market for financial bonds.Most agencies are financial services providerswhich concentrate on particular financingfunctions, such as housing or the promotion ofsmall and medium-sized enterprises. In mostcases these agencies are, at least implicitly,equipped with a state guarantee and thereforehave first-class credit ratings. More than 80%of the agencies active in the European bondmarket are rated AAA. Since the introduction ofthe euro this segment has been dominated by thebenchmark programmes of KfW and FreddieMac, but this dominance has weakened asFreddie Mac retreated from the eurodenominated market in the wake of anaccounting irregularity in 2003.

Issuance by agencies has increased slowly butsteadily in recent years, both in absolute termsand as a percentage of total issuance. In 2003agency issuance accounted for 4% of totalissuance in the euro bond market. The averagevolume of a single agency issue has increasedcontinuously since 1999, and issues below€500 million accounted for only 10% of thismarket segment in 2003. In 2003 about 60% ofagency issues had a maturity of between threeand seven years. In the past three years 85% to90% of newly issued agency bonds were fixedrate coupon bonds.

8 Informative charts about structural developments are containedin the statistical annex to European Commission (2003); moredetailed figures are presented in the statistical annexes to theEuropean Commission’s monthly and quarterly bond marketnotes.

9 Fabozzi/Choudhry (2004), p 174.10 Fabozzi/Choudhry (2004), p 174.

25ECB


3 MARKETOVERVIEWASSET-BACKED SECURITIES (ABSs)

The strong growth in asset-backed securities(ABSs) began as early as the mid-1990s. Evenafter the economy began to slump in 2001, theshare of ABSs in issuing activity largely heldfirm. Many banks and other financial servicesproviders use the asset securitisation functionof ABSs specifically to remove risk from thebalance sheet and improve balance sheet ratios,which means that, in difficult market phases,the demand for securitisation may in fact tend torise. ABSs are generally long-term variable rateinstruments with an excellent credit rating. Theproportion of issues with a maturity of morethan ten years has increased considerably since1999 and stood at 70% in 2003. The percentageof floating-rate issues in the same year wasmore than 90%. Credit rating has been a majorfactor in the success of ABSs in the market. Thecredit quality of an ABS issue is completelyindependent of the rating of the originator, withthe result that ABS issuance has profited fromthe “flight into quality” since 2001. Thepercentage of AAA-rated ABSs rose to nearly100% by as early as the end of 2002 in the lightof difficult market conditions and investors’pronounced risk aversion. In 2003 ABSsaccounted for just over 4% of all euro areaissuing activity, and this figure has remainedlargely unchanged over the past three years.ABS are particularly widespread in Italy, acountry in which this instrument is also usedintensively by general government. Other euroarea countries, such as France and Germany, areunderrepresented in this segment because thesecountries have well-developed markets forPfandbriefe and/or covered bonds. This clearlyreduces the need for mortgage-backed securities(MBSs), a key subset of ABSs.

CORPORATE BONDSThis segment saw the strongest growth in thethree years following the introduction of theeuro, despite the fact that the size of the euroarea corporate bond market continues to pale incomparison with that of the US market. Theoutstanding volume of euro-denominatedcorporate bonds grew by 95% from end-1998 toend-2003, compared to a growth rate of only

37% in the case of financial bonds (excludingcovered bonds). The corporate segmentparticularly highlights the differences incorporate financing structures between theUnited States and Europe. Although the debateabout the advantages and disadvantages ofbank-based versus market-based financialsystems has not been settled finally,11 it is anundisputed fact that bank-based economies alsobenefit from a developed corporate bondmarket. Corporates are the segment of the bondmarket which benefits most from the trendtowards disintermediation. One expectedconsequence of the new regulatory regime ofBasel II is that bank loans will be granted onmuch stricter conditions, at least for companieswith lower credit ratings. Even in cases wherebank loans are readily available, companieshave an interest in broadening their financingbasis in order to increase their flexibility. Forexample, bond financing offers betteropportunities for companies to incur long-termdebt with a maturity above ten years as banksmight be reluctant to grant loans with such longmaturities.

Corporate bonds tend to have the lowest creditratings in the bond market and are therefore themost strongly affected by economic downturns.This was reconfirmed after the 2001 economicslump, although a key structural difference tothe US corporate bond market lies in the lowpercentage of speculative grade issues in theeuro area. In the euro area, these make up onlyaround 10% of the corporate bond market,compared with around 40% in the USA. If oneincludes the fact that market capitalisation in theeuro area is relatively low compared with theUSA, Europe may be said to be an emergingmarket for “junk bonds”. This holds true,although the divide between investment andnon-investment grade issues is not as strong asit is in the US, as investor guidelines in Europeare less strict.12 After the beginning of the

11 For an overview, see A Demigürc-Kunt and V Maksimovic,“Funding growth in bank-based and market-based financialsystems: Evidence from firm-level data”, Journal of FinancialEconomics, Vol 65, Issue 3, September 2002, pp 337-363.

12 Fabozzi/Choudhry (2004).


economic downturn in 2001, many prominentcompanies (among them British Airways, ABB,Ericsson and Ahold) became “fallen angels”,that is they were downgraded from investmentgrade to non-investment grade.

The proportion of smaller issues (up to €500million) has decreased considerably since theintroduction of the euro (from 40% to below

20%). Nevertheless, the percentage of issuesabove €2 billion, which was more than 30% in2001, also decreased to less than 10% in 2003.A study13 by Paul Harrison in 2001 stressed the

1 To be more precise, Merrill Lynch offers so-called option-adjusted spreads (OAS spreads, def ined against the swap curve) for theirrespective indices. The corporate bond yield spread is therefore calculated as the difference between the OAS spreads on thecorporate bond index in question and the respective direct government bond index.

2 This coincides with global developments in capital markets after the shock of the terrorist attack on the USA on 11 September 2001.3 Looking at short-term interest rate conditions in the money markets, the peaks in the three-month Euribor occurred in November 2000

and a smaller one in May 2002. From then on until December 2003, short-term interest rates declined strongly (from 3.46% to2.15 % per annum). While the decline in short-term interest rates started earlier, the positive development in liquidity conditionsprobably contributed to the spread compression observed after October 2002.

Box 3

RECENT DEVELOPMENTS IN CORPORATE BOND SPREADS IN THE EURO AREA

Taking the aggregate euro area government bond market as a reference point, euro area corporatebond yield spreads are calculated as the difference between yields on corporate bond indices and the

yields on the relevant euro area governmentbond indices.1 Since this investigation of recentcorporate bond spread developments concentrateson spread tendencies in the cross-sectionaldimensions – first for the maturity spectrum,then for the rating spectrum – the governmentbond reference category will be adjusted tomatch the respective cross-sectional dimensions.

For the maturity dimension of investment-grade corporate bond spreads, each spreadis calculated against a government bondindex with a corresponding maturity range. AsChart A shows, one can identify three distinctphases since the start of EMU. The firstphase – until roughly the end of 2000 – ischaracterised by an increasing trend incorporate bond spreads across the maturityspectrum. The subsequent period until aboutthe end of October 2002 saw two pronouncedpeaks in spreads, with a downswing in-between. The first clear jump in spreadsoccurred in September 2001, with the peak

following in October 2001.2 The second, much stronger peak came in October 20023 (for theaggregate corporate bond index, the spread amounted to 120 bp at the end of October 2002,

Chart A Investment grade corporate bondspreads by maturity ranges

(in bp)

Sources: Merrill Lynch and Deutsche Bundesbank calculations.

0

20

40

60

80

100

120

140

160

0

20

40

60

80

100

120

140

160

1999 2000 2001 2002 2003

aggregate1-3 years3-5 years5-7 years7-10 years

Euro area corporate bond indices (investment grade):yield spreads against euro area government bond indices,by maturity buckets

13 Paul Harrison, “The impact of market liquidity in times of stresson the corporate bond market: pricing, trading, and theavailability of funds during heightened illiquidity”, paperpresented at the BIS Third Joint Central Bank Conference on RiskMeasurement and Systemic Risk, 2001.

27ECB


3 MARKETOVERVIEW

compared to 92 bp at the end of October 2001). The last phase – from November 2002 until endof 2003 – finally saw a strong compression in corporate bond spreads across the maturityspectrum. Spreads did not decrease to the levels of early 1999, but – apart from the 1-3 yearsmaturity bucket – were lower than at the beginning of 2001. Overall, investment-gradecorporate bond yield spreads across the maturity spectrum show a great deal of co-movement inthe last phase, especially after September 2001. Since then, the particular behaviour of thecorporate bond yield spread for the 1-3 yearmaturity range seems to have vanished, whileits share in the face values of the indices in the1-10 year range increased (compared with theend of 2000, see Chart B).

Chart C highlights the development ofcorporate bond spreads along the ratingdimension only. In this case, all spreads aredefined relative to the AAA-rated directgovernment bond index. The spread for theAAA-rated corporate bond index thuscaptures the higher risk quality of corporatebonds versus government bonds belonging tothe same rating class. A slight decline in theoverall yield spread trend since 2001 is bestseen in AAA-rated corporate bonds. Sincelower ratings can be expected to correspond tohigher risks within the corporate bond indexsector, corporate bond spreads againstgovernment bond yields with AAA-ratingincrease with lower corporate bond ratings. Inaddition, with lower ratings the cyclical

Chart B Shares in face values of corporate bond indices, by maturity ranges


1-3years

3-5years

5-7 years

7-10years

1-3years 3-5

years

5-7 years

7-10years

Chart C Corporate bond spreads in therating dimension

(in bp)


Euro area corporate bond indices (investment grade):yield spreads against AAA-rated government bonds,by ratings of corporate bonds

1999 2000 2001 2002 2003

AAA-ratedAA-ratedA-ratedBBB-rated

0

40

80

120

160

200

240

280

0

40

80

120

160

200

240

280

Shares in face values in December 2000,euro area corporate bond indices by maturity ranges

Shares in face values in December 2003,euro area corporate bond indices by maturity ranges


development over time increases. Focusing on the 2001-2003 period, the peaks of September/October 2001 and October 2002 show up strongly for the spreads of corporate bonds with Aand BBB ratings, i.e. at those times the risk compensation worked as a multiplier ofdevelopment across the rating spectrum. The same general pattern of more pronounced peakswith lower ratings can be replicated when the spreads are defined with respect to the AAA-corporate bond index.

The degree of risk in terms of default and recovery are, of course, of a much larger scale for theso-called high-yield corporate bond segment. However, this segment has experienced dynamicgrowth in recent years.4 Chart D therefore compares the corporate bond spreads – nowcalculated against AAA-rated corporate bonds – for the lower grade investment ratingcategories and for the high-yield corporate bond sector. The spreads in the investment gradesector are clearly dwarfed by those in the high-yield sector. However, for the high risk, high-yield sectors with B ratings and below, the peaks in September/October 2001 are higher thanthose in October 2002. Moreover, the general spread compression observed for the corporatebond sector after October 2002 was particularly strong for the very high risk segmentcomprising high-yield corporate bonds with CCC ratings and below. The yield spreadcompression in general, but in particular for the high yield sector, points to a specialdevelopment in capital markets since October 2002. This phase of bond yield spreadcompression has probably been influenced by factors such as the high liquidity in markets andthe search for yields in capital markets. These factors have also been linked in the literature tothe yield spread compression observed in the latter phase for other high risk bond categories,such as for emerging markets bonds.5

4 Nevertheless, it is still small compared with the investment-grade corporate bond segment.5 See International Monetary Fund, Global Financial Stability Report, April 2004.

Chart D Yield spreads against AAA-rated corporate bonds

(in bp)


Euro area corporate bond indices (investment grade): yieldspreads against euro area AAA-rated corporate bond index,by rating classes

Euro high yield indices: yield spreads against euro areaAAA-rated corporate bond index, aggregate and by ratingclasses

1999 2000 2001 2002 2003

AA-ratedA-ratedBBB-rated

0

50

100

150

200

250

0

50

100

150

200

250

1999 2000 2001 2002 2003

aggregateBB-ratedB-ratedCCC and lower

0

1,000

2,000

3,000

4,000

5,000

0

1,000

2,000

3,000

4,000

5,000

29ECB


3 MARKETOVERVIEWimportance of liquidity for the corporate bond

market. He showed that issuance is morestrongly curtailed in the case of a liquidityshock than after credit quality shocks. Thestudy also pointed to the importance of liquidityfor the composition of the bond market. Ifliquidity is restricted, investors emphasise thesize and “familiarity” of issues, and so forsmaller and less prominent companies marketaccess becomes difficult. The proportion oflonger maturities of newly issued bonds rosedistinctly during the period from 2001 to 2003.Obviously the issuers tried to take advantage ofthe comparably favourable financing conditionsin recent years. In the first half of 2004 therewere also a lot of buybacks and bond exchangesin the corporate bond market. In most cases theaim of these activities was to issue bonds witheven longer maturities in order to lock in thelow interest rate level. In 2003 more than 80%of the newly issued corporate bonds were fixedrate coupon bonds. This proportion has beenlargely unchanged in recent years.

The percentage of issuers not residing in theeuro area (most of which are based in the UnitedStates and the United Kingdom) is, at 40%, veryhigh in the corporate bond segment. It is worthnoting that corporate bonds are often issued atinternational financial centres. This shows thatintegration to form a truly single euro areacapital market is still far from being a reality.The most important banking locations wherecorporate bonds are issued are London andLuxembourg. In the case of bank debtsecurities, the percentage of offshore issues ismuch lower since they are still generally subjectto significant national legislation, especiallythat governing the issuing of Pfandbriefe.Legislators in individual countries haveaccordingly been particularly active, especiallyin their quest to create a legal framework for theissuing of covered bonds based on the GermanPfandbrief.


In the past few years, the stronger competitionowing to the introduction of the euro hasaccelerated the process of market infrastructurereshaping and has involved trading, clearingand settlement stages. The different componentsof the financial market places developed newservices and rationalised ownership structures.Strong synergies, the need to lower costs andthe drive to strengthen the position of the mainmanagement companies spurred integrationbetween the trading platforms and thesettlement systems.

THE TRADING SECTOR

THE PRIMARY MARKETBetween 2001 and 2003 trends in the evolutionof the primary market infrastructure includedmore extensive use of the internet andwidespread recourse to syndication ingovernment securities issuance. Thedevelopment of electronic systems – whichenhance speed, reliability, transparency andcost-effectiveness, contributing to securing thecheapest borrowing terms – was less importantthan in the secondary market, principallybecause most European countries had alreadyfully automated government debt issuanceprocesses.

At present, almost every Member State has anelectronic auction or tap issuance system; theremaining countries are planning to implementsuch a procedure in the near future (forexample, the United Kingdom, Czech Republicand Slovakia). Moreover, in the past three yearssome countries improved their auction systemsin order to handle different types of operations(e.g. buy-back and exchange operations), toreduce the time between the auction cut-off andthe awarding of bids (the “awarding time”,which has become an important parameter ofcompetition among sovereigns)14 and to makeaccess easier for non-resident operators byreducing some of the technical requirements forparticipation.

In the period under review, syndication wasused – especially by the smaller countries – in

4 MARKE T I N F R A S T RUC TUR Eorder to achieve a high initial outstandingvolume and thus rapidly build up liquidity in aparticular issue. However, bigger countries alsohave begun to use this placement method morefrequently in order to reduce risks related to theissuance of innovative products or, moregenerally, when major uncertainty exists aboutinvestor demand.15 In fact, such a placementsystem helps to remove much investorreluctance by means of reduced priceuncertainty and outright provision of highliquidity.

In the syndication process, electronic systems –in particular the internet – are used in severalways: issuers and lead managers16 candisseminate information quickly to a largenumber of potential investors; investors candirectly enter bids into the system through thesales representatives; electronic book-buildingprovides the issuer and the syndicate with real-time information on the bidding process, thusenhancing pricing transparency. Finland,Portugal and Spain have used the internet as atool in syndicate transactions, but theirexperience indicates that, although the investorshave the technical facilities needed to enter bidsdirectly into the system, contact with the salesteam continues to be very important.

The direct placement of government securitieswith retail investors, used by Spain andSweden, although technically possible, has alsoremained limited, principally for reasonsrelating to the high cost and the limitedknowledge about this investment alternative.

THE SECONDARY MARKETThe euro denominated bond secondary markethas been characterised by the growing use ofmultilateral electronic trading systems, related

14 The shortness of the awarding time reduces the risk for bidders,as they are informed more quickly about their positions. As aresult, the probability of unexpected moves in market prices isalso reduced.

15 Italy, for example, has been using syndication for initial issues ofindex-linked and ultra-long bonds.

16 The lead managers are a pool of primary dealers active in thedomestic market, which is selected by the issuer, and in charge ofproviding the demand.

31ECB


4 MARKETINFRASTRUCTURE

to their benefits in terms of lower costs, higherliquidity, transparency and easier cross-bordertrading. The trend was clearly visible in themore homogenous and liquid government bondsector.

In the first phase, because of the large size ofthe OTC bond market, competition amongtrading markets caused the proliferation of newtrading platforms, these taking the form ofregulated markets and alternative tradingsystems (ATSs).17 Consolidation initiativeshave begun to emerge only recently,accompanied by moves away from the use ofsmaller less successful trading platforms.18

Platform vendors have contributed toencouraging the adoption of electronic tradingby adding or enhancing features and services intheir trading systems such as connectivity forautomated trade processing, pre-trade servicessuch as research, analysis and calculationsoftware, post-trading linkages and servicestailored to the syndicated underwriting of newissues.

In its last review of electronic transaction systemsfor fixed income markets, the Bond MarketAssociation identified 77 systems operating in theUnited States and Europe in late 2003, as opposedto 11 in 1997. Of these platforms, 31 were basedprincipally in Europe.19 All the systems identifiedsupported the government bond market.However, over the past three years, due tocompetition, platform vendors and traders havealso accelerated their adoption of electronicexecution for less liquid products such ascorporate debt securities, asset-backed andmortgage-backed securities.20 The latter trendwas less significant for inter-dealer platforms. Infact, in relationships between intermediariesconcerning less liquid financial instruments,traditional trading channels continue to belargely predominant, despite the existence ofdevoted segments on several platforms. Onthe basis of the BMA survey and informationfrom European central banks, Table 4 lists theprincipal electronic bond trading systems usedin the EU.21

Among dealer-to-dealer platforms, MTSand Icap/BrokerTec are the most widespread;other important platform are Eurex Bond andE-Speed.

In particular, MTS Group, the first wholesaleelectronic market in the euro area, reinforcedits leadership in government securitiestransactions and promoted the integration of theeuro denominated bond market by broadeningthe range of securities traded and servicesoffered and by exporting its platform to otherEuropean countries: between 2001 and 2003 thenumber of national markets using the MTSplatform rose from 5 to 12. Among the new EUMember States, Poland announced in October2003 the implementation of a domesticgovernment bond market based on the MTSsystem. Like MTS Italy, all the other domesticMTSs are quote-driven and based on theobligation for market makers as a whole toquote two-way prices during the day. The rulesof each market, however, such as conditions ofaccess, obligations of market-makers, list oftraded securities etc are established byshareholders in accordance with national law.Because all national MTS use the same platformas MTS Italy, participants in the market canemploy the same workstation to access all theMTS markets they have joined, thus benefitingfrom economies of scale. Moreover, some MTS

17 According to the FESCO def inition, an ATS is “an entity which,without being regulated as an exchange, operates an automatedsystem that brings together buying and selling interests – in thesystem and according to rules set up by the system’s operator – ina way that forms, or results in, an irrevocable contract.” This isa broad definition which captures any trading functionalityregardless of whether the functionality operates bilaterally ormultilaterally.

18 In May 2001 the acquisition of the dealer-to-customer platformBondClick by MTS Spa, to create BondVision, was widely seen asthe starting point of a consolidation of the sector. Among thesubsequent events witnessing the new trend we can mention thetake-over of the British regulated market Coredeal Ltd by MTSSpa, with the subsequent integration of Coredeal and Euro MTSplatforms (2002), and the closing of Jiway, the regulated marketof OM Group (2002).

19 This does not include systems focused principally on retail orindividual investors.

20 At the end of 2003 the majority of the systems were platforms onwhich multiple types of bonds can be traded.

21 It must be underlined that the high number of platforms, theirheterogeneity and the dynamism of the sector makes it difficult togive an exhaustive picture.


markets (e.g. Belgium, Finland and Denmark)drew up an agreement for reciprocal access,which makes the operators of each marketdealers on the other markets, too.

Some MTS markets were provided with newfunctionalities and services. For example, since2003 the Italian, Finnish and Portugueseplatforms have allowed exchange operations tobe managed, and MTS Spain and MTS Portugalhave activated a segment dedicated to T-bills.

Other developments concerned EuroMTS, the“super-wholesale” element of MTS Group,based in London, where benchmark governmentbonds of several euro area countries are traded.In the past two years MTS Group, incollaboration with Euronext, started thediffusion of new bond indices on EuroMTS,EMTX and EMTXi, which respectively providea synthetic indication about the trend of thefixed income and inflation-linked governmentsecurities market in the euro area.

Finally, in 2003 a new platform, NewEuroMTS, was activated to trade euro-denominatedgovernment securities issued by the new EUMember States in order to promote the effortsof those countries towards integration into theeuro bond market. At present, nine bonds(issued by Poland, Hungary, Lithuania, theCzech Republic, Slovakia) are listed on theplatform.

The other principal dealer-to-dealer systemused in Europe, BrokerTec, is a global fixedincome platform founded in 1999 and recentlyacquired by ICAP, the world’s largest inter-dealer broker. The system providesparticipating dealers with execution andstraight-through processing. It allows thetrading of a large range of US fixed incomesecurities and European government bonds(Belgian, Dutch, French, German, Spanish andAustrian).

A number of euro area countries have seen thebeginning of electronic repo trading on dealer-to-dealer platforms, which are already quite

well developed in many Member States(especially on Icap/BrokerTec and MTS Italy).In Germany, Eurex Bonds introduced a repotrading facility via the “euro repo” web-basedplatform; in Spain a repo segment was activatedon MTS Spain. According to ISMA,22 togetherwith the development of European repo market,the share of electronic trading continued togrow steadily over the past three years,reaching 20% of the total value of repocontracts at the end of 2003. Electronic tradingis not yet widely used for swaps, despite theexistence of two platforms (E-MIDER and AT-FOX), owing to the fact that swap operationsare not sufficiently standardised.

The application of electronic trading torelationships between intermediaries andinstitutional investors has also showedsignificant developments in recent years, withprogressive diffusion and the use of dealer-to-customer platforms, of which TradeWeb andBondVision are the most important.23

In recent years the supervisory authorities havecontinued to examine the implications of thegrowing use of electronic trading systems forthe structure and functioning of the financialmarket. From the supervisory perspective,crucial aspects are: (i) the risk of liquidity beingsplit up between different platforms, whichwould undermine efficient price formation;(ii) the tendency of intermediaries to carryout in-house netting of orders of oppositesigns received from customers (known asinternalisation); (iii) the difficulty ofdistinguishing clearly between regulatedmarkets, other trading platforms and trading

22 International Securities Market Association, European repomarket survey No 6, conducted December 2003, published inMarch 2004, p 6.

23 TradeWeb is an auction system which links 24 f ixed-incomesecurities leaders with more than 1,500 buy-side institutions inNorth America and Europe. Products traded on the platforminclude different types of fixed-income securities, bothEuropean and US. BondVision is a regulated market, launched byMTS Group in August 2001, which allows primary dealers totrade euro government securities over the internet directly withinstitutional investors by means of a competitive auction. Theplatform, used in many EU Member States, has registeredsignificant growth regarding both operators and transactionssince its launch.

33ECB



activity performed by individual intermediaries;and (iv) the disaster recovery capabilities toguarantee the continuity of services.

In July 2002 the Committee of EuropeanSecurities Regulators (CESR) published the“Standard for Alternative Trading Systems”,under the Investments Services Directive(ISD), aiming to ensure the protection of boththe users of ATSs and the integrity of themarket.

In April 2004, the ISD was amended by theadoption of the Financial Instruments MarketsDirective (FIMD). For the first time, takinginto consideration the consequences of theATS development, a Directive establishesa comprehensive regulatory frameworkgoverning the organised execution of investors’transactions by regulated markets, other tradingsystems (called multilateral trading facilities orMTFs24), and banks and investment firmspractising systematic internalisation. This newenvironment ensures adequate transparency inprice formation, whatever the venue,establishing pre-trade transparency obligations,which, however, are limited to equities. Theextension of transparency obligations to bondscan be adopted by individual EU Member Statesfor their investment firms and is to bediscussed, as a general rule, within two yearsfollowing approval of the Directive. The impactof the new Directive on the bond market willprobably depend on that decision.

THE POST-TRADING SECTOR

The evolution of European post-trading systemshas continued to be directed towards theimprovement of stability and the search forhigher levels of operational efficiency.

In this context, clearing and settlement systemshave been pushed towards a common paradigmcharacterised by increasing integration ofservices providers, stronger diffusion ofschemes for settlement in central bank money,more extended use of central counterpartyfunctions, and activation of new services such

as greater use of services such as securitieslending and guarantee management.

THE SETTLEMENT SIDEThe search for efficiency gains has reinforced adynamic movement towards consolidationamong EU settlement providers. Progress inthis process has been achieved, both in the formof structural changes and strategic measures.Consolidation has involved mergers ofinstitutions providing similar services(“horizontal consolidation”) and mergers ofinstitutions providing different but integratedservices (“vertical consolidation”).

In the period under review, at a local level,reforms towards horizontal consolidation wereadopted in Spain, where the two centraldepositories, SCLV (for equities and corporatebonds) and CADE (the book entry system forpublic debt), were unified to establish a singlecentral securities depository (CSD) for allinstruments (Iberclear). In Italy the nationalcentral depository (Monte Titoli) extended itsfunctions to settlement activities, which hadbeen managed by the central bank until 2003.In the United Kingdom the settlement ofgovernment bonds was integrated into thesecurity settlement system for equities andbonds, CREST, in 2000. The settlement ofmoney market securities was integrated intoCREST in 2003.

At the EU level, horizontal consolidation,which had already started when ClearstreamInternational25 and Euroclear Group26 were set

24 An MTF is defined as a multilateral system, operated by aninvestment firm or by a person who manages the business of aregulated market (the “market operator”), which brings togethermultiple third-party buying and selling interests in financialinstruments – in the system and in accordance with non-discretionary rules – in a way that results in a contract inaccordance with the provisions of Title II of the Directive. Froma micro structural point of view, the definition is less broad thanATS, as it excludes bilateral systems.

25 Clearstream International is the holding company for theinternational central securities depository (ICSD) ClearstreamBanking Luxembourg and for the German CSD ClearstreamBanking Frankfurt, the result of the merger between Cedel andDeusche Börse Clearing.

26 The group originated from the merger of Euroclear and thenational CSDs SICOVAM (France) and later on NECIGEF (TheNetherlands).


up, continued with the merger of CREST withthe Euroclear Group in September 2002.

At the beginning of 2004 another initiative wasthe signing of a letter of intent between theFinnish and Swedish CSDs (APK and VPC,respectively) concerning their merger to create ajoint CSD group within the Nordic area(NCSD). The creation of a Finnish-SwedishCSD is aimed at facilitating the harmonisationof rules and settlement processes and promotingthe development of a common technologyplatform in order to increase the efficiency andhence the competitiveness of the Nordic regionas a financial market.

As far as vertical consolidation is concerned,similar interventions were implemented (i) inGermany and in Italy, where the companiesmanaging the stock exchanges (Deutsche BörseAG and Borsa Italiana SPA) became the main orexclusive shareholders of the national clearingand settlement systems; and (ii) in Spain, wherea single holding company Bolsas y MercadosEspanoles (BME) was incorporated to integrateboth the Spanish stock markets and the post-trading systems.

At the EU level, Deutsche Börse AG took overClearstream International completely in 2002.Thus at present Deutsche Börse constitutes awholly integrated securities processing chaincomprising trading, clearing and settlement,thereby maintaining a high degree ofinteroperability of its diverse trading systemsand of its CSDs with other CSDs.

Regarding technical innovations in procedures,new settlement models were introduced inGermany and in Italy in 2003. Basic features ofthe German procedure are prefunding whichresults in the elimination of unwinding risks aswell as earlier finality in the morning beforeTARGET opening. The new Italian system isaimed at combining the advantages of net andgross settlement, moving forward thetransaction settlement to the early morning ofthe settlement day, reducing liquidity needs andpermitting the efficient use of collateral in

respect of intraday credit from the central bank.In the United Kingdom the settlement system,CREST, moved to full delivery-versus-payment(DVP) in central bank money in 2001. Asnoted, the integration of the settlement ofmoney market securities into CREST took placein 2003. In Sweden, the securities settlementsystem (VPC) made a substantial change in itssettlement procedures. In November 2003 VPCreplaced its net settlement procedures withgross settlement combined with automaticcollateralisation thereby eliminating the risk ofunwinding.

Finally, many countries (e.g. Greece, Poland,Slovakia, Malta, Hungary) are taking (or havealready taken) various steps, such us improvedlinkages and intra day multi-batching asopposed to end-of-day processing, which willallow the shortening of the trade settlementcycles and the introduction of real-time grosssettlement and DVP.

THE CLEARING SIDEDemand for clearing and central counterpartyservice have developed rapidly in Europe in thepast few years as a means of reducingoperational and credit risks, enhancingefficiency in the usage of capital and loweringtransaction costs. The demand was partlyrelated to the increasing use of electronictrading platforms with trader anonymity. Inaddition, the greater use of centralcounterparties is often cited as a factor able toreduce the heterogeneity of national systems.Moreover, larger efficiency gains would occurif further cross-border consolidation of centralcounterparties were to take place.

In this field, the initiatives to improve theefficiency and stability of systems includedvarious merger combinations among servicesproviders and the gradual extension of centralcounterparty activities to the cash market, bothfor equities and bonds.

The most important initiative for theconsolidation of the European marketsinfrastructure was the merger in December 2003

35ECB



of two of Europe’s leading central counterparties,London Clearing House Ltd of the UK andClearnet SA of France, to form the LCH.ClearnetGroup (“LCH.Clearnet”). The combined rangeand penetration of the two companies makes thegroup the largest and most diversified centralcounterparty in Europe. The group is planning tofurther expand its bond market coverage. Theharmonisation of the technical platforms of LCHand Clearnet, which is already under way, willfacilitate an expansion of cross-border servicesand the group has stated publicly that it is open tofurther mergers.

Eurex Clearing and Cassa di Compensazione eGaranzia, the central counterparties for theGerman and Italian domestic derivativesmarkets respectively, launched centralcounterparty activities on the cash equitymarket in 2003. In addition, Cassa diCompensazione e Garanzia signed an agreementwith Clearnet to establish a central counterpartyservice for Italian government transactions onMTS Italy.

Eurex Clearing is going to expand its businessto include OTC bonds and repos in the nearfuture. In Spain, a new company, Meffclear,was created in 2003 for the purpose of clearingpublic debt transactions.

INTEGRATION OF EUROPEAN POST-TRADINGSYSTEMS AND REGULATORY INITIATIVESEven if European clearing and settlementsystems consolidated on a national basis andreached significant progresses in theirfunctionality, the rather limited process ofintegration developed at the EU level, and theconsequent fragmentation of post-tradinginfrastructures, still represents an obstacle tocross-border transactions, due to theircomplexity and high costs.

In recent years the authorities and marketparticipants have stepped up their efforts toestablish principles, common methods andstandards to promote competition, efficiencyand security in the provision of post-tradingservices.

Two reports, known as the Giovannini reportswere published in 2001 and in 2003, with theaim of identifying inefficiencies in EU clearingand settlement arrangements and devising astrategy to eliminate them. The first reportidentified 15 barriers as the sources of thoseinefficiencies, based on market practice/regulatory requirements, tax procedures andissues of legal certainty. The second reportproposed a strategy for the removal of the 15barriers, based on an appropriate sequencing ofactions, a clear allocation of responsibility(between private sector agents and publishauthorities) and realistic deadlines for eachaction (consistent with the deadline set for thefull implementation of the FSAP). The reportalso analysed the possible integration modelsfor EU clearing and settlement systems,showing that alternative structures couldemerge in the consolidation process and that, atthis stage, an ex-ante assessment of thesestructures does not allow preference to be givento a particular solution.

In 2003 another important contribution camefrom the Joint Working Group set up by theEuropean System of Central Banks (ESCB) andthe Committee of European SecuritiesRegulators (CESR) which, on the basis of theG10-IOSCO recommendations, published aconsultive paper in order to define commonEuropean standards for stability, efficiency,transparency and investor protection to guidesupervisory authorities, CSDs, SSSs, centralcounterparties and other institutions thatprovide similar services.

The final report, titled “Standards for securities,clearing and settlement in the European Union”was approved by the Governing Council and theCESR in October 2004.

From a legal point of view, a significant step isrepresented by the Collateral Directive. TheDirective aims to help to create a clear anduniform pan-EU legal framework for the use ofcollateral to limit credit risk in financialtransactions, lowering credit losses andencouraging cross-border business.


Dailyaverage

turnoverin EUR

System/ Managing User Regulated Type of Type of System millionsMarket name company country market1) trading 2) operations type3) (2003)4)

AIAF Mercado de Renta Spain Yes Wholesale Cash/Rep D2D 8Fija S.A.

AUTOBAHN Deutsche Bank Germany No Wholesale Cash D2C N/A

BLOOMBERG Bloomberg L.P. EU No Wholesale/Retail D2C N/A

BOND VISION MTS Group EU Yes Wholesale Cash D2C 680

BUDAPEST STOCK Budapest StockEXCHANGE (MMTS) Exchange Ltd Hungary Yes Retail Cash D2D N/A

MTS-POLAND MTS CETO SA Poland No Wholesale/Retail Cash/Rep D2D 173

EBOS II Bratislava Stock Slovakia Yes Wholesale/Retail Cash/Rep D2D, 107Exchange D2C

ESTONIAN CSD Hex Tallin Group Estonia Yes Wholesale D2D N/A

E-SPEED Cantor Fitzgerald EU No Wholesale/Retail Cash/Rep D2D N/A

EUREX BONDS Deutsche Bourse A.G. Germany Yes Wholesale Cash/Rep D2D 1,000

EUROMOT Borsa Italiana Spa Italy Yes Retail Cash D2D 15

EUROMTS MTS Group EU No Wholesale Cash/Rep D2D 2,400

EURONEXT LISBON Euronext Group Portugal Yes Retail Cash D2C 5

HDAT Bank of Greece Greece Yes Wholesale Cash D2D 2,725

ICAP/BROKERTEC Icap EU No Wholesale Cash/Rep D2D N/A

INSTINET Reuters EU No Wholesale/Retail D2D, N/AD2C

MARKET AXESS Market Axess Ireland No Retail Cash D2C N/A

MATS Malta Stock Exchange Malta No Retail Cash D2D 1

MOT Borsa Italiana Spa Italy Yes Retail Cash D2D 566

MTS AMSTERDAM MTS Group Netherlands No Wholesale Cash D2D 500

MTS ASSOCIATED MTS Group Belgium No Wholesale Cash D2D 593MARKET Denmark 500

MTS AUSTRIA MTS Group Austria No Wholesale Cash D2D 115

MTS FINLAND Division of MTS Finland No Wholesale Cash D2D 250

MTS FRANCE MTS Group France No Wholesale Cash D2D N/A

MTS GREECE MTS Group Greece No Wholesale Cash D2D 220

MTS GERMANY MTS Group Germany No Wholesale Cash D2D 870

MTS IRELAND MTS Group Ireland No Wholesale Cash D2D N/A

MTS ITALY MTS Group Italy Yes Wholesale Cash/Repo D2D 8,400

MTS PORTUGAL MTS Group Portugal Yes Wholesale Cash D2D 503

MTS SPAIN MTS Group Spain No Wholesale Cash/Repo D2D 724

SAM Bourse de Louxemburg Luxembourg Yes Retail D2C 1.47

SAXESS OMX Exchanges Sweden Yes Wholesale Cash D2D N/A

SENAF Senaf S.A. Spain Yes Wholesale Cash/Repo D2D 1,002

TLX TLX S.P.A. Italy Yes Wholesale/Retail Cash D2D 46

TRADE WEB Thomson Financial EU No Wholesale/Retail Cash D2C N/A

TRADING SYSTEMOF THE RIGA STOCKEXCHANGE (AS400) Riga Stock Exchange Latvia Yes Retail Cash D2D 1

Table 4 Principal electronic systems for bonds used in Europe

Sources: Bond Market Association; Bond study questionnaires; web sites.1) As defined in the Investment Services Directive and Financial Markets Instruments Directive.2) Wholesale/retail.3) Dealer-to-dealer system (D2D)/dealer-to-customer system (D2C).4) Cash operations only.

37ECB



Moreover, in April 2004, the EuropeanCommission, taking into account the describedreports and initiatives, adopted a Communicationwhich set out, for the first time, overallCommission policy on this subject and presentedpossible courses of action to improve thecross-border post-trading environment. TheCommunication proposes the preparation of aframework directive on clearing and settlement bylate 2005.

Another interesting development is the HagueConvention on the law applicable to certain rightsin respect of securities held with an intermediary.This international treaty determines the exactjurisdiction of securities in custody. After theratification of this treaty, the law of the country ofthe intermediary will apply, in line with the placeof the relevant intermediary approach (PRIMA),and no longer the law of the country of the issueror of the CSD or of the owner of the securities.


Various trends observable in the developmentof the euro bond market. This chapter focuseson innovative instruments and technologiestaking centre stage in this process.

5.1 COVERED BOND MARKET

The European covered bond market haswitnessed interesting developments in recentyears. While the issuance of covered bondsdeclined until 2001, mainly due to the sharpreduction in the issuance of GermanPfandbriefe, a recovery started as of 2001,bringing the absolute issuance volume back tothe 1999 levels of nearly €250 billion in 2003.The total volume of covered bonds outstandingin the EU at the end of 2003 surpassed €1.5trillion. However, apart from the rising volumessince 2001 as well as the continuous productinnovations, the interesting feature in thismarket segment is the growing share ofissuance from European countries other thanGermany, whose covered bonds onetheless stilldominate the market. Not only are well-established covered bond markets showingincreased issuance, but new markets have alsobeen developed or are about to come into being.This is accompanied by the modernisation ofexisting covered bonds legislation in severalcountries, while other countries have adopted acovered bonds legislation, or will soon do so.These developments show the dynamism in thecovered bond market in a pan-European context.

DEFINITION AND TYPES OF COVERED BONDS

Covered bonds are used by banks to refinanceloans secured by mortgages or loans to thepublic sector. In this respect, covered bonds canbe defined as full recourse debt instrumentssecured (covered) by collateral pools, namelymortgage assets and/or claims against publicsector entities. From the issuer’s point of view,the covered bonds form part of the liabilities ofthe institution. They are linked to a certainamount of cover assets, but normally not to aspecific set of assets. In contrast to asset-backed and mortgage-backed securities, the

5 MARKE T T R END Scover assets remain on the balance sheetof the originator (or, in a few cases, theyare segregated through the transfer to aseparate entity, for example in the form ofa limited liability partnership). Coveredbonds, therefore, constitute “on-balance-sheetsecuritisation”. Since covered bonds are fullrecourse creditworthiness of the issuer. As aconsequence, the rating of the issuer is astarting point for the rating to be attached to thecovered bond. However, the range of differenttypes of covered bonds is wide, and some typescome close to resembling structured financeproducts as in some cases securitisationtechniques are used in order to achieve greaterde-linkage from the issuer/originator.

In general, a distinction can be made betweenthe following three types of covered bonds.First, regular, or “plain vanilla” covered bonds,for which asset quality, cash flow adequacy andcounterparty risk are determined by a legalframework. Second, structured covered bonds,which are structurally enhanced in order tofurther reduce credit risk and, by de-linkingfrom the fundamental credit rating of the issuer,achieve a higher rating. This credit enhancementcan, for example, consist in a higher thanlegally required over-collateralisation, i.e. thecover assets exceed the issued covered bonds.Finally, the third type of covered bonds mightbe called “replicated” covered bonds, meaningthat the covered bond framework is determinedby private law, as a legal framework for coveredbonds does not exist in all countries.

THE EUROPEAN COVERED BOND MARKET

Issuance of covered bonds represents around15% on average of the gross issuance of euro-denominated bonds, corresponding to around€220 billion per year on average since 2001. Arather substantial increase in issuance activitywas observed in 2003. Overall, the amountoutstanding of covered bonds was over €1,550billion at the end of 2003. Covered bonds havebeen mainly issued with maturities between twoand ten years. Around 80% of total issuance isfixed rate coupons. Around 65% of the covered

39ECB


5 MARKETTRENDS

bonds are rated AAA. The share of non-ratedbonds remains relatively high at 15%, due to theshare of smaller-sized German Pfandbriefe.

German Pfandbriefe account for by far thelargest share of covered bonds, although theshare has declined.27 Since 1999, and inparticular since 2001, the decline in issuance ofGerman Pfandbriefe is, at least partly, offset byan increasing issuance of covered bonds inother European countries. These developmentsunderline the dynamism in the covered bondmarket in a pan-European context.

Apart from Germany, the most noteworthyissuance of covered bonds – mainly related tothe well-established mortgage markets in therespective countries – is found in Denmark andSweden28 and in France and Spain. There hasalso traditionally been a relatively sizeableissuance of covered bonds in Austria andLuxembourg. Finally, Ireland launched theissuance of covered bonds at the start of 2003and the United Kingdom in July 2003. Whereasthe other countries mentioned have a legalframework for covered bonds, issuance in theUnited Kingdom is carried out under a privatelegal structure based on UK common law andcontract law, i.e. these covered bonds are anexample of “replicated” covered bonds.29

Developments with respect to a covered bondframework – either to enhance or introducecovered bonds legislation or to follow theexample of the United Kingdom – are presentlyunder way in Italy, Belgium, the Netherlandsand Portugal.30

SUPPLY FACTORS: THE ISSUER SIDE

Covered bonds are a refinancing instrumentused by credit institutions to fund theirmortgage and/or public sector loans. Given theimportance in terms of the size of its mortgageloan business in particular, a substantial part ofthe asset side of a bank’s balance sheet requiresthe best means of funding. In view ofdisintermediation trends, issuance of coveredbonds via the capital market as a means ofraising funds offers an alternative to retail

funds, i.e. deposits. Furthermore, coveredbonds offer the possibility of extending thematurity profile of the liability side of a bank’sbalance sheet. With mortgage loans generallyhaving a long-term orientation, covered bondsenable the credit institution to better balance thematurity profile between the asset and liabilitysides of its balance sheet, which is alsodesirable from a more general financial stabilitypoint of view. In the effort to maximise returnson equity, another advantage offered by coveredbonds is their lower funding costs compared toasset-backed securities. While the creditinstitution generally remains exposed to thecredit risk, market risk, prepayment risk and

27 In this respect, implications might also arise from the abolition ofthe German Landesbank guarantee mechanisms as of July 2005.

28 It is noted that only part of the overall Danish and Swedishcovered bonds issuance is denominated in euro.

29 The further development of the UK covered bond market willalso be influenced by the final opinion of the UK FinancialServices Authority, which signalled at the end of August 2004that it might restrict covered bond issuance to 4% of a bank’s totalassets.

30 Germany: Pfandbriefe (Hypotheken- and ÖffentlicherPfandbrief), Denmark: realkreditobligation, Greece: ktematekesomologies, Spain: cédulas hipotecarias/cédulas territoriales,France: obligation foncière, Ireland: covered bond, Italy:obbligazione fondiaria, Luxembourg: lettre de gage, Austria:Pfandbriefe (Hypotheken- and Öffentlicher Pfandbrief),Portugal: obrigações hipotecárias, Netherlands: pandbrief,Finland: hypoteekkilaina, Sweden: bostadsobligation, UnitedKingdom: covered bond.

Chart 8 The European covered bond market

(%)

Sources: European Mortgage Federation; own calculations.

DK12

DE74

ES3

LU1

IE0.5

AT0.5

SE4

UK0.5FR

4.5


other risks inherent in the underlying mortgageloans, covered bonds offer cheaper financingbecause the bonds are covered not only by theunderlying mortgage (and public sector) loans,but also by the credit institution’s reserves andother own funds. Lower funding costs are alsorelated to the fact that covered bonds oftenenjoy a higher liquidity than asset-backedsecurities (in particular in the case of GermanJumbo Pfandbriefe and, due to increasingissuance size, also in other countries). Ingeneral, as covered bonds so far only representa fraction of the overall amount of mortgageloan and public sector financing, growthpotential for covered bonds will prevail forsome time to come.

DEMAND FACTORS: THE INVESTOR SIDE

From an investor’s point of view, coveredbonds are based, as a minimum, on thecreditworthiness of the issuing creditinstitution. However, they involve theadditional security of a claim on the underlyingmortgage or public sector loan, as well aseither a strict legal framework or contractuallyfixed credit enhancements. This makes coveredbonds an attractive high-quality investmentopportunity. Furthermore, investors seekalternative ways to invest their capital, whichis also offered by covered bonds. Morespecifically, covered bonds present anadditional instrument for asset classdiversification and, in view of the issuanceactivity in several countries, also forgeographical diversification (within and outsidethe euro area) of the investment portfolios ofinstitutional investors. In a more generalcontext, this is also fostered by the trendtowards savings disintermediation, i.e.households’ orientation towards capital marketinvestments with horizons longer thantraditional bank deposits, a development whichis induced by, inter alia, both the demographicsof an ageing population and the necessity toundertake private pension investments.

CURRENT DEVELOPMENTS

Apart from the mentioned numerous initiativesin several countries regarding covered bondslegislation or the creation of new covered bondmarkets two further aspects relevant for futuredevelopment can be highlighted.

First, implications for the future developmentof the European covered bond market arise fromthe transposition of Basel II through the newCapital Adequacy Directive in the Europeancontext. At present, the only legislativereference to covered bonds is contained inArticle 22 (4) of the Directive on undertakingsfor collective investment in transferablesecurities (the UCITS Directive). The CapitalAdequacy Directive in turn makes reference tothe relevant provisions contained in the UCITSDirective. Today, the existing Capital AdequacyDirective stipulates as a general rule thatsecurities in accordance with Article 22 (4) ofthe UCITS Directive may be risk-weighted at10% at the discretion of national supervisoryauthorities. The European Commission workedon the definition of covered bonds and the risk-weighting in the new Capital AdequacyDirective, both under the internal ratings-basedapproach as well as under the revisedstandardised approach. The outcome, also bycomparison with asset-backed (and mortgage-backed) securities, will influence the futuredevelopment of the European covered bondmarket.31

Second, issuance activity in many Europeancountries, which is accompanied, as recalledabove, by numerous initiatives regardingcovered bonds frameworks (legislation orprivate structures), might also present investorswith a certain difficulty. Due to the fact thatthere is no harmonised pan-European coveredbond framework, investors need to keepthemselves informed about the various issuance

31 The draft new Capital Adequacy Directive was released on14 July 2004. The EU Commission would like to introduceprivileged treatments for covered bonds. The implementationdate for the proposed Directive is foreseen for end 2006, with atransition period by 2007.

41ECB


5 MARKETTRENDS

structures, which might also require knowledgeof local features. In this respect, an initiative ofthe European Mortgage Federation can bementioned. The European Mortgage Federation(which at present represents around 70% of themarket capitalisation of European coveredbonds) announced on 1 July 2004 the creationof a “European Covered Bond Council”, i.e. aEuropean platform for covered bonds thatshould bring together covered bond issuers,credit analysts, investment bankers, ratingagencies and other interested marketparticipants. This Council, officially to belaunched in November 2004, will focus on thedevelopments shaping the economic andregulatory environment of covered bonds at theEU level.

5.2 SECURITISATION MARKET

Off-balance-sheet term securitisation did nottake off in Europe until the late 1990s. It hasseen impressive growth rates since then and hasnow become an established asset class in theEuropean fixed income market (see Chart 9).According to Moody’s,32 total (funded andunfunded) issuance volumes rose to €268billion in 2003 and expectations are thatsecuritisation issuance will continue to growand even outstrip corporate bond issuance in2004.

KEY ASSET CLASSES AND INSTRUMENTCHARACTERISTICS

The goal of off-balance-sheet securitisation isto ensure that the credit quality of an asset-backed security (ABS) is based solely on thequality of the assets and the credit enhancementbacking the obligation, without any regard tothe originator’s own creditworthiness. For thispurpose, the originator or owner of the financialassets transfers those assets to a bankruptcy-remote special purpose vehicle (SPV). Adifference is made between cash structures,whereby the net proceeds from the notes issuedby the SPV are used to purchase the pool ofassets33, and synthetic structures, that used

credit risk derivatives to achieve the redit risktransfer from the risk shedder to an SPV.Synthetic structures can be both “funded”through the issuance of credit-linked notes,whereby the risk taker has to provide upfrontfunding in the transaction, or “unfunded” ifthey rely on credit default swaps. In the lattercase funding is provided by the risk taker onlyupon occurrence of a credit event. In Europe thede-linkage of the credit risk of financial assetsfrom that of the originator has been an effectiveway to bring the demand for and the supply ofhighly rated capital market instrumentstogether. Around 80% of all Europeansecuritisation issuance has been rated AAA.

Chart 9 European term securit isation

(volume in EUR billion)

Source: Moody’s.

0

50

100

150

200

250

0

50

100

150

200

250

2001 2002 2003

fundedunfunded

32 Issuance data includes rated true-sale transactions as well asfully and partially funded synthetic transactions of assetsdomiciled in Europe and issued by European originators. In thecase of the latter, only rated super senior swaps or credit defaultswaps are included. Since the statistical framework for off-balance-sheet securitisation has not yet been established and themarket is very much a rated market, rating agencies have so farbeen the main data source. Moody’s has the broadest datacoverage among the rating agencies since it includes not onlypublicly rated transactions in its issuance volumes, but alsoprivate placements rated by Moody’s. However, since unratedsynthetic structures are not captured by the data, volumesunderestimate the actual credit risk transfer that has taken placevia securitisation structures.

33 This needs to be conducted in a manner that results in a “truesale”, i.e. removing the assets from the bankruptcy or insolvencyestate of the originator.


In terms of asset classes, the residentialmortgage-backed securities (RMBS) market isthe most established and dynamic marketsegment, accounting, at €133.5 billion,for around 50 % of the overall Europeansecuritisation market in 2003 (see Chart 10).There is a trend towards a bigger issuance size,fungible issues and multi-currency deals inorder to meet foreign investor demand, andsecondary market liquidity has started toimprove. Those trends are most pronounced inthe United Kingdom and the Netherlands buthave started to surface also in other EuropeanRMBS markets, in particular in Spain and Italy.With the exception of Germany, true-saletransactions dominate in Europe. The secondbiggest asset class is collateralised debtobligations (CDOs) with issuance volumes of€70.9 billion in 2003. The European CDOmarket is essentially synthetic (92% syntheticissuance in 2003 according to Moody’s). Themarket started off with balance sheet CDOs,driven by regulatory capital arbitrage, but todaysynthetic arbitrage CDOs outpace balance sheetCDOs. In terms of issuers, the CDO market hasbeen dominated by large, internationally activebanks. One of the main attributes of theEuropean CDO market has been its rapid abilityto innovate. Recent developments includethe shift towards single-tranche CDOs, acontinuous expansion of CDO collateral typesand the development of credit indices thatpermit highly standardised trades. Non-mortgage based ABSs form the third biggestasset class (€37.9 billion issuance volume in2003). Consumer assets are the main underlyingasset class (46%), driven by multiple issuancefrom UK credit card master trusts. Public assettransactions rank second, dominated bytransactions of the Italian Treasury, which hasbeen the most active European government userof securitisation techniques. Other asset classesinclude commercial mortgage-backed securities(CMBSs) and more exotic asset classes such asinventories, future cash flows or whole-business securities (WBS), non-performingloans, private equity and project finance, whichare mainly a feature of the UK market.

Typically, “pass-through” structures are usedfor assets such as mortgage loans, consumerloans and trade receivables, whereby periodicpayments are passed through to the investor. Asa consequence of the early amortisation featuresin those deals, prepayment risk can besubstantial. In recent years, a growing numberof ABSs in Europe have been offered withfloating, rather than fixed, rates of interest.

LARGE CROSS-COUNTRY DIFFERENCES

With respect to the degree of development of thesecuritisation market there are still largedifferences between European countries. Oneexplanatory factor is that legal, regulatory, taxand accounting rules applicable to securitisationtransactions differ widely between variousEuropean jurisdictions. In certain common lawjurisdictions, such as the United Kingdom,different types of securitisation structures havebeen able to evolve without legislativeinterventions. In some other jurisdictionscharacterised by civil legal codes (France,Spain, Italy, Portugal), specific laws had to beadopted to allow the securitisation market todevelop. In jurisdictions such as Germany,legal and tax arrangements have long preventedthe evolution of true-sale securitisation,and transactions have been predominantly

Chart 10 European term securitisationvolumes by asset type in 2003

(volumes in EUR billion)

Source: Moody’s.

fundedunfunded

0

20

40

60

80

100

120

140

0

20

40

60

80

100

120

140

ABS CDO CMBS RMBS WholeBusiness

43ECB


5 MARKETTRENDS

structured synthetically to circumvent thosearrangements. In some countries (Greece,Luxembourg), specific securitisation laws havebeen introduced only recently, while in othercountries (Finland and the central Europeancountries) there is still a lack of a domesticsecuritisation law. However, the examples ofBelgium and Sweden show that a favourablelegal and tax framework alone is not sufficientto foster the development of the domesticsecuritisation market.

In addition, the driving forces behindsecuritisation may vary widely. So far issuersin Europe have mainly been banks, withcorporates accounting only for a smallpercentage of overall securitisation activity.Securitisation undertaken by banks can bemotivated by the desire to free up regulatory oreconomic capital, to reduce the cost of funding,to improve the risk management of the balancesheet by reducing the exposure to a specificsector or class of debtors, to achieve greaterdiversification in asset sources or to tap intonew investor groups, and to create tradablesecurities that could serve as acceptablecollateral in the credit operations of theEurosystem.34

For example, the relative lack of securitisationissuance volumes in some continental Europecountries may be explained by the fact thatbanks in those countries enjoy a morecomfortable funding position. In countries withan established tradition of covered bondstructures (such as Germany and Denmark),covered bonds have served as an efficient andrelatively cheap funding mechanism formortgage banks in residential mortgagemarkets. In Germany, in particular, the driverfor securitisation has therefore been regulatorycapital relief and not reduced funding costs. Inother countries, such as Belgium and Sweden,the retail deposit base has remained strong,reducing the pressure on banks to diversifyfunding channels. Moreover, in Belgium, theabove-average credit rating of the largerBelgian banks enables easy access to the capitalmarket.

Moreover, the existence of large commercialand mortgage banks has so far been a necessaryprerequisite for the securitisation market todevelop. Only large banks reach the criticalissuance size which justifies the highstructuring costs of securitisation transactionson a stand-alone basis. The lack of critical massis an important hindrance to the securitisationmarket in smaller countries such as Belgium andSweden, as well as in central Europeancountries. Smaller banks would have to pooltheir portfolios to achieve enough critical mass.The Spanish regional savings banks have takenthis route. Since 1999, several Spanishdomestic multi-seller deals have come on to themarket. However, with the exception of Spain,multi-originator transactions have been rare inthe European securitisation market.

Overall, the securitisation market remainshighly fragmented in Europe. Only syntheticarbitrage CDO transactions have moved awayfrom single-jurisdiction deals towardstransaction-backed multinational portfolios,and there are number of synthetic pan-EuropeanCMBS deals (20% of the CMBS sector in2003).

FACTORS SHAPING THE FUTURE

On the supply side, growth prospects aremixed. On the one hand, securitisation has notyet achieved its full potential in a number ofcountries. For example, true-sale securitisationis still underdeveloped in Germany. Thefunding needs of German banks have becomemore pressing over the past two years,however, and interest is growing in tappingalternative funding sources via true-salesecuritisation. In order to improve theconditions for true-sale transactions inGermany, the True Sale Initiative (TSI) waslaunched in June 2003 by thirteen Germanbanks under the lead-management of KfWBankengruppe. The bundled lobbying power ofthe TSI achieved exemption for German

34 In the Netherlands a few sizeable programmes were launched in2002 and 2003 for this purpose.


resident SPVs from trade tax on bankreceivables. However, the original ambitions ofthe initiative to set up a multi-seller platformthat would buy and pool bank loans and issuehighly liquid, standardised securities have beenscaled down. It remains to be seen how manydeals the TSI securitisation infrastructure willattract. Nonetheless, expectations are thatGerman banks will increasingly tap the true-sale securitisation market in one way oranother. At the same time a number of countries(Italy, the Netherlands, Spain, France, theUnited Kingdom) are expected to make a forayinto the synthetic market, following in thefootsteps of Germany. In addition, moregovernments seem to be becoming interested inusing off-balance-sheet securitisation as afunding tool. The Portuguese governmentrecently launched a sizeable public assetsecuritisation. Moreover, eastern Europeanmarkets may supplement the issuance bytraditional markets, even though they still havea long way to go.

On the other hand, the treatment of ABSs underBasel II and International Financial ReportingStandards (IFRS) accounting rules may reducethe attraction of off-balance-sheet securitisationto issuers, since they remove the incentives touse them for regulatory or economic capitalmanagement purposes. The RMBS market maybe particularly affected if securitisationbecomes predominantly a funding vehiclebecause, as a funding tool, RMBS will have toface the competition of (structured) coveredbond markets.

The demand side has seen a continuousbroadening of the investor base. According toMerrill Lynch, it includes not only hedge fundsand arbitrage-driven ABS CDO managers, butalso a far broader array of institutionalinvestors such as pension funds, insurancefirms and of traditional bank investors.However, it remains to be seen whether theincreased demand represents a structural shift,with investors deciding to move a portion oftheir overall portfolio permanently into the ABSmarket, or whether it has been driven mainly by

a search for yield, following the tightening inthe corporate bond markets. The real test of howcommitted investors are to the securitisationmarket will come when credit markets start tosoften and spreads start to widen again.

All-in-all, off-balance-sheet securitisation islikely to continue to grow. However, it remainsto be seen whether explosive growth rates seenover recent years can be maintained.

5.3 CORPORATE AND HIGH-YIELD BOND MARKET

Since mid 2003 European corporate bondspreads have come down to exceptionally lowlevels. In the high-yield bond segment(characterised by high risk), spreads were attheir all-time lows in April 2004. They haven’trecovered much since then. It stands to reasonthat the global low interest rate environmentmay have promoted this trend. Yield-seekingbehaviour may have led investors to pursueriskier investment strategies. One mighttherefore ask how bond spreads could beexpected to react to a potential turnaround ininterest rates, which has already started at theshort end of the yield spectrum in the US andGreat Britain. The determinants of themovement of spreads need to be studied beforethis question can be answered with any measureof reliability.

MARKET SITUATION

For quite some time both absolute and relative35

corporate bond spreads have been fallingsharply in the euro area. This movement hasbeen particularly evident in the area ofspeculative bonds. As of 30 September 2004,euro area high-yield bond spreads, at 326 bp,were only around one fifth of their September2001 peak (1,578 bp). The investment gradesegment saw a trend reversal in October 2002across maturities. At the end of September2004, instruments with a residual maturity

35 The relative corporate bond spread is the ratio between thespread and the government bond yield.

45ECB


5 MARKETTRENDS

of one to five years, at an average of 41 bp,were at around 29% of their October 2002levels (141 bp).

When assessing the development of corporatebond spreads and their possible impact onfinancial markets and the real economy, it needsto be borne in mind that in the euro area bank-based financing is still the dominant source ofcorporate finance. In the near future it willprobably continue to be the most importantsource of external financing in many Europeancountries.

One possible reason for this is that the issuanceof bonds is primarily viable for companies withlarger financing needs, whereas economicstructures in many European countries aredominated by small and medium-sizedenterprises. Some euro area countries – likeFrance – nevertheless have a long tradition ofcorporate financing via the bond market. InSeptember 2003 the outstanding amount ofcorporate bonds as a percentage of GDP inFrance amounted to 23%. This figures almoston par with the USA’s corporate bonds, whichtotal 26% of GDP.36 In contrast, the amount ofcorporate bonds outstanding in Germany wasaround 7% of GDP at the end of that year. The

high-yield segment that – due to its highlyvolatile movements – is particularly preferredby market watchers amounted to only 6% of theEuropean corporate bond market at the end of2003.

Most countries did not witness the developmentof a broad corporate bond market until the eurowas introduced. Growth rates in thesecountries, however, were impressive. In Italycorporate bond markets grew by more than1,100% between 1998 and teh third quarter of2003. In Germany markets grew by more than600% in the same period. In the United Statesmarket growth was only 22%.37 These dynamicshave probably been propelled by theintroduction of the euro, catalysing theintegration of the euro area financial markets.Increased financing needs, caused by thetechnology boom and corporate restructuring,and the increasing importance of institutionalinvestors on the demand side of the corporatebond market further triggered the remarkablerates of growth. Not least, European bondmarkets have benefited from the fact that

Chart 11 Spreads on corporate bonds ineuro area by bond grade

(end-month data)

Source: Merrill Lynch.

corporate bonds, AAAcorporate bonds, BBBeuro-denominated high yields

0

400

800

1,200

1,600

0

400

800

1,200

1,600

1998 1999 2000 2001 2002 2003 2004

Chart 12 Spreads on investment-gradecorporate bonds in the euro area bymaturity(end-month data)

Source: Merrill Lynch.

1-5 years to maturity5-7 years to maturity7-10 years to maturity

20

40

60

80

100

120

140

160

20

40

60

80

100

120

140

160

1998 1999 2000 2001 2002 2003 2004

36 Deutsche Bundesbank, “Recent developments in the corporatebond market”, Monthly Report, April 2004.

37 Deutsche Bundesbank, “Recent developments in the corporatebond market”, Monthly Report, April 2004.


corporate bonds had already proved successfulin the United States, which in this regard can beconsidered a pioneer country.

QUANTITATIVE MODELLING OF BOND SPREADDYNAMICS

The detailed results of the quantitativemodelling of bond spread behaviour are postedin the annex A.38 Observation of the spreads ofBBB-rated bonds with seven to ten years tomaturity39 yields the following findings40 forthe period between June 1999 and December2003:

– Within the longer-term equilibriumrelationship, changes in the three-monthmoney market rate cause spreads to move inthe same direction, along the lines ofmacroeconomic portfolio modelling.41

Portfolio theory42 suggests that lower levelsof feasible portfolio returns (given a certainamount of risk) tend to be accompanied byshifts to riskier portfolio investments. Thecurrently very low interest rate levels aregiving market participants an incentive toshift their assets to relatively risky but morelucrative assets in order to yield optimalrisk-return patterns. Such a demand shiftresults in an increase in the prices ofcomparatively risky assets and in diminishedyield spreads.

– Corporate spreads are driven by a rise in thecorporate sector debt to earnings ratio (dte-ratio), which serves as an indicator for thesupportability of indebtedness and solvencyof the companies in the market. In line withoption price theory43, yield spreads, taken as

risk premia, rise with the probability of thelender default.

– Accelerating GDP growth increases thespreads of BBB-rated corporate bonds in theshort term. Increasing economic growthboosts expected yields on investments in thestock market and expected back payments –due to reduced probability of default – in thehighly speculative high-yield corporate bondsegment. This in turn leads to a shift awayfrom BBB investment-grade bonds.

– An increase in relative gross issues ofcorporate bonds tends to reduce the corporatespread. This unexpected result can beexplained by the increase in market liquidityresulting from increased bond issuance

1999 2000 2003

Bonds issued by non-financial corporations 4.3 5.3 7.0Of which high yields 6.0Loans by domestic banks to non-financial corporations 39.5 41.6 41.9Central government bonds 52.3 51.3 52.4

Sources: ECB, Bloomberg, Merrill Lynch, Deutsche Bundesbank.

Table 5 Loans and market capital isation of bonds in the euro area in % of GDP

(annual averages)

38 Allowing for potential endogeneity of several variables as wellas for potential non-stationarity of time series, a vector errorcorrection model is created. The modelling closely follows thatin G. DeBond, “Euro area corporate debt securities market: firstempirical evidence”, Working Paper No 164, European CentralBank, August 2002.

39 The spreads of corporate bonds against government bonds wereapproximated by data from the Merrill Lynch Global IndexSystem, adjusted for the 10-year swap spread. BBB-rated bondsare at the lower end of investment-grade bonds. Their primaryattraction lies in having greater liquidity than speculative gradebonds. In addition, their volumes developed more consistentlyduring the observation period than those of the other ratingsegments.

40 For the estimate of the spread, the model has a goodness-of-fit of58%. The adjusted R2 is 40%.

41 J. Tobin and W. C. Brainard, “Pitfalls in Financial ModelBuilding”, American Economic Review, Vol 58, No 2, May 1968,pp 99-122.

42 H. Markowitz, “Portfolio Selection”, Journal of Finance, Vol 7,No 1, 1952, pp 77-91.

43 R. C. Merton, “An Intertemporal Capital Asset Pricing Model”,Econometrica, Vol 41, No 5, September 1973, pp 867-887. For amodif ied version of the Merton model see F. A. Longstaff andE. Schwartz, “A Simple Approach to Valuing Risky Fixedand Floating Rate Debt”, Journal of Finance, Vol. 50, 1995,pp 789-821.

47ECB


5 MARKETTRENDS

and leading to increased attractiveness ofinvestment in corporate bonds. The currentbroadening and deepening of the Europeancorporate bond market is expected to continuein the future. This gives reason to believe thatits dampening impact on spread movementswill continue unabated.

Supplementary estimates across segments andmaturities show that in a pronounced highinterest rate environment investors are willingto add BBB-rated corporates to their portfoliosonly if they receive much higher spreads thanon AAA-rated corporates or financialcorporates. Greater volatility of short-terminterest rates thus tends to have a greater impacton spreads for BBB-rated corporate bonds thanon those for AAA-rated corporate bonds.

POSSIBLE IMPLICATIONS

The rapid growth and considerable rise in theliquidity of the corporate bond market arebecoming more important for monetary policymakers for several reasons:

– Corporate spreads are increasingly becoming aqualified information variable for monetarypolicy, thanks to rapid and precise dataavailability combined with high informationcontent. Empirical studies show that spreadmovements and the term structure of corporatebonds may complement the term structure ofgovernment bonds in terms of informationalcontent.44 Given the tendency for externalcorporate financing to be increasingly orientedtowards the capital market, this marketsegment, in monetary policy terms, will comeunder yet more intense scrutiny in the future.

– A maturing corporate bond market, as anincreasingly feasible way of financing, addsto the diversification of corporate financing,possibly leading to an optimisation of capitalallocation. In particular, it could help, forexample, to cushion the effects of potentialshocks in the banking sector.45 A stablefinancial market environment with lowtransaction costs and low interest rate

volatility – with regard to the sensitivity ofspreads – will have a decisive influence onmovements in the European bond market asan additional, stable financing channel.

– It is advisable to keep a close eye on thecorporate bond market with regard tofinancial market stability. Capital flows tendto be more volatile in bond markets than theyare in bank-based financing. However, thecorrelation between the two is low.46

– Econometric modelling of bond spreadbehaviour suggests that spread movements arepositively correlated with the interest ratelevel. They therefore work as an amplifierin the monetary transmission process.Interest rate rises exert an additional restrictiveimpact through corporate bond spreads. Theconsequence might be that interest rates risingsharply from today’s very low levels will alsolead to a considerable increase in spreadpremia.

5.4 CREDIT DERIVATIVES MARKET

The market for credit derivatives has expandedrapidly in recent years, and credit derivatives areon the way to becoming one of the mostsuccessful financial innovations in recent history.This chapter covers the development of the globaland European credit derivatives market,highlighting the integration of credit markets in

44 See, for example, B. Bernanke, “On the Predictive Power ofInterest Rates and Interest Rate Spreads”, NBER, Working PaperNo 3486, 1990.

45 See E. P. Davies, “Multiple avenues of intermediation, corporatef inance and f inancial stability”, IMF, Working Paper, 01/115,2001. It should be noted at this juncture, however, that a scenariowith greater difficulty of access to the capital market owing to arecession is probably more likely than that of crisis situationsrelating to bank-oriented raising of external capital. On thissubject, see P. Artus, “Rating, cycle économique, cyclefinancier”, CDC IXIS Flash 2001-221, Banque de France, 2001,“Le cycle financier facteurs amplificateurs et réponsesenvisageables par les autorités monétaires et financières”,Bulletin de la Banque de France, No 96, November 2001,pp 41-65, and H. Hesse and H.-H. Kotz, “Financial Cycles, RealCycles and Monetary Policy”, R. Pethig and M. Rauscher,“Challenges to the World Economy”, Festschrift for HorstSiebert, Berlin, 2003.

46 Deutsche Bundesbank, “Zur Regulierung der europäischenWertpapiermärkte”, Monthly Report, July 2004.


Europe. Particular attention is paid to creditdefault swaps (CDSs), which are the mostimportant instrument among credit derivatives.CDSs are used for various purposes, such astrading and market making, hedging andmanagement of credit risks related to regulatoryand economic capital. The recent development ofCDS indices will further enhance transparencyand liquidity in the market.

CHARACTERISTICS OF CREDIT DERIVATIVES

With credit derivatives, credit risk can beunbundled from other risks embedded in afinancial instrument, thus allowing the separationof credit risk from the underlying creditrelationship. This allows the transfer of risk toother sectors that lack direct originationcapabilities. Credit derivatives are typicallytraded over the counter, taking into account thespecific needs of both counterparties in thecontract. Credit derivatives comprise a variety ofinstruments, such as CDSs, total return swaps,which encompass all the economic risks involvedin a credit transaction, and credit spread options.

CDSs are credit protection contracts in which thebuyer of protection can insure himself for aspecific period against the likelihood that a givenissuer will default by paying a periodic premiumto the protection seller during the contract periodor until the occurrence of a credit event. The sellerof protection agrees that, in the case of a definedcredit event, he will make a contingent paymentamounting to the difference between the par valueof the underlying asset and its market value afterdefault. A CDS transaction does not generallyentail a direct contractual relationship with thereference entity. A specific reference asset is onlystrictly required in the case of a cash settlement.In the case of a physical settlement the buyer ofprotection typically has a “cheapest to delivery”option allowing the market to provide importantmitigation of the squeeze risk, which is stillrelatively high in the repo market for cashinstruments.

In addition to merely separating off credit risk,credit derivatives make it possible to combine

credit risk exposure in new ways; for example,CDSs provide the basis for more complexstructured credit products. Syntheticcollateralised debt obligations (CDOs) can beformed by a portfolio of single CDS contracts.Such securitised instruments have rapidlygained in popularity. The smooth transitionbetween securitised products and creditderivatives is symbolised by hybrid products,most notably credit-linked notes (CLNs). WithCLNs the principal debt will be repaid only ifthe defined credit event does not occur.

STRUCTURE OF THE CREDIT DERIVATIVESMARKET

The British Bankers’ Association (BBA)estimates that the credit derivatives market grewworldwide from a total notional amount of USD180 billion in 1997 to USD 2 trillion at the end of2002. According to the recently published BBACredit Derivatives Report 2003/2004, thenotional amount is estimated to USD 3.5 trillionby the end of 2003 and is expected to rise to USD5.0 trillion by the end of 2004. The BBA reportreveals that single-name CDSs make up roughly51% of the overall credit derivatives market. Thisshare is expected to decline until 2006 in favourof an increasing use of CDS indices, which are avery recent phenomenon in the market. Thelargest share of credit derivatives is written oncorporate assets, which has partly arisen from theincrease in synthetic securitisation. These figuresare in line with other studies assessing globalcredit derivatives markets.47

47 According to the recent market survey by the InternationalSwaps and Derivatives Association (ISDA), the global notionalamounts of CDSs outstanding actually grew from USD 2.2 trillionat the end of 2002 to USD 3.6 trillion at the end of 2003 (see ISDAMarket Survey, 2004). A Standard & Poor’s study put the value ofthe CDS market at USD 3 trillion at the end of the first quarter of2003 (see Demystifying Banks’ Use of Credit Derivatives,Standard & Poor’s, December 2003).The rating agency Fitch has conducted a credit derivativessurvey of around 200 financial institutions active mainly asprotection sellers (see Global Credit Derivatives: A QualifiedSuccess, Fitch Ratings, September 2003). The survey showed thatthese entities sold USD 1.7 trillion worth of protection, of whichabout one-third originated in Europe. On balance, about two-thirds of European banks were net sellers of protection. Inparticular, Germany’s Landesbanks and a number of banks in theBenelux countries were active as protection sellers, taking risksfrom larger, global banks based mainly in the USA.

49ECB


5 MARKETTRENDS

The growing popularity of credit derivativesstems from the increasing importance of riskmanagement for financial institutions. Most ofthe large global investment banks and securitieshouses trade for their own account to buyprotection for their credit portfolios or againstcounterparty risk arising in other OTCderivatives transactions (related to interest rateswaps, for instance). In addition, financialinstitutions can reduce credit concentration andregulatory capital without affecting theunderlying credit relationships by purchasingcredit protection. As active market participants,banks also provide their customers with marketliquidity.

Protection sellers may have various objectivesof which the most prominent are making profiton the credit derivative’s premia anddiversification of credit risks (i.e. taking riskswhich are still not well represented in aportfolio). Market participants are increasinglyadopting portfolio theory considerations intheir investment decisions focusing on risk/return profiles. Historically low yields havealso boosted the search for more attractiveinvestments. Credit derivatives provide theopportunity to invest in higher yield segmentsassociated with sufficient liquidity by sellingcredit protection. Globally, the insurance sector

is a net risk taker. Insurance companies can sellprotection on the assets side through investmentin securities such as CDO or CLN and on theliabilities side by entering into single name orportfolio CDSs.

With respect to European markets, the activitiesof EU banks in risk transfer were assessed bythe Banking Supervision Committee (BSC) ofthe ESCB.48 The BSC survey, which coversover 100 banks, revealed that in 2002 and 2003the credit risk transfer markets continued toperform quite well with regard to market size,liquidity and innovation. Regarding netpositions there is a regionally mixed picture.Whereas Belgian, Spanish, French, Italian,Dutch, Portuguese, Swedish and medium-sizedIrish banks are mainly net buyers of protection,banks from Denmark, Greece, Luxembourg,some Austrian banks, smaller regional Germanbanks and large Irish banks were found to benet protection sellers.49

Chart 13 Global distribution of instruments

(%)

Source: BBA Credit Derivatives Report 2003/2004.

end 2001end 20032006e

0

10

20

30

40

50

60

0

10

20

30

40

50

60

1 Single-name CDSs2 Portfolio/synthetic CDOs3 Index trades4 Credit linked notes5 Total return swaps

6 Asset swaps7 Spread options/swaptions8 Basket products9 Equity linked products

1 2 3 4 5 6 7 8 9

Chart 14 Underlying reference entit ies

(%)


end 2001end 20032006e

1 Corporate assets2 Financials3 Sovereigns (emerging markets)

4 Sovereigns (non-emerging)5 Other

0

10

20

30

40

50

60

70

0

10

20

30

40

50

60

70

1 2 3 4 5

48 See European Central Bank, “Credit risk transfer by EU banks:Activities, risks and risk management”, May 2004.

49 Some Member States also published surveys on their nationalmarkets. According to these reports, the surveyed banks inGermany and France account for €397 billion in risk taking(selling protection) and €393 billion in risk shedding (purchasingprotection).


The aforementioned studies show that the bulkof transactions are cross-border deals takingplace predominantly within the global bankingsystem, with large US banks as the mostimportant counterparties. The vast majority ofreference obligations are rated A or better andCDSs account for the largest share of trades.However, the information available fromprivate and public data sources is still notadequate to evaluate the redistribution of creditrisk and the resulting risk concentration.

MARKET EFFICIENCY AND TRANSPARENCY

It is essential for the further development of thissegment that credit derivatives markets functionsmoothly. However, market imperfection couldbe the result of a number of different factors.

One potential drawback might be the fact thatprotection buyers, when they are also theoriginators of credit, have inside informationabout the risk they transfer and therefore apotentially unfair advantage. In addition, thepossibility of easily shedding credit risks viacredit derivatives might lower the risk aversionof banks, potentially increasing theirwillingness to invest in riskier business alsoassociated with a more expansive credit policy.

Transparency with respect to the content andrisks of credit derivatives contracts is crucial toavoiding adverse selection problems.

With regard to a review of CDS markets,transparency was promoted by the introduction ofCDS indices constructed from CDS portfoliosconsisting of diverse single name CDSs. Themost frequently traded index families were DowJones TRAC-X and iBoxx, which are composedof numerous regional and sectoral sub-indices. Inmid-2004 the two competing sets of indices weremerged into Dow Jones iTraxx indices. The newindices bring together 20 market makers and theyare transparent, rule-based and administeredby a jointly owned private company, IIC. Theimprovements in standardisation, transparencyand the increase in market size recently persuadedderivatives exchanges to give up theirreservations against the introduction of futures oncredit derivatives. It may be presumed that at theend of 2004 futures on iTraxx CDS indices willbe tradable on the big futures exchanges aroundthe world. In addition, institutional investorswhich, owing to investment restrictions, are nottrading in credit derivatives at present might beattracted by greater transparency and theregulatory environment of an exchange.

Chart 15 Buyers of credit protection

(%)


end 2001end 20032006e

1 Banks2 Securities houses3 Hedge funds4 Mutual/pension funds

5 Insurance companies6 Corporates7 Gov./credit agencies

0

10

20

30

40

50

60

0

10

20

30

40

50

60

1 2 3 4 5 6 7

Chart 16 Sel lers of credit protection

(%)


end 2001end 20032006e

1 Banks2 Insurance companies3 Securities houses4 Hedge funds

5 Mutual/pension funds6 Corporates7 Gov./credit agencies

1 2 3 4 5 6 705

1015202530354045

051015202530354045

51ECB


5 MARKETTRENDS

In order to promote the evolution of a standardisedmarket, the International Swaps and DerivativesAssociation (ISDA) released its Credit DerivativesDefinitions in 1999 (last amended in 2003). Themajority of credit derivatives are documentedaccording to ISDA definitions. The fast-growingvolumes and the efforts to standardise contractualterms helped the credit derivatives market tooutperform parts of the corporate bond market withrespect to liquidity. Furthermore, credit derivativesmarkets are characterised by relatively lowtransaction costs. These qualities positively affectprice discovery capabilities, which is an importantfunction of secondary markets. Moreover,integration of credit markets contributes to thepromotion of the price discovery process byfacilitating the adjustment of prices in differentmarkets to new information.

The following paragraphs look at CDSs andbond markets in the context of existing inter-relationships and their price discoverycapabilities.

RELATIONSHIP BETWEEN CDS AND BONDMARKETS

Due to arbitrage relationships, credit spreads andCDS prices should be closely linked. However,various features of bonds and CDS run counter tothe arbitrage assumptions. For instance, variousbonds include options, which make them callable,puttable, convertible, subordinated or structured.Moreover, taxation and liquidity restraints maydistort arbitrage relationships. Small issues andimplied options often hamper adequate pricing ofcorporate bonds and the possibilities ofborrowing and short-selling those bonds arelimited. As a result, temporary or non-genericfactors affect corporate bond prices. Some ofthese shortcomings can be overcome throughcredit derivatives, and the shortening of positionsbecomes quite easy when buying protectionthrough CDSs. However, there is also acounterparty default risk associated with CDSs,and physically settled CDSs contain a cheapest-to-delivery option which allows the protectionseller to choose between the delivery of a numberof different bonds in the event of a default. Thus,

the yield spread between a risky bond and a risk-free security and CDS spreads cannot always befully attributed to credit risk and do notnecessarily coincide.

Chart 17 illustrates the relationship between CDSand bond spreads for Vattenfall, the Swedishutility. At a first glance the two time series areclosely linked, although the CDSs seem to show ahigher volatility.50 This is a pattern also observedfor many other companies.

EMPIRICAL EVIDENCE

An empirical analysis will shed more light onthe degree of integration and price-settingbehaviour of both markets. Hasbrouck (1995)and Gonzalo and Granger (1995) haveempirically analysed price discovery processesby means of common factor models using thevector error correction approach.51 Blanco et al(2004) applied these models to the CDS andcorporate bond markets.52

Chart 17 CDS premia and credit spreads forVattenfal l

(bp)

Source: Bloomberg.

CDS premiacredit spread

01020304050607080

01020304050607080

Nov. Dec. Jan. Feb. Mar. Apr. May Jan. Feb. Mar. Apr. MayJune July Aug. Sep. Oct. Nov. Dec.

2002 2003 2004

50 The standard deviation of CDS spreads (13.15) is slightly higherthan for credit spreads (12.65).

51 See J. Hasbrouck, “One Security, Many Markets: Determiningthe Contribution to Price Discovery”, Journal of Finance, Vol 50,1995, pp 175-199; and J. Gonzalo and C. W. J. Granger,“Estimation of common long-memory components incointegrated systems”, Journal of Business and EconomicStatistics, Vol. 13, 1995, pp 27-35.

52 See R. Blanco, S. Brennan and I. W. Marsh, “An empiricalanalysis of the dynamic relationship between investment-gradebonds and credit default swaps”, Bank of England, WorkingPaper No 211, 2004.


The basic idea behind these approaches is thatprice movements are responses to newinformation. Price discovery means the veryfirst response to new information and impliesthat the market, which is leading in pricediscovery, is moving first as new informationare revealed. Thus, the leading market is settinga new price (in this case, a new risk premia)deviating from the price equilibrium that existsbetween both markets before the newinformation came up. The second market isfollowing the leading market, moving to thenew price with a time lag and a new priceequilibrium between both markets isestablished.

The results of an empirical analysis presented inAppendix B are in line with the findings ofBlanco et al (2004) suggesting that the CDSmarket dominates the price discovery process ina large segment of corporate credit markets,although the bond market is not insignificant inthis respect.

These discrepancies in price discovery might becaused by different information inflowsintroduced by different market participants withdifferent incentives in both markets.

CONCLUSION

The remarkable development of creditderivatives markets in Europe and the ongoingintegration of European credit marketscontribute to the evolution of liquid marketsfacilitating the efficient pricing and trading ofcredit risks. The majority of credit derivativetransactions are cross-border deals taking placepredominantly within the global bankingsystem. Recent innovations will furtherimprove the potential for growth in the creditderivatives market. In this context, the newiTraxx indices and the intended launch offutures on these indices will offer investorsdiversification and accessing exposure tomarket direction with a single liquidtransaction.

CDSs, which also provide the basis for morecomplex structured instruments, fulfil animportant function in secondary credit marketswith respect to price discovery and riskallocation. The possibility of transfering creditrisk in liquid markets and an efficient pricesetting process facilitates the efficientallocation of credit risk.

5.5 INFLATION-PROTECTED FINANCIALINSTRUMENTS

Inflation-linked bonds is a small but growingsegment of the euro bond market. Most of theEU national treasuries which have alreadyissued some inflation-linked bonds are tendingto increase their issuance (United Kingdom,Sweden, France, Italy and Greece), whereas theGerman Treasury is expected to start to issue in2005. In parallel, the market for inflation-linkedderivatives has picked up over the past threeyears, expanding the hedging and tradingopportunities of inflation risk.

THE GROWING MARKET FOR INFLATION-LINKEDBONDS

An inflation-linked bond (ILB) is a bond whosecoupons and principal are linked to a consumerprice index. For ILBs issued within the euroarea the linking index is the HICP excludingtobacco or, in the case of one of the two FrenchILBs (OATi), the French CPI index excludingtobacco. The coupons and the principal arelinked to the price index with a lag due to theproduction time of the indices.53 Within the euroarea, ILBs have a guaranteed principal, whichmeans that investors are protected fromdeflation. ILBs are by nature long-term bonds,maturing between 2008 and 2032.

Issuance of ILBs is an old practice which washistorically often used in a high inflationenvironment by governments which wanted to

53 This time lag (usually three months) exposes investors to inflationwhen the bond is close to maturity. In fact, the indexationmechanism will not take account of the evolution of inflation inthe three months preceding the maturity date.

53ECB


5 MARKETTRENDS

meet a demand from investors for real yield.54

Nonetheless, examples of issuance remainedisolated. The auction of an indexed Gilt by theUK Treasury in 1981 was the first attempt tocreate a modern and liquid market for ILBs. Thesecond turning point in the growing ILB markettook place in a low inflation environment, withthe decision of the US and French Treasuries toissue ILBs in 1997 and 1998 respectively. Sincethen, within the euro area, Italy, Greece andAustria each issued ILBs in 2003 and Germanymay issue one in 2005. Of the non-euro areaMember States, Poland (1992), Sweden (1994)and Hungary (1995) have experienced issuanceof ILBs.

Within the EU, Treasuries have tended toincrease their issuance of ILBs over the pastthree years in response to a strong demand forsecure long term assets by investors who have ahigh aversion to inflation risk. In absoluteterms, the most active issuers over the last two

years have been the French and ItalianTreasuries, although Greece has also financed asignificant share of its issuance via ILBs.

As regards the inflation-linked bonds issued bynon-central-government agents, only the UKmarket reaches a significant size. Most of thetime, non-government issuers are utilities oragencies. However, the development of this ILBmarket is restrained by the lack of liquiditywhich leads investors to demand a highliquidity premium, pushing up the relativefunding cost of ILBs for non-governmentissuers, especially corporates.

54 Apparently the first inflation-linked bond was issued inMassachusetts in 1780. From that year until the 1970s the othermain ILBs were issued by the governments of Finland (1945),Sweden (1952), Iceland (1955), Israel (1955), Brazil (1964),Chile (1966), Colombia (1967) and Argentina (1972). For furtherdetails on the emergence of ILBs, see R. Shiller, “The Inventionof Inflation-Indexed Bonds in Early America”, NBER WorkingPaper No 10183, December 2003.

UK France Italy Sweden

Market value (EUR billions or equivalent) 130 65 23 21Number of indexed bonds 9 6 2 5

Table 6 Outstanding amounts of the main European ILBs at the end of May 2004

Source: UK Debt Management Office (DMO), end of May 2004.

Source: Bondware.1) Data in 2004 up to end-June.

Chart 18 Issuance of ILBs by EU centralgovernments

(EUR millions)

02,0004,0006,0008,000

10,00012,00014,00016,00018,000

02,0004,0006,0008,00010,00012,00014,00016,00018,000

2000 2001 2002 2003 2004 1)

UKFRSEITGR

Chart 19 Issuance of ILBs by EU centralgovernments as a percentage of the total

0

5

10

15

20

25

30

0

5

10

15

20

25

30

2001200220032004 (H1)

UK FR SE IT GR


Even with regard to government bond issuance,the ILBs have a reputation for being less liquidthan the traditional sovereign bonds. Forinstance, the bid-ask spread can be set,depending on the bonds, four or five times aswide for an ILB as for a traditional bond. Therelative lack of liquidity is due, of course, to thesize of issuance, but also to the technicalcharacteristics of the ILB. For instance, allother things being equal, a longer indexationlag reduces liquidity because arbitrageopportunities are more difficult to assess andhedging strategies are also trickier toimplement.

In July 2004, the electronic trading platformEuroMTS launched a new market dedicated tothe trading of inflation-linked securities. Anindex of ILBs was also created at the same time.The index currently includes nine bonds: oneGreek, two Italian and six French inflation-linked bonds. It should provide a reliablebenchmark and is likely to contribute topromoting the liquidity of the bonds included inthe index.55

THE REASONS FOR THIS DEVELOPMENT

There are three basic reasons for issuing ILBs:to reduce the cost of funding, to broaden theinvestor base, and to manage the issuer’s assetsand liabilities.56

As far as the first reason is concerned, byissuing an ILB a treasury or a corporate cansave the inflation risk premium. Indeed,according to the Fisher relation betweennominal and real interest rates, the nominal rateis a sum of the real rate, the inflation expectedby the market and a premium which is mainly aninflation risk premium. By neglecting the otherpremia (especially the liquidity premium), thelinear approximation of the Fisher relation isthat:

nominal rate = real rate + expectedinflation + inflation risk premium,

thus:

break-even rate57 = nominal rate– real rate

= expected inflation+ inflation riskpremium

The inflation risk premium remuneratesuncertainty about the ex post real yield of anominal bond, but in the case of an ILB thisuncertainty is almost completely ruled out forinvestors. Therefore, the issuer can basicallysave the inflation risk premium because heaccepts the inflation risk.

A treasury can also take advantage of ILBs if itdeems market inflation expectations to be toohigh. This is sometimes the case when marketsfall victim to adaptive expectations or a nominalillusion which could lead to an over-estimationof the inflation break-even rate, especiallyduring periods of inflation slowdown.58

Nevertheless, the growing efficiency of thebond market and the decrease in the break-evenrate toward levels more or less in line with thelong-term inflation expectations no longerallows for such opportunistic issuance.

The second reason for issuing ILBs is thepossibility of attracting long-term institutionalinvestors (e.g. pension funds) which have a

55 All euro-denominated inflation-linked government bonds of aminimum of €2 billion supported by at least eight market makersand with at least one year to maturity will be eligible for tradingon the platform.

56 Other justifications for issuing ILBs are also sometimesmentioned by some treasuries, such as improving the credibilityof their anti-inflationary policy (although this argument isquestionable insofar as indexation can make it easier to spreadinflationary pressures and treasuries are not responsible forkeeping inflation low) or offering a public instruent useful formanaging macroeconomic risk and assessing inflationexpectations.

57 The precise definition of the break-even rate is: break-evenrate = (1 + nominal rate) / (1+ real rate) – 1.

58 From this point of view, the issuance of the f irst indexed Gilt bythe UK Treasury in 1981 was a very cheap funding.

55ECB


5 MARKETTRENDS

specific interest in interest-linked instrumentsfor assets and liabilities management (ALM)purposes. In addition, this aim of broadeningthe investor base often explains the choice ofinflation for the reference index. Bonds linkedto the national CPI, as opposed to the HICP,allow for a better ALM for local investors at theexpense of a reduction in the overall number ofpotential investors. The different issuingstrategies of euro-area governments illustratethis point. Among central government issuers,only France has issued two types of ILBs(indexed on the national CPI and the HICP).59

Other countries have so far chosen to use theHICP, without ruling out the possibility ofissuing ILBs linked to to their own CPI.

According to the third reason, issuers are alsoattracted by ILBs in order to manage theirbalance sheet exposure. Agents such asgovernments, which have revenues linked toinflation (mainly through VAT), may beinterested in linking the cost of their funding inorder to smooth the cyclical pattern of deficitthrough the inflation cycle.

As regards risks taken by issuers, the mainspecific drawback of ILBs for issuers comesfrom the fact that investors demand a liquiditypremium to invest in ILBs because of therelative shallowness of this new market.60 Fromthe investor’s point of view, adding an ILB to aportfolio can improve the risk-return trade-offof the portfolio and provide suitable tools withwhich to manage the assets and liabilities ofinvestors.

Introducing an ILB into a portfolio improvesthe risk-return attributes of the portfolio bydiversification. Indeed, thanks to theintroduction of an ILB in a portfolio theefficient frontier is significantly positivelyshifted because ILBs usually have lowcorrelation coefficients with other assets. Thus,empirical evidence suggests that the volatilityof the price of an ILB is usually lower than thatof a traditional bond. However, ILBs haveanother advantage for investors: they are inessence risk-free assets since their price

depends solely on the variability of the realyield curve. Indeed, ILBs enable the investorsto lock in an ex ante real yield. This feature ofILBs is valuable for agents who have to managethe saving of their clients (insurancecompanies, pension funds, etc) over the longterm in order to maintain the purchasing powerof their capital. In this context a “buy and hold”strategy enables inflation-adverse agents toprotect themselves efficiently against inflationrisk.

The other origin of the strong demand forindexed products comes from the need to managethe assets and liability items on the balance sheet.One of the standard factors of demand consists inthe desire to match liabilities that are correlated toinflation. This is typically the case for definedbenefit pensions funds, as retirement benefitsdepend on the final salary, or for property andcasualty insurance companies, given thatindemnities closely follow price levels.Furthermore, the new accounting frameworkshould further encourage investors to betterimmunise their balance sheet against inflationrisk. Indeed, under those new rules, if thevariations of liabilities are not wellhedged by a parallel move on the asset side,the shareholders’ equity would decreasesignificantly. Furthermore, regulation caninfluence the demand for inflation protection. Forinstance, the recent partial linkage to the FrenchCPI of the remuneration of the French tax-exempt saving accounts (such as “livret A”),which represent very large amounts,61 gave astrong incentive to the banks managing ILBs tolook for inflation protection as a means ofensuring an effective ALM.

59 The Lazio region in Italy recently launched an ILB linked to theItalian CPI.

60 The cost effectiveness of the ILBs for Treasuries is very difficultto assess owing to the uncertainty about the inflation rate in thelong term. Nevertheless Sack and Elsasser estimated that thedifferential cost of the US issuance programme of ILBs incomparison with a nominal equivalent programme was aroundUSD 3 billion between October 1997 and January 2003. See B.Sack and R. Elsasser, ”Treasury inflation-indexed debt: a reviewof the US experience”, FRBNY Economic Policy Review, May2003.

61 The total volume of indexed savings accounts is around €300billion. These accounts have been linked to the three-monthEuribor and the French CPI in equal measure since 1 July 2004.


The risks borne by buyers of ILBs include: thepossibility of there being a slight bias in marketexpectations of inflation, the fact that thecoupons are linked to inflation with a time lag,the mismatch between the linking index andthe relevant index for the investor, etc.Nevertheless, these drawbacks seem limitedcompared with the flaws of the other hedgingstrategies such as investing in the moneymarket, in equities, in real estate or in gold, allof which provide very poor and impure inflationhedging. Institutions have therefore turned toILBs and inflation derivatives for more preciseand effective protection. However, the demandfor ILBs has so far clearly outpaced supply,putting a limit on the effectiveness of inflation-related ALM through ILBs. This situation isunlikely to improve in the coming years giventhe increasing protection needs associated withpopulation ageing and the accompanyinggrowth in pension funds assets,62 even ifinflation derivatives may help to bridge the gapbetween the demand and the supply of indexedproducts.

THE EMERGENCE OF THE INFLATIONDERIVATIVES MARKET

The derivatives market is mainly a swap marketalthough more sophisticated products, such asoptions on inflation, are now available.

Basically, an inflation swap is a contractbetween an inflation “receiver” who pays afixed rate versus an inflation-indexed rate(floating or fixed). The inflation “payer” paysthe indexed rate, thereby bearing the inflationrisk. In this market, the final inflation payersare the same as in the ILBs market – utilities,retailers and other agents whose revenues arelinked to inflation. However in the case of aswap transaction, a hedging bank manages themismatch of needs between the counterparties,implementing structured inflation transactions.

The inflation swap market has developedstrongly in countries in which ILBs are

62 Notably in the countries, such as Germany and France, thatrecently introduced pension funds.

A: inflationpayer

B: inflationpayer

Box 4

EXAMPLES OF INFLATION SWAPS

Let us imagine a swap between A, an inflation payer, and B, a person who wants to hedgeinflation risk. In this example the swap is a zero coupon, the index is the HICP and the maturitydate is 30 June 2014, the starting date is 1 July 2004, the indexation lag is equal to three monthsand the fixed leg of the swap is equal to 2%.

On 30 June 2014:A pays: [HICP March 2014 / HICP March 2004]*notionalB pays: [(1+2 %)^10-1]*notional

Inflation swaps can also pay regular coupons. In this case, which is more common, the swapsgenerally look like this:

Fixed rate + inflation

Euribor or fixed-rate + spread

57ECB


5 MARKETTRENDS

available for hedging the position of theinflation payer, which is why most of thecontracts are linked as ILBs to the euro HICPexcluding tobacco, to the UK RPI, or to theFrench CPI excluding tobacco. The tradesusually have a maturity within a range of 5 to15 years.

The growth of this market is a consequence ofthe needs of investors and issuers not beingsatisfied by the ILB market. As regards theprotection seller, the main advantage of thederivatives and swaps market comes from theconvenience, the speed and the low barriers toentry of a swap contract by comparison with abond issue. Thanks to derivatives, protectionbuyers can obtain a very flexible product,possibly tailor-made, which directly fits theinflation exposure of investors in terms ofmaturity, underlying index, size and timing ofcash flows. Buyers of the fixed leg of the swapdo not have to manage the maturity mismatches,the reinvestment of coupons, the credit risk ofissuers and so on. To summarise, the derivativemarket provides efficient tools to enhance theprocess of matching the needs of the differentmarket participants, but what are thedrawbacks?

The main disadvantages associated with thederivatives market are the lower liquidity andthe lower break-even point offered by swaps.

From this point of view, the existence ofindexed bonds for certain maturity datesprovides an anchor which underpins theliquidity of both the swaps and the ILBs forthose specific points of the curve.

The relative lack of liquidity makes the hedgingof short-term inflation risk through the swapmarket very tricky. Market participants wishingto take short-term positions on the level of theHICP can use the OTC options market that hasbeen managed by Deutsche Bank and GoldmanSachs since May 2003.

Nevertheless, apart from the inflation swapmarket, the other derivatives markets are still intheir infancy; their development could offset thescarcity of ILBs and complete the set ofinstruments available for efficient hedgingagainst inflation risk.

5.6 REAL-TIME INDICES AND EXCHANGE-TRADEDFUNDS

Several means exist by which financialinnovation can contribute to lowering the cost ofinvesting in diversified portfolios of corporatebonds. One possible approach is through thedevelopment of exchange-traded funds (ETFs),which allow a diversified portfolio to be boughtor sold through a single transaction. Anotherway in which innovation could help to lowertransaction costs would be through thedevelopment of OTC and exchange-tradedderivatives based on portfolios of corporatebonds, with delivery taking place either in cashor through ETFs. A prerequisite for thedevelopment of ETFs and related derivativecontracts is the existence of indices whoseintegrity is beyond doubt among marketparticipants and that comply with certain market-oriented characteristics: price computation, indexfeatures, bond eligibility criteria and rebalancingshould be rule-based, transparent and replicable.Moreover, fair prices should be transparent andmade available in (almost) real time. This chapterhighlights the development of both real-timeindices and (fixed income) ETFs.

Chart 20 Turnover in the secondary marketfor French ILBs and for the euro inf lationswap market

Sources: Agence France Trésor and ICAP plc.

0,0May

2001 2002 2003 2004July Sep. Nov. Jan. Mar.May July Sep. Nov. Jan. Mar.May July Sep. Nov. Jan. Mar.

5,0

10,0

15,0

20,0

25,0

30,0

0,00,51,01,52,02,53,03,54,04,55,0

OATi and OAT€ i turnover (EUR billion, left-hand scale, monthly data)inflation swaps brokers turnover (EUR billion, right-hand scale, monthly data)


IMPEDIMENTS TO THE DEVELOPMENT OF THECORPORATE BOND MARKET

Recent developments and trends in thecorporate bond market were highlighted inChapters 3 and 5. In addition to the introductionof the euro, there are many elements that haveplayed a role in fostering or hindering thedevelopment of the corporate bond market. Anenvironment of low nominal yields, forinstance, has encouraged an appetite for theyield pick up provided by non-sovereign bondsand thus bolstered demand for corporate debt.However, several elements that may havehindered the development of the corporate bondmarket are the lower level of liquidity providedby these instruments, the higher transactioncosts incurred by investors and the difficulty ofhedging easily and cheaply positions held inthis market.

Corporate bonds tend to incur highertransaction costs than government bonds andother benchmark bonds, such as JumboPfandbriefe. The main reasons for that are thelower outstanding amounts of each bond and thedifficulties for dealers trying to hedge positionsin an environment of asymmetric, discontinuousand not always publicly available information.The first issue also affects bonds issued bygovernments of “smaller” euro area countries.63

Beyond the question of liquidity as such is thequestion of the number of transactions that aninstitutional investor needs to perform to buildup a fully diversified corporate bond portfolio,as opposed to a government bond portfolio. Forasset managers who measure their performanceagainst bond indices this implies additionaltransaction costs as well as costs associatedwith the difficulty of monitoring a largernumber of issuers and instruments andassociated with the complexity of replicatingmarket indices which in most cases have notbeen designed for this purpose.

One additional problem raised by investment ordealing in corporate bonds is the question of theavailability of appropriate hedging instruments.Bond futures and interest rate swaps (IRSs)

have so far been the most commonly usedinstruments for hedging positions held in thebond market. Bond futures and IRSs, however,mainly allow the hedging of market risk and notcredit risk exposure. In fact, when using themto hedge corporate bond portfolios, one incurs abasis risk which obviously increases with thevolatility of the underlying spreads betweengovernment and corporate bonds.64 All otherthings being equal, this credit risk is likely todeter investors from taking positions incorporate bonds or to lower the price of thesebonds through the incorporation of a premiumcompensating for this particular risk.65

Important financial innovations in the creditderivatives market which provide tools toaddress in particular the remaining credit riskare described in Chapter 4.4. In this context itshould be recalled that iTraxx indices are basedon credit derivatives and allow the completeseparation of other risks embedded in cashportfolios (like interest rate risk) from creditrisk, which is not the case for ETFs which arefunded investment tools.

The following paragraphs describe recentinnovations targeted at a specific set ofinvestors, such as private individuals or lesssophisticated asset managers, in order to helpthem to tackle some of the aforementioneddrawbacks of the bond market.

DEVELOPMENTS IN THE INDEX SECTOR

An important development, whoseconsequences have probably not yet allmaterialised, is the development of real-timeindices, i.e. market indices computed andpublished in real time throughout the tradingday. The importance of market indices tosupport investment in any market segment iswell known. Indices serve several purposes.

63 By way of illustration, the overall public debt of many new EUMember States amounts to less than the standard size of one bondissued by the largest sovereign issuers. The same applies, ofcourse, to corporate issuers.

64 Note that this applies also to investment in government bondsother than those underlying the futures contracts.

65 There are of course other ways to hedge positions in corporatebonds, such as swaps, CDSs, etc.

59ECB


5 MARKETTRENDS

They serve to convey to investors the objectiveof a fund. They also serve risk management andperformance measurement purposes.

However, indices can also fulfil other roles.They may be used as references in the draftingof financial contracts, in particular derivatives.This is standard practice for instance in theswap market, where references such as theEONIA or EURIBOR are customarily used tothat end. In principle, the possibility of usingindices as references for cash-settled futurescontracts may also be considered or, indeed, forthe development of any other type of derivativeinstrument that can be used to hedge positionsin any market.

In fact, in any market, the availability ofhedging instruments appears to be crucial to thedevelopment of the underlying market. Withoutthe possibility of offsetting rapidly and at lowcost the specific risks induced by entering into atransaction with customers, intermediaries (e.g.market makers) would be far less able or willingto provide liquidity for the secondary market.Hedging instruments, in practice, often tend tobecome the actual reference for the pricing ofthe instruments that they are used to hedge, asopposed to the other way round. It may beargued, for instance, that the “true” benchmarkfor the euro area government bond market is theBund futures contract rather than the underlyinggovernment bond.

The development of futures based on onecorporate bond or a basket of corporate bondswould in principle facilitate the management ofmarket risk associated with dealing in corporatebonds but would in practice be difficult becauseof the small outstanding amount of eachcorporate bond and/or the costs associated withthe management of a basket.

Tradable indices or futures contracts settledin cash and based on the value of an indexwould seem to provide a convenient solutionhere. They would broadly reflect the behaviourof prices of corporate bonds and wouldtherefore be convenient for hedging a large

range of instruments. They would also berelatively free of the problems (e.g. squeezes ofthe underlying asset) sometimes associated withfutures contracts. This requires, however, thecomputation and publication (preferably in realtime) of indices with a high degree of integrityand reliability.

Against this background, the development ofsuch indices must be seen as having potentiallysignificant consequences. One series of suchreal-time indices which cover sovereign andnon-sovereign issuers has been developed byiBoxx, a fixed income provider established forthat purpose. iBoxx, which was founded byDeutsche Börse and seven investment banks,currently computes real-time price and totalreturn indices for investment grade corporatebond indices in euro and pound sterling onthe basis of indicative quotations providedprimarily by a growing number of dealers (nineat present). The indices have been jointly set upwith the aim of addressing the tracking andhedging concerns by selecting liquid bonds,limiting their number per index and excludingspecial bond types.

The limitation of such an innovative processis that iBoxx index prices do not resultfrom market makers’ binding quotations orexecuted trades. Furthermore, an importantcomplementary step which has been missing sofar points to market jointly agreed rules forcomputing asset swap spreads, based on thesame swap yield curves. In other words totalreturn indices, which represent the performanceof a market, should be also expressed in termsof spreads providing an “objective” measure tobe referenced for financial contracts (similarlyto CDS indices). This would, inter alia, solvethe computational problem raised by thedifferent methods existing for computing assetswap spreads but most importantly would helpmarket participants to reduce mispricingbetween the cash and the derivative creditmarkets through more efficient arbitrage.Beyond the application of total return indices,price indices are useful for retracing marketmovements. The potential of asset swap spread


indices, which should be tradable by a largenumber of dealers with jointly agreed rules,may be somehow comparable to the recentlyexploded CDS indices.

EuroMTS and Euronext have also jointlydeveloped indices for European governmentbonds, which are published in real time (theEuroMTS index or EMTX). While this initiativehas not yet extended to non-sovereign bonds,this may take place in the future (e.g. on thebasis of the EuroCredit MTS coverage). Theextension to real time indices of non-sovereignbonds would provide a similar level oftransparency and integrity as indices for bondmarkets that have existed in equity markets foryears and have allowed the development ofhedging instruments in these markets.

DEVELOPMENT OF FIXED INCOMEEXCHANGE-TRADED FUNDS

A development which has resulted from thesetting-up of related real-time indices andwhichmay be instrumental in allowing thecorporate bond market to develop further is theintroduction of fixed income exchange-tradedfunds (ETFs).

ETFs are similar to mutual funds in that theyrepresent a fractional ownership in anunderlying portfolio of securities. Typically,the underlying portfolio is built up with theview of replicating a given market index.

A particularity of ETFs, however, is that theytrack the performance of market indices or sub-indices, allowing investors to buy or sell anentire market in one trade, through a singlesecurity. They are traded on an exchange orover the counter, in a way similar to individualstocks. This implies that investors do notnormally purchase shares (or redeem them)from the fund itself. They purchase shares from– or sell them to – dealers or other investors,without incurring any transaction for theunderlying portfolio. Another key element isthat prices are available for ETFs at any timeduring trading hours and are updated on a real-time basis (according to demand and supply).

The first ETFs were based on equities. On29 January 1993 the first ETF tracking theS&P 500 index was launched. The ETF markethas developed rapidly since then, especially inrecent years. By the end of June 2004 theEuropean ETF market totalled about €21 billionassets under management (AUM) accountingfor some 10% of the global ETF market. Afterthe United States, the European market is themost rapidly growing ETF market, almostmatching the Japanese market in size and clearlyoutrunning it in terms of trading volumes.

Fixed income ETFs, which are a relativelyrecent phenomenon, have only really started todevelop in the past few years. In the USA thefirst one was launched in July 2002. The firstquarter of 2003 saw the launch of the first fixed

Country/region End-June 2004 End-March 2004 End-December 2003

USA 145,882 132,517 119,659Europe 21,027 20,483 16,230Japan 24,254 23,478 21,937Canada 4,212 4,717 4,078Hong Kong 3,221 3,291 3,137Taiwan 1,085 1,124 923ETF total 201,948 188,403 168,333

Sources: Morgan Stanley ETF Strategies; Bloomberg.

Table 7 ETF assets under management

(EUR millions)

61ECB


5 MARKETTRENDS

income ETFs available to European investorsand by the end of June 2004 the volume of fixedincome ETFs amounted to €2.2 billion.66 TheEuropean market clearly has the potential todevelop into a relatively liquid marketaccessible to a growing number of investors,i.e. to asset managers, hedge funds, andinstitutional and individual investors andtraders.

COMPARATIVE ADVANTAGES OF EXCHANGE-TRADED FUNDS

Fixed income ETFs provide private investorswith several benefits over and above what isavailable from the more traditional investmentvehicles. Diversification is a major concern forany investor seeking exposure to corporatebonds. However, adequate diversification isdifficult to achieve for small portfolios. For thisreason, many private investors seekingexposure to bond markets tend to favour funds.Compared with traditional funds one of thebiggest advantage of ETFs is the very low levelof fees.67

As ETFs are index-tracking strategies, the fundholdings and risk exposures are completelytransparent to the investor, giving the investorsthe return on a specific market. Apart fromlower fees, ETFs trade intra day and can be soldshort, unlike traditional funds. This allowsinvestors to accept shorter-term, or immediate,exposure in a market or to use ETFs to hedgethemselves against specific short-term eventrisks.

Many institutional investors gain their exposureto bonds not by investing directly in the bondsthemselves but rather by holding CDSs or byutilising futures or swaps within their portfolios.However, with futures contracts, the investordoes not receive any coupon income and eachquarter he must roll from one contract to the next,incurring transaction and administrative costs.Moreover, no futures contracts exist on corporateportfolios and it is therefore difficult to gain largeexposure given the liquidity available.

As long as the bulk of fixed income investmenttook the form of government bonds – especiallyin countries where governments typicallyconcentrate issuance on a few benchmark issues– the advantages of ETFs in this sector wereperhaps limited. This may be changing,especially in Europe, for the following reasons:

– Demand for corporate bonds implies thepurchase of very diversified portfolios withmany small issues. This is a situation moreakin to investment in equity markets, andtherefore investments in ETFs may be moreattractive as ETFs seek to give investorsexposure to a diversified credit portfolio,through a tradable product;

66 The f irst was the eb.rexx German government bond ETF,launched in February 2003 by Index change and primarilytargeted at German retail investors. The second was the iBoxxEuro Corporate Bond ETF, launched in March 2003 by iShares,primarily targeted at institutional investors, but accessible toretail investors, too.

67 According to Fitzrovia, the average total expense ratio (TER) forfixed-income ETFs is 17 basis points per annum, which is farbelow the average 99 basis points charged by fixed-income fundsin Europe.

Type of exposure No of ETFs Total assets (EUR millions) % total

Regional euro area 13 7,255 34.5 European country 16 6,828 32.5 Fixed income 12 2,156 10.2 Regional Europe 6 1,762 8.4 International 13 1,705 8.1 European sectors 33 869 4.1 Euro area sectors 10 271 1.3 Other 2 159 0.8 Global sectors 3 16 0.1

Table 8 European ETF assets by type of exposure as at 30 June 2004

Sources: Morgan Stanley ETF Strategies; Bloomberg.


– EU (and euro area) enlargement implies thatbroad government bond indices now alsoencompass a much larger number of smallerbonds, as opposed to a US Treasury index,for example. There may therefore also be aparticular interest in ETFs in the Europeangovernment bond sector.

FACTORS AFFECTING THE DEVELOPMENT OF THEETF MARKET

Beyond the arguments mentioned above, whichmay foster the development of the fixed-incomeETF market, there are also various obstaclesand market forces that may obstruct itsdevelopment:

– An ETF is not a product that banks are keento offer as it cannibalises their highermargins products. This is manifestly evidentobserving the limited number of banksinvolved in this business.

– The constant duration may be rather a rigidfeature for investors and it exposes them to

market risks that a more active managementmay limit. This is more relevant for fixedincome ETFs than equity ETFs becausecoupon cash flows play a larger role thandividends.

– Fixed income ETFs offer fewer advantagesthen equity ETFs in terms of differential feecompared with normal index funds.

– At present, ETFs make up a very nichesegment and therefore a limited productinnovation (in a way confirmed by the USfixed income ETF market) because it is notreally supply driven.

To allow this product innovation to be really amarket innovation, educational effort is neededto change the approach of financial advisors andfund managers: instead of concentrating ontactical allocation and thus specific individualassets (which can be easily and cheaplyreplicated with ETF), financial advisorsand banks would have to steer their clientstowards strategic allocation and therefore

Feature ETF Closed-end fund Ordinary mutual or open-end fund

Pricing Real-time share price. Portfolio Real-time share price. End-of-day End-of-day net asset value.values updated throughout day. net asset value. May trade at aMay trade at tiny discounts or substantial discount or a steeppremiums. Pricing very transparent. premium.

Trading shares Trades as a stock. Can use limit Trades as a stock. Can use limit Transactions normally occur at theorders to buy or sell at your price orders to buy or sell at your price or end-of-day net asset value. Can onlyor better. Can also trade intraday. better. Can only trade once a day. trade once a day.

Average 0.15% 1.05% for managed government 0.35% for index-based funds;expense ratios and corporate bond funds. 1.05% for managed government and

corporate bond funds.

Management Index-based Actively managed adding a layer Actively managed adding a layer ofof management risk. management risk.

Portfolio Limited at present to sovereigns Wide assortment of taxable and Wide assortment of taxable andchoices and high-grade corporate portfolios. tax-free portfolios. tax-free portfolios.

Transaction Brokerage commission, bid-ask Brokerage commission, bid-ask Generally none if no-load fund iscosts spread, and potential price impact spread, and potential price impact purchased through fund company.

of large trades. of large trades.

Leverage Unleveraged Many use leverage to boost yields, Unleveragedwhich adds volatility.

Can be sold Yes No Noshort

Transparency Portfolio of bonds disclosed Portfolio of bonds must be reported Portfolio of bonds must be reportedmonthly or even daily. at least semi-annually. at least semi-annually.

Table 9 The features of ETFs and traditional funds

63ECB


5 MARKETTRENDS

the relative weight that should be attributedto equities, credit, currency, etc., includingthe geographical and sectorial exposure. Thiswould allow final investors to save managementcosts and pay for the real added value providedby asset managers or financial advisors interms of strategic allocation. Mainly non-sophisticated investors would gain a lot fromthis process.

At present the agents which appear reallycommitted to this educational effort for fixedincome ETFs are mainly the stock exchangeswhich are already quoting the instruments(Borsa Italiana, Euronext Group and DeutscheBörse), the few fixed income ETF managers(mainly IndEXchange, BGI and Lyxor Int AM)with BGI as the only issuer of corporate ETFs,and electronic trading platforms like EutoMTS,Eurex Bonds and Bond-Vision.

The involvement of electronic trading platformsshould be regarded favourably as it could alsoresult in the extension of underlyinginstruments to non-sovereign bonds, inparticular corporate bonds, and in a movementof institutional investor trading from OTC tomore transparent markets. At the same time, theinvolvement of exchanges is very important tofurther broaden the investor base and developinnovative derivative products (e.g. options andfutures), enhancing the availability ofinstruments used by more sophisticatedinvestors for trading and hedging purposes.

A combination of the elements described abovewould naturally lead to increased liquidity forthe underlying bonds, resulting in a virtuouscycle for real-time non-sovereign indices,corporate in particular, with fairer prices.

CONCLUSION

The development of real-time indices andexchange-traded funds in the bond market – andin the corporate bond segment, in particular – isinteresting to monitor because it may allow twoimpediments to the development of this marketto be mitigated: the lack of convenient hedging

instruments and the relatively high transactioncosts associated with fully diversifiedportfolios composed of many small issues.

More specifically, the development of these twoproducts may increase the attractiveness of thecorporate bond market for investors andintermediaries, and therefore trigger thedevelopment of this market.

It is mainly in the interest of private investorsthat these products are developed, in particularwith varied range in the fixed income non-sovereign sector, for various currencies andregional exposure. At the same time,innovations in the strategical investmentapproaches will lead more and more assetmanagers to benefit from these products,fostering suppliers (including banks) to bemore proactive and therefore transforming thisproduct innovation into a real marketinnovation.

5.7 RATING AGENCIES

Since the advent of the euro, rating agencieshave been playing a pivotal role in thedevelopment of the euro bond market, asproviders of independent credit assessments onbond issuers’ creditworthiness. Yet, thecoverage of credit ratings in Europe is stillunder-developed compared with the UnitedStates as a consequence of a greater reliance onbank intermediation. This section focuses onthe coverage of credit ratings in Europe,compared with the USA, but also on thecoverage in terms of European sectors andacross European countries. While highlighting acatching up effect, the section also discussesthe factors that might further the use of ratingsin Europe in the future.

CREDIT RATINGS AND THE GROWTH OF THEEURO BOND MARKET

The European bond market has been growingvery rapidly since the launch of the euro. One ofthe factors underpinning this growth has been


the increasing coverage and use of credit ratingsprovided by rating agencies. Yet, credit ratingsare still significantly less frequently available inEurope compared to the US. Both generalstructural factors and specific European driversexplain the role of rating agencies in theEuropean bond market.

Traditionally, the use of credit ratings hasalways expanded hand in hand with the growthof debt capital markets. This was for instancethe case in the US during the twentieth century,as the number of rated issuers grew (and attimes retreated) in response to the ups anddowns of the bond market68. It reflects theessential role of credit ratings69 in a market-based economy, which is all about transparencyand the gathering, treatment and disseminationof information in the market. By providinginvestors with an independent opinion on anissuer’s creditworthiness, credit ratings allowthe imbalance of information between issuersand investors to be reduced. Hence ratings canhelp to broaden the issuer’s investor base,potentially adding the benefit of morefavourable financing conditions. Furthermore,even sophisticated investors may not have theexpertise or knowledge to analyse the myriad ofpotential debt investments available to themacross different industries and countries.Having a simple and easily understandablesummary of a credit assessment in the form of arating symbol (e.g. AAA) assists them inselecting the bonds they want to invest in.Finally, ratings contribute to the process ofprice formation in the bond market, as shown byseveral academic contributions on the linkbetween ratings and credit spreads70.

The advent of the euro and the integration ofEuropean financial markets conferred an evenmore determinant role to credit ratings. Byeliminating currency risk, the use of the euroallowed bond investors to focus on credit riskwhile the enlargement of their investmentuniverse increased their need for simpleindicators of this risk. Indeed, investorsdiversified their portfolios and turned to bondsfrom the eleven other euro area countries.

However, as they knew issuers from thesecountries less than issuers from their homecountry, they required credit ratings as acondition for them entering the market. Hence,the euro naturally broadened the investor baseof issuers but at the expense of increasedcompetition, a good rating increasinglybecoming a quasi-prerequisite for raising fundsin the euro bond market.

However, the slow financial disintermediationprocess in Europe explains why, despite thestrong drivers to a greater use of ratings, thelatter remains limited compared to the US.According to the Bank for InternationalSettlements, bank loans represent around 50 to70% of banks’ financial assets in Europeagainst 25% in the US. Banks theoretically havethe capacity to assess and monitor internallytheir debtors’ creditworthiness and hence havelittle need for external credit ratings.Furthermore, the traditional approach ofcorporate finance in Europe i.e. based on closeand continuous relationships between creditorsand borrowers, was perceived in manycountries as mitigating the need for anindependent credit assessment. Therefore,corporations started to ask for credit ratingschiefly when they started to tap the bond marketeither to get more financing or to obtain it onbetter terms.

Finally, the landscape of the use of creditratings in the European bond markets would notbe complete without mentioning the dramaticgrowth of the structured finance market (i.e.ABSs, MBSs and CDOs) since the late 1990’s.This market is in nature a fully rated market, inthe absence of a real “issuer”. Therefore, therating is essential as an opinion on the credit

68 For more details, see Richard Sylla, “A Historical Primer on theBusiness of Credit Ratings”, conference on “The Role of CreditReporting Systems in the International Economy”, World Bank,March 2001.

69 On the expanding use of ratings, see Banque de France, FinancialStability Review, June 2004, and ECB working paper No 16,“Market dynamics associated with credit ratings: a literaturereview”.

70 For example R. Cantor and F. Packer, “Sovereign Credit Ratings”,Current Issues in Economics and Finance, Vol 1, No 3, FederalReserve Bank of New York, June 1995, pp 1-6.

65ECB


5 MARKETTRENDS

quality of both the underlying assets and thestructured transaction.

ASSESSING CREDIT RATING COVERAGE INEUROPE

Measuring credit rating coverage in Europeentails many practical difficulties. Deriving aratio of coverage of credit ratings is not trivial,starting with the difficulties with the numerator,i.e. what is meant by credit rating and endingwith the denominator, i.e. to which universe ofcompanies should the number of ratedcompanies be compared71.

Regarding the numerator, credit ratings in thisstudy will be understood as issuer ratingsawarded by the three major rating agencies. Whileindustry sources count more than 150 ratingagencies72, only three of them (Moody’, S&P,Fitch) have the capability to rate all types ofissuers on a global scale. Additionally, the choiceof the issuer rating, as opposed to asset ratings,simplifies the data gathering, and still offers aconsistent measure since it corresponds to therating of a senior unsecured bond of the issuer.Moreover, it should be borne in mind that acomparison between the three ratings agencies’figures is of limited significance due to the

differences in rating methodologies. Turningnow to the denominator, it can take many forms(see Table 10). The number of rated companiescan be compared to the number of issuers in thebond market (including unrated bond issuers) orto the number of companies listed on the stockexchange, as these two broad categories canbe viewed as proxies for market-financedcompanies. The comparison with the number ofcompanies with a turnover above a certainthreshold can also be meaningful, as the higherthe size and the turnover of the company, the morelikely it is to ask for a rating and/or issuemarketable debt.

Unsurprisingly, the coverage in Europe appearssubstantially less extensive than in the US.When considering the EU-wide turnoverthreshold for defining large-size enterprises(€50 mn73), around 10% of large-sizeenterprises are rated (vs. 65% in the US).However, this proportion increases to close to

71 For more insights into the difficulty of assessing ratingpenetration, see Basel Committee on Banking Supervision, reportof the working group on “Credit ratings and complementarysources of credit quality information”, August 2000.

72 For more information about other rating agencies, see BaselCommittee supra.

73 €50 million is also used for distinguishing a large-size companyin the Basle II framework.

EU-15 EU-15 EU-15 USA USA USAMoody’s S&P Fitch Moody’s S&P Fitch

No of rated companies 1) 1,223 1,352 650 2,991 4,189 1,230As a % of bond issuers 2) 57 63 30 60 84 25As a % of listed companies 23 25 12 57 80 23As a % of companieswith turnover above € 50 m 10 11 5 65 92 27

Table 10 Coverage of issuer ratings in Europe and the US in July 2004

Source: Moody’s; S&P; Fitch; World Federation of Stock Exchanges; Bureau Van Dijk; TRACE database; Bloomberg.Notes: Issuer Ratings are understood as long-term senior issuer ratings. Financial Strength Ratings (for banks and insurance companies)are excluded. General government issuers (i.e. central, regional and local governments) are excluded. Published ratings awarded on thebasis of public information (sometimes referred to as “unsolicited ratings”) are included. This type of rating mainly concerns theinsurance and corporate sector in Europe, depending on the agencies.1) In the table, the number of Long-term issuer ratings covers differences in methodologies as regards in particular the insurance sector.Whereas S&P grants a Long Term issuer rating to any insurance company to which it grants a Financial Strength rating – FSR –, Moody’sand Fitch do not systematically grant long-term ratings to insurance companies to which they have granted an FSR but only to those boundto issue bonds. FSR ratings for the insurance sector are respectively in EU-15 (515 for S&P, 128 for Fitch and 106 for Moody’s). The highnumber of S&P insurance ratings also comes from a higher proportion of ratings based on public information (sometimes referred to as“unsolicited ratings”).2) It is possible that data on bond issuers are subject to some double counting effects due to different name spelling which deflatessomewhat the calculated percentage.


25% when taking as a reference the number oflisted companies. More puzzling is the still highnumber of unrated issuers in the European bondmarket (e.g. according to Bloomberg, 237unrated issuers from the industrial sector).European corporates, unlike their UScounterparts, can still rely on “domestic namerecognition” for raising funds without beingrated. Also, publicly-owned companies can stillrely on State sponsoring.

CREDIT RATING COVERAGE BY SECTORThe industry sector profile of rated entities inEurope has changed substantially over the pastten years, showing increased diversification interms of issuer types. It nevertheless remainsfairly concentrated relative to the US, notablyon the banking sector.

The coverage by sector in Europe evidencesthree trends:

– While all central governments have beenrated for years if not decades, the number ofregional and local governments asking forratings has been rapidly increasing in thewake of decentralisation in some MemberStates (e.g. Italy);

– Banks and other financials (includinginsurance) still represent a large proportionof rated issuers in Europe, togetheraccounting for 40% of the total (vs. 25% in

Other financialsBanks (including insurance) 1) Corporates General government

Moody’s S&P Fitch Moody’s S&P Fitch Moody’s S&P Fitch Moody’s S&P Fitch

EU-15 472 304 395 82 515 41 669 533 214 133 182 79US 295 400 484 480 1,636 222 2,216 2,028 524 n/a n/a n/a

Source: Moody’s; S&P; Fitch.Notes: For the US, general government is not relevant for comparison purposes as it would correspond to the “US public finance” sectorincluding states, municipalities, universities etc, e.g. some 22,000 issuers for Moody’s.1) In the table, the number of Long-term issuer ratings covers differences in methodologies as regards in particular the insurance sector.Whereas S&P grants a Long Term issuer rating to any insurance company to which it grants a Financial Strength rating – FSR –, Moody’sand Fitch do not systematically grant long-term ratings to insurance companies to which they have granted an FSR but only to those boundto issue bonds. FSR ratings for the insurance sector are respectively in EU-15 (515 for S&P, 128 for Fitch and 106 for Moody’s). The highnumber of S&P insurance ratings also comes from a higher proportion of ratings based on public information (sometimes referred to as“unsolicited ratings”).

Table 11 Issuer ratings by sector and region in July 2004

the US). Banks are massively rated, giventhe importance of ratings-based marketpractices (e.g. assessing creditworthinessand limits in financial transactions on thebasis of ratings);

– Corporate rating coverage is clearly laggingbehind compared to the US, but it is quicklyexpanding. For example, the number ofEuropean corporates rated by Moody’s hasmore than doubled between 1997 and 2003from 242 to 512. Part of the lag with the UShas regulatory grounds since any corporate’sissuance above USD 50 millions must berated by at least two Nationally RecognizedStatistic Rating Organizations (NRSRO).

CREDIT RATING COVERAGE: COMPARISONSBETWEEN EUROPEAN COUNTRIESIn the same way that rating penetration isdifferent in the United States and the EU, it isalso different within the EU. This is vastly aresult of the existence of different financialstructures (see Chart 21) with more or lessfinancial disintermediation – as measured bycorporate bond issuance as a percentage ofGDP. The size of the country seems to play arole as well, as an issuer can hardly issue inthe bond market below a critical mass of EU500 million or €1 billion per issue. Overall,three groups of countries emerge in decreasingorder of rating penetration: a first group ofsignificantly disintermediated economies (the

67ECB


5 MARKETTRENDS

United Kingdom, the Netherlands); a secondgroup of large-size but bank-based economies(DE, FR) and finally a group of smaller bank-based economies.

DRIVING FACTORS OF THE USE OF CREDITRATINGS IN EUROPE

Albeit remaining modest, the coverage of creditratings in Europe is expanding very quickly74.Yet, rating agencies estimate that there is still ahuge potential for growth. Moody’s estimatesthat there are 1,500 unrated institutions inEurope with annual revenues of at least€1 billion. Notwithstanding the continuousdisintermediation trend, other structural factorsmight drive the use of ratings in Europe in theyears ahead:

– The ongoing internationalisation/europeanisation of fixed income portfolios:As the geographical diversification ofportfolios is not over, it will remain a strongsource of demand for ratings in the Europeanbond market. At year-end 2002, the IMFCoordinated Portfolio Investment Survey(CPIS) indicates that euro area investorsrepresent on average 64% of foreigninvestors in euro area bonds.

– The regulatory demand for ratings: In thepast, regulatory demand for ratings has

Chart 21 Number of issuer ratingsand disintermediation rate in EU-15in July 2004

Sources: Moody’s; S&P; BIS; IMF.Note: General government issuers are excluded.

0

50

100

150

200

250

300

350

400

0

5

10

15

20

25

30

35

no Moody’s rated issuers (left-hand scale)no S&P rated issuers (left-hand scale)corporate bond issuance (in % of GDP; right-hand scale)

AT BE DK FI FR DE GR IE IT LU NL PT ES SE UK

given a strong support for the use of ratings.Currently, ratings are required by regulatorsin a variety of fields75. In the next few years,the reform of the capital accord (Basle II)will surely affect the need for and use ofratings, although it is still difficult to predictprecisely in which way.

– New market-based methods to assess creditrisk: New methods derived from equityprices, subordinated bond prices or CDSsare emerging as quasi-substitutes for ratings.They reflect the “market” view of thecreditworthiness of a corporation, ratherthan the opinion of a particular agency. Theyare increasingly used, including by theagencies themselves for cross-checkingpurposes. However, it remains to be seenwhether such measures can be meaningfulindependently from ratings.

5.8 IMPACT OF THE FSAP ON THE BOND MARKETAND PROGRESS TOWARDS INTEGRATION

This chapter highlights some major results ofthe Financial Services Action Plan (FSAP) ofrelevance for the European bond market.Furthermore, it summarises the current state ofcompletion of the overall FSAP.

The FSAP, adopted by the EuropeanCommission in 1999, and endorsed by theLisbon European Council in March 2000,presents to date the most ambitious initiative tofoster the integration of capital markets and toachieve a single market for financial services inthe EU. The four strategic objectives underlyingthe FSAP relate to the single EU market forwholesale financial services, open and secureretail markets, state-of-the-art prudential rulesand supervision, and wider conditions for anoptimal single financial market (namely tax and

74 This is also evident from the geographical split of revenuegeneration by rating agencies. Moody’s generates only 25% of itsrevenues in Europe, but European revenues increased onaverage by 37% annually in the period from 1997 to 2003.

75 See European Parliament, “Report on role and methods of ratingagencies”, Draft report, January 2004 and Banque de FranceFinancial Stability Review June 2004.


corporate governance issues). The mostimportant measures of relevance for theEuropean bond market that have so far beenadopted, are the following.

– Directive 2002/47/EC on financial collateralarrangements. This Directive was adopted on6 June 2002 with a transposition deadlinethat passed on 27 December 2003. It isrelated to the part of the FSAP on measuresagainst systemic risks in securitiessettlement. The aim of the Directive was firstto eliminate the substantive legal differencesbetween Member States’ civil andinsolvency law which impede cross-bordersecurities transactions. The Directive alsoseeks to resolve the main problems affectingcross-border use of collateral in wholesalefinancial markets. It proposes to abolishexisting administrative burdens andcomplexities, creating a clear framework oflegal certainty in the field of collateral by(a) ensuring that an effective and simpleCommunity regime exists for the creation ofcollateral; (b) providing limited protection ofcollateral arrangements from some rules ofinsolvency law, particularly those that wouldinhibit the effective liquidation of collateralor cast doubt on the validity of techniquescurrently used; (c) creating legal certaintywith regard to cross-border provision ofcollateral, in the form of book-entrysecurities, by extending the principlesalready applied under the Settlement FinalityDirective to determine where such securitiesare located; (d) restricting the imposition ofonerous formalities on either the creation orthe enforcement of collateral arrangementsand ensuring effective agreements permittingthe collateral taker to re-use the collateral fortheir own purposes under pledge structures,the classic way of providing collateral.

– Directive 2003/71/EC on prospectuses of4 November 2003. The transposition deadlinefor this Directive is 30 June 2005. ThisDirective belongs to the part of the FSAP onmeasures for raising capital on an EU-widebasis and addresses issues related to the

prospectuses, i.e. the disclosure documents,containing key financial and non-financialinformation which a company makes availableto potential investors when it issues securitiesto raise capital and/or when its securitiesshould be admitted to trading on exchanges.This Directive will make it easier and cheaperfor companies to raise capital throughout theEU on the basis of approval from a regulatoryauthority (“home competent authority”) inone Member State. It will reinforce protectionfor investors by guaranteeing that allprospectuses, wherever issued in the EU,provide them with the clear and comprehensiveinformation that they need.

– Directive 2004/39/EC of 21 April 2004on markets in financial instruments. Thetransposition deadline for this Directive isthe end of May 2006. This Directive, whichcorresponds to the part of the FSAP onmeasures for establishing a common legalframework for integrated securities andderivatives markets, updates the formerInvestment Services Directive and is a corecomponent of the FSAP. Its objective is togive investment firms an effective “singleEU passport”, allowing them to operatethroughout the EU on the basis ofauthorisation in their home Member State.Whilst stimulating investment within the EU,the legislation also aims to ensure investorprotection by providing more preciserules on the conduct of investment firms’business, reinforced “best execution”obligations, and a comprehensive pre andpost-trade transparency regime. For the timebeing, this obligation is restricted to shares,but it could be extended to bonds in thefuture. The new Directive also sets out toensure fair competition between regulatedmarkets (stock exchanges), multilateraltrading facilities or alternative tradingsystems and off-exchange transactions. Inthis respect, investment firms and banks willbe allowed to “internalise” client orders, i.e.to execute them in-house without goingthrough a regulated market in accordancewith the requirements of the directive.

69ECB


5 MARKETTRENDS

– Commission Communication on Clearingand Settlement issued on 28 April 2004. Thisconsultative communication outlines thedirection for further work in order to achievean efficient, integrated and safe market forsecurities clearing and settlement which iscrucial for an integrated and efficientEuropean capital market. The ultimate targetis a framework Directive for pan-Europeanclearing and settlement, aimed at increasingthe efficiency and safety of cross-borderclearing and settlement, while at the sametime ensuring a level playing field among thedifferent providers of clearing and settlementservices. As already highlighted by the twoGiovannini reports of 2001 and 2003, thepresent market infrastructure for cross-border securities settlement poses barriers tofurther integration. The Commissionidentified three core principles for theproposed framework Directive, that are to bemore clearly defined in accordance with thecomments received in the consultation.These core principles encompass (i) thegranting of comprehensive rights of non-discriminatory access to and choice ofclearing and settlement systems in the EU;(ii) a common regulatory and supervisoryframework for the authorisation,recognition, regulation and supervision ofsecurities clearing and settlement systems;and (iii) appropriate governancearrangements and transparency requirementsfor securities settlement systems and centralcounterparties, as these can become a sourceof instability for the financial system in theevent of a higher degree of consolidationthroughout the EU. The ECB, among otherparties that were invited to submit theirviews, published the Eurosystem’s responseon 29 July 2004. In its answer, the ECBmade reference to, among other things, theESCB/CESR standards for securitiesclearing and settlement systems that couldusefully provide the basis for the upcominglevel 2 comitology work.

– The Proposal for a Directive on transparencyrequirements for securities issuers approved

by the ECOFIN Council on 11 May 2004.Entry into force of this Directive isscheduled for autumn 2004, with atransposition deadline two years later. ThisDirective improves financial reporting bysecurity issuers and should also lead tobetter dissemination of information onissuers among Member States.

As such, the measures foreseen in the FSAP,with a final deadline of implementation by2005, have been delivered in full and on time.Naturally, regulation alone does not guaranteean integrated single market. The measuresconcluded under the FSAP have to beimplemented at Member State level in aconsistent and timely manner, and nationalsupervisory practices should converge. It isstill too early to finally evaluate to what extentthe FSAP has achieved its objectives. With thisin mind, in 2003 the European Commissioninitiated a monitoring process to assess the stateof integration of European financial markets.One line of assessment is being followed withthe help of four expert groups of marketparticipants. The four sectoral groups relatingto banking, insurance, securities and assetmanagement were set up in November 2003 anddelivered their respective reports in May 2004,open for public comment until 10 September2004. Another line of assessment is theCommission’s first annual “FinancialIntegration Monitor” report published in May2004. It provides initial documentation ofchanges in the level of cross-border integrationin key financial segments over recent years.Furthermore, in June 2004 the FinancialServices Committee, a forum composed ofrepresentatives of the Ministers of Financefrom the Member States, a collective view ofintegration of financial services in the EU andon the remaining challenges, for considerationby the ECOFIN Council.

As concerns the euro area bond market, a recentstudy proposing a number of measures toquantify the state and evolution of financialintegration in the euro area came to theconclusion that a rather high degree of


integration can be observed.76 Nonetheless,further integration may be possible. First,yields of similar or identical credit risk andmaturity have not entirely converged, possiblydue to differences in liquidity and availability ofrelated derivative markets. Second, whilegovernment bond yields are now mainlydetermined by common news, idiosyncraticfactors continue to be relevant, in particular in

the two-year and five-year maturity segments,but less so for ten-year government bonds (seeBox 5).

76 See Baele et al, ”Measuring financial integration in the euroarea”, European Central Bank, Occasional paper No 14, April2004. It is noted that this study refers to the euro area financialmarkets and not to the euro-denominated financial markets. Thelatter are located both within and outside the euro area and arethe subject of the MOC bond market study.

Box 5

DEGREE OF INTEGRATION IN THE EURO AREA MARKETS FOR GOVERNMENT BONDS

One way of monitoring the degree of integration in financial markets is to do so directly on thebasis of market prices. The more closely the market prices for comparable financial instrumentsconverge (law of one price), the more integrated the financial markets are deemed to be. In fullyintegrated markets, the market participants have access to the same information and face thesame transaction costs. Furthermore, the pricing should depend solely on the structure of thefinancial instrument and not, for example, on the place of issuance or custody.

The extent to which the euro area markets for government bonds are already integrated isillustrated by the reaction of European government bonds to fluctuations in the price of selectedbenchmark bonds. It is assumed that the price movements of both the benchmark bonds and theother bonds in the euro area are based on common factors affecting prices. The chart belowshows the average regression coefficients which are measured by the reaction of Europeangovernment bonds to a 1% price change in German government bonds. The estimates relate to

national price indices calculated by Bloombergfor European government bonds with amaturity of seven to ten years. The regressioncoefficients are the result of estimates usingweekly data and a moving two-year window,and, owing to the design of the estimationapproach, they can be measured directly aselasticity.

The nearer the coefficient is to 1, the moreclosely the European government bonds trackmovements in the price of the Germanbenchmark bonds. It is striking that sincemonetary union came into force in 1999,

movements in European government bond yields have increasingly coincided. The elasticity ofEuropean government bonds with regard to movements in the price of the German benchmarkbonds is currently 0.98. Viewed in terms of prices, the integration of the euro area markets forgovernment bonds is remarkably far advanced.

Average elastic ity of European governmentbonds with respect to German sovereigns

Sources: Bloomberg, own calculations.

0.75Jan.

1995 1996 1997 1998 1999 2000 2001 2002 2003July Jan. July Jan. July Jan. July Jan. July Jan. July Jan. July Jan. July Jan. July Jan.

2004

0.80

0.85

0.90

0.95

1.00

0.75

0.80

0.85

0.90

0.95

1.00

71ECB


ANNEX A

A NN EX A

QUANT I TAT I V E ANA LY S I S O F CORPORAT E BONDS PR E AD DYNAM I C S

Vector error correction estimates: Sample: June 1999 – December 2003; 55 observations; t-statistics in [ ]Cointegration restrictions: B(1,1)=1, B(1,3)=0, B(1,5)=0; LR test for binding restrictions (rank = 1):Chi-square(2) 2.455Probability 0.293

Cointegrating equation: CointEq

BBB-spreads(-1) 1.0003M-MONEY(-1) -78.527

[-9.244]GDP_GROWTH(-1) 0.000DTE-RATIO(-1) -66.640

[-6.094]ISSUES (-1) 0.000C 312.433

D(BBB- D(3M- D(GDP- D(DTE-Error correction: SPREADS) MONEY) GROWTH) RATIO) D(ISSUES)

CointEq -0.488 -0.000 3.05E-05 0.003 -0.001[-3.885] [-0.111] [ 0.033] [ 3.387] [-0.857]

D(BBB-SPREADS(-1)) 0.307 -0.002 0.000 0.000 0.001[ 2.101] [-1.077] [ 0.264] [ 0.266] [ 0.584]

D(BBB-SPREADS(-2)) -0.143 0.002 -0.001 -0.000 0.002[-0.988] [ 1.294] [-1.017] [-0.110] [ 1.378]

D(BBB-SPREADS(-3)) 0.206 -0.002 -0.000 -0.001 0.001[ 1.424] [-1.458] [-0.403] [-0.463] [ 0.606]

D(3M-MONEY(-1)) -6.519 0.579 0.149 0.258 0.036[-0.314] [ 2.876] [ 0.970] [ 1.551] [ 0.168]

D(3M-MONEY(-2)) -23.705 -0.101 -0.051 0.292 -0.175[-1.233] [-0.541] [-0.359] [ 1.896] [-0.883]

D(3M-MONEY(-3)) -35.721 0.106 -0.133 0.288 -0.001[-1.745] [ 0.533] [-0.880] [ 1.757] [-0.003]

D(GDP-GROWTH(-1)) 76.347 -0.104 0.785 -0.308 -0.018[ 3.187] [-0.448] [ 4.423] [-1.605] [-0.073]

D(GDP-GROWTH(-2)) -2.457 0.105 0.101 -0.100 -0.025[-0.083] [ 0.367] [ 0.460] [-0.422] [-0.082]

D(GDP-GROWTH(-3)) -15.557 0.147 -0.141 -0.227 0.170[-0.689] [ 0.672] [-0.844] [-1.256] [ 0.733]

D(DTE-RATIO(-1)) 7.516 -0.165 0.165 -0.000 -0.462[ 0.415] [-0.943] [ 1.232] [-0.003] [-2.479]

D(DTE-RATIO(-2)) -9.668 0.128 -0.140 0.206 -0.013[-0.509] [ 0.696] [-0.998] [ 1.353] [-0.066]

D(DTE-RATIO(-3)) 22.204 -0.118 -0.041 0.024 -0.287[ 1.321] [-0.723] [-0.326] [ 0.177] [-1.659]

D(ISSUES(-1)) -23.164 -0.0433 -0.006 -0.011 -0.713[-1.578] [-0.305] [-0.053] [-0.090] [-4.722]

D(ISSUES(-2)) 4.481 -0.026 0.030 -0.355 -0.436[ 0.26] [-0.159] [ 0.235] [-2.584] [-2.471]

D(ISSUES(-3)) -3.263 -0.068 0.001 -0.205 -0.608[-0.197] [-0.427] [ 0.006] [-1.549] [-3.581]

C 0.456 0.004 -0.009 -0.001 0.032[ 0.203] [ 0.161] [-0.513] [-0.045] [ 1.364]

R-squared 0.580 0.387 0.669 0.457 0.518 Adj. R-squared 0.403 0.128 0.529 0.229 0.316Jarque-Bera test for multivariate normality: p-value=0.1091; LM-test (AR=4): p-value=0.4779.

Notes: BBB-SPREADS: Merrill Lynch BBB corporate bond spreads, defined as the sum of the option-adjusted spread of BBB corporatebonds with 7-10 years to maturity and the 10-year swap spread in basis points. 3M-MONEY: 3-month EURIBOR. GDP-GROWTH: annualgrowth rate of EU real GDP. Monthly values are interpolated from quarterly observations. DTE-RATIO: debt to earnigs ratio of the non-financial corporate sector, approximated by the log ratio of the sum of non-financial corporate debt (loans to and outstanding amount ofdebt securities issued by non-financial corporations) relative to a proxy of “corporate earnings” (DJ Euro Stoxx50 divided by thecorresponding price-earnings ratio). ISSUES: relative gross issues of debt securities issued by non-financial corporations relative tothose issued by the government (central government plus other government).


The empirical analysis is restricted to financial and non-financial companies, which are the mostimportant reference entities in the European CDS market. In order to explore long-termrelationships and short-term dynamics in pricing credit risk, vector error correction models(VECMs) using Johansen’s procedure to test for the existence of cointegrating relationships areestimated for each reference entity:

ANNEX B

EMP I R I C A L E V I D ENC E O F P R I C E D I S COV ERY I NTH E CD S AND BOND MARKE T S

N obs C ααααα (t-stat) λλλλλ 1 (t-stat) λλλλλ 2 (t-stat) GG

Barclays 501 -8.599 -2.027 -0.007 0.073 0.91(-6.83) (-1.35) (4.73)

Commerzbank 714 19.46 -0.998 0.015 0.070 1.27(-17.44) (1.61) (7.68)

Deutsche Bank 624 1096.4 -45.78 0.000 0.005 0.99(-7.18) (-0.25) (7.17)

San Paolo 522 31.82 -1.088 0.000 0.072 1.00(-4.10) (0.07) (3.73)

Alcatel 651 125.2 -1.446 -0.076 0.020 0.21(-28,25) (-5.01) (2.10)

Deutsche Telekom 503 54.05 -1.628 -0.003 0.027 0.90(-20.05) (-0.24) (3.34)

France Telecom 399 1.623 -0.976 0.002 0.080 1.03(-24.61) (0.14) (3.95)

Telefonica 507 -21.2 -0.798 -0.018 0.042 0.70(-12.35) (-1.46) (3.79)

DaimlerChrysler 507 -36.7 -1.056 -0.061 0.017 0.21(-8.87) (-2.96) (1.15)

Renault 505 0.59 -1.000 -0.088 0.082 0.48(-31.26) (-3.14) (3.48)

Parmalat 270 -50.5 -1.005 -0.171 0.066 0.28(-20.40) (-3.17) (3.07)

Endesa 508 12.660 -1.448 -0.055 0.022 0.29(-18.82) (-2.47) (1.60)

RWE 408 63.83 -2.796 -0.016 0.066 0.81(-18.21) (-1.26) (4.03)

Vattenfall 408 8.848 -1.277 0.004 0.050 1.09(-8.38) (0.39) (2.90)

Total 518 6.835 -2.035 -0.004 0.098 0.96(-7.70) (-0.63) (4.97)

Note: European corporates’ CDS and bond data are taken from Bloomberg. CDS spreads for five years’ maturity were chosen, since thisis the benchmark maturity in the CDS market. The corresponding corporate bond yields were calculated as a weighted interpolation ofbond yields with maturities of longer and shorter than five years in order to derive bond yields for five years’ maturity for each point intime. The bonds included are exclusively liquid euro-denominated bonds with fixed rates which are not puttable, callable, convertible,subordinated or structured. Bond spreads are calculated as the difference between the generated bond yields and the five-year euroswaps rate. All data refer to mid-market prices on a daily basis covering the period from October 2001 to June 2004.

Table 12 Estimation results

)( 1,1

1,1

11,1,1, titCSi

iitCDSi

itCStCDStCDS pppcpp εδβαλ +∆+∆+−−=∆ −=

−=

−− ∑∑

with �1t

and �2t as independent, identically distributed (IID) residuals.

The coefficients �1 and �

2 measure the speeds at which the CDS market and the bond market,

respectively, adjust towards equilibrium. Hence, if �1 turns out to be (significantly) negative, CDS

prices adjust towards the long-run relationship. This means, that the CDS prices are following the

(1)

(2) )( 2,1

2,1

21,1,2, titCSi

iitCDSi

itCStCDStCS pppcpp εδβαλ +∆+∆+−−=∆ −=

−=

−− ∑∑q q

q q

73ECB


ANNEX B

bond prices temporarily and the bond market is therefore leading in terms of price discovery.Whereas, if �

2 is (significantly) positive, the bond market reacts to deviations from equilibrium

and the CDS market accounts for price discovery.

The Johansen cointegration test indicates for 68% of the analysed companies a cointegratingrelationship at the 5% level. For 32% of reference entities no unique cointegrating relationshipcould be found, which may be caused by the aforementioned factors affecting the arbitragerelationship or, to be more speculative, might be interpreted as indicating imperfect integration.The estimation results – including the coefficients of the cointegrating vector [1, a , c] – arepresented in Table 12 and refer to entities with a unique long-term relationship. The lag length forthe dynamic part of the VECMs is determined by Akaike’s information criterion.

An illustrative way to present the contribution to price discovery is the Gonzalo and Grangermeasure GG = �

2 / (�

2 - �

1). For GG < 0.5 the bond market leads with respect to price discovery,

and for GG > 0.5 the CDS market makes the greater contribution to price discovery.

At 87%, the CDS market plays an important role in price discovery (�2 is significantly positive),

whereas the bond market contributes to price discovery in 33% of all cases (�1 is significantly

negative). The Granger and Gonzalo measure indicates that the CDS market is leading in pricingcredit risk in 67% of reference entities, whereas the bond market is more relevant in 27%. In oneregression (Renault) the results are ambiguous.


G L O S S A RYAlternative trading systems (ATSs): systems that offer other means of trading than establishedexchanges. They operate electronically (lowering transaction costs) and focus on services that arenot always provided by established exchanges (e.g. central limit order book, after hours trading ordirect access for institutional investors).

Arbitrage: profiting from differences in prices when the same security, currency or commodity istraded in two or more markets.

Asset-backed securities (ABS): bonds which are generated by “special purpose vehicles” inorder to transform illiquid assets of a certain corporation (the “originator”) into transferablesecurities. ABS are issued in several tranches with different credit quality.

Assets under management: assets managed by a financial institution which are beneficiallyowned by clients.

Bank certificates of deposit (CDs): short-term securities issued by banks.

Benchmark: value used as a reference or means of comparison for measuring the performance ofan investment.

Benchmarking: basing the investment allocation on an industry standard and/or on a fixedsecurities index.

Bid-ask spread: differential prevailing in the market between the bid price and the offered price.

BIS: Bank for International Settlements.

Bon à taux fixe (BTF): French Treasury bill.

Bond rating: ranking of a bond’s quality in terms of default risk. Bonds are rated from a high ofAAA (highly unlikely to default) to a low of D (issuer already in default).

Bonos del Estado (Bonos): Spanish Treasury bonds with an original maturity of between two andfive years.

Bons du Trésor à taux fixe et à intérêt annuel (BTAN): negotiable fixed-rate medium-termFrench Treasury notes with annual interest. On issue their maturity is either two or five years.

Broker: firm which operates in a market on behalf of other participants to arrange transactionswithout being a party to the transactions itself.

Bubill: German Treasury bill.

Buoni Ordinari del Tesoro (BOT): Italian Treasury bill.

Buoni Poliennali del Tesoro (BTP): Italian Treasury bonds with an original maturity of three tothirty years.

Capitalisation: see Market capitalisation.

75ECB


GLOSSARY

Central counterparty: an entity which interposes itself as the buyer to every seller and as theseller to every buyer of a specified set of contracts.

Certificati di Credito del Tesoro zero coupon (CTZ): Italian government debt instrument issuedat discount with an original maturity of up to two years.

Certificati di Credito del Tesoro (CCT): Italian Treasury floating rate securities with a seven-year original maturity.

Central securities depository (CSD): a facility for holding securities which enables securitiestransactions to be processed by book entry. Physical securities be immobilised by the depositoryor securities may be dematerialised (i.e. so that they exist only as electronic records). In additionto safekeeping, a central securities depository may incorporate comparison, clearing and settlementfunctions.

Clearing: the process of transmitting, reconciling and, in some cases, confirming the paymentorder and the securities transfer prior to settlement. In the context of repos, this can have threeseparate aspects: confirmation/matching, netting and clearing with the central counterparty.

Cointegration: in econometrics non-stationary time series are cointegrated if there exists a linearrelationship that is stationary. The stationary linear combination is called the cointegratingequation and may be interpreted as a long-run equilibrium relationship among the variables.

Collateralised debt obligation (CDO): a structured debt security backed by a portfolio of assets;see synthetic collateralised debt obligation.

Commercial paper (CP): short-term obligations with maturities ranging from 2 to 270 daysissued by banks, corporations and other borrowers. Such instruments are unsecured and usuallydiscounted, although some are interest bearing.

Confirmation/matching: the process of ensuring that the two counterparties agree with regard tothe terms of the transaction – price, asset(s), value dates, settlement data, including relevantaccount numbers – before the payment and transfer orders are sent for settlement.

Convertible bond: bond exchangeable for equity at a set price.

Corporate bond spread: difference between the interest rate of fixed income instruments issuedby corporations and the rates of debt obligations without default risk.

Counterparty: the opposite party in a financial transaction.

Credit default swap: credit derivative transaction in which the buyer of protection can insurehimself against default by a particular issuer for a specific period by paying a periodic premium tothe protection seller.

Credit derivative: an OTC derivative designed to transfer credit risk from one party to another.

Credit risk: the risk that a counterparty will not settle an obligation for the full value, either whendue or at any time thereafter.


Credit linked note (CLN): debt security which is fully paid back only if an agreed credit eventhas not occurred, otherwise the repayment is reduced by the agreed contingent payment.

Credit spread option: option to buy a bond at the future market price and obtaining the priceassociated with an agreed “strike spread”.

Cross margining: netting of margin requirements for the simultaneous sale and purchase ofsecurities.

Crossing network: system which matches buy and sell at a price determined in another market.Such systems play no part in price discovery.

Currency risk: the risk that the operations of a business or the value of an investment will beaffected by changes in exchange rates.

Dealer: firm whose primary business is entering into transactions on both sides of wholesalefinancial markets and seeking profits by accepting risks in these markets.

Defined benefit insurance policies: policies for which the benefits are defined ex ante.

Depository: an agent with the primary role of recording securities either physically orelectronically and keeping records of ownership of these securities.

Derivative: a financial contract whose price depends on the value of one or more underlyingreference assets, rates or indices. For analytical purposes, all derivatives contracts can be dividedinto basic building blocks of forward contracts or options or combinations thereof.

Disintermediation: the investment in or borrowing from financial markets directly, without theuse of financial intermediaries such as banks.

Electronic trading: in broad terms, this refers to any use of electronic means of sending orders(bids and offers) to the market where their automated execution is performed.

EMU: Economic and Monetary Union.

End-user: an entity that takes positions for investment or hedging purposes. An end-user oftendeals on one side of the market only.

EU: European Union.

EURIBOR: the euro area interbank offered rate for the euro, sponsored by the European BankingFederation (EBF) and the Association Cambiste Internationale (ACI). It is an index price sourcecovering dealings from 48 prime banks.77

Euro overnight index average (EONIA): the overnight rate computed as the euro area interbankoffered overnight rate for the euro. It is computed as a weighted average of all overnight unsecured

77 Number of panel banks as of November 2004.

77ECB


GLOSSARY

lending transactions in the interbank market, initiated within the euro area by the contributingpanel of 48 prime banks.78

Euronext: company created on 22 September 2000 through the merger of the exchanges ofAmsterdam, Brussels and Paris. The Lisbon and Porto exchanges joined later.

European Master Agreement: legal contract sponsored by the European Banking Associationand the European Savings Association which aims to consolidate into a single set of harmoniseddocuments various master agreements used within the euro area and certain neighbouringcountries, particularly for repurchase transactions and securities lending.

European System of Central Banks (ESCB): the European Central Bank and the national centralbanks of the EU Member States.

Eurosystem: the European Central Bank and the national central banks of the EU Member States,which have adopted the euro.

Exchange traded funds (ETFs): a fund that is traded on a stock exchange like a share.

Financial services action plan (FSAP): elaborated and adopted by the European Commission in1999, the action plan was endorsed by the Lisbon European Council in March 2000, which set adeadline for its implementation of 2005 at the latest. Its purpose is to improve the single market infinancial services in order to reap the full benefits of the euro. The action plan suggests indicativepriorities and time-scales for legislative and other measures to tackle three strategic objectives,namely ensuring a single market for wholesale financial services, open and secure retail marketsand state-of-the-art prudential rules supervision.

Futures: agreement to buy or sell a specific amount of a commodity or financial instrument at aparticular price on a stipulated future date.

General collateral: collateral which, owing to its homogeneous features, is broadly accepted.

Hedge fund: private investment partnership whose offering memorandum allows for the fund totake both long and short positions, to use leverage and derivatives, and to invest in many markets.

Hedging: strategy to offset investment risk.

High-yield bonds: bonds rated BB and below.

Home bias: tendency of investors to limit their holdings to the domestic market.

Investment grade bonds: bonds rated at least BBB.

International central securities depository (ICSD): a central securities depository which clearsand settles international securities or cross-border transactions in domestic securities.

Junk bond: high-yield bond with a credit rating of BB or below.

78 Number of panel banks as of November 2004.


Liquid (market): the three aspects of liquidity are tightness in bid-ask spreads, depth andresiliency. It is characterised by the ability to transact in a market without significantly movingprices.

M&A: mergers and acquisitions.

Market capitalisation: value of a corporation as determined by the market price of its issued andoutstanding common stock.

Market maker: dealer obligated to quote buy and sell prices in return for certain privileges withina market (sometimes used to refer to anybody who provides quotes).

Monetary financial institutions (MFIs): financial institutions forming the money-issuing sectorof the euro area. They include the ECB, the NCBs of the euro area countries, and credit institutionsand money market funds located in the euro area.

Mortgage bond: bond issue secured by a mortgage on the issuer’s property, the lien on which isconveyed to the bondholders by a deed of trust.

Mutual funds: investment company that raises money from shareholders and invests the proceeds(also investment funds).

Net asset value: is calculated for an investment fund by taking the market value of all securitiesowned plus all other assets such as cash, subtracting all liabilities, then dividing the result (totalnet assets) by the total number of shares outstanding.

Netting: the process of offsetting cash or securities positions. Through netting, the grosspositions are reduced. This is particularly true for the cash side, as all cash is fungible, this is notnecessarily true for assets.

Non-collective investment funds: funds not managed on a collective basis but on behalf of anindividual institution, or possibly of an individual. Important institutional investors includepension funds and insurance companies, which might delegate the management of funds to anexternal fund manager or within the financial group to which it belongs. By contrast with collectiveinvestments, “non-collective investments” may be characterised as private placements notmarketed to the general public but established by direct contracts between the investor and the fundmanager.

Non-stationarity: see stationarity.

Obligaciones del Estado: Spanish Treasury bonds with initial maturity of more than five years.

Obligations assimilables du trésor (OAT): French fungible Treasury bonds with originalmaturities from seven to thirty years.

Obligations linéaires-lineaire obligaties (OLO): Belgian fungible medium and long-termTreasury bonds with original maturity of up to thirty years.

OECD: Organisation for Economic Cooperation and Development.

79ECB


GLOSSARY

Option: an option is a financial instrument which gives the owner the right, but not the obligation,to buy or sell a specific underlying asset (e.g. a bond or a stock) at a predetermined price (the strikeor exercise price) on or up to a certain future date (the exercise or maturity date). A call optiongives the holder the right to purchase the underlying asset at an agreed exercise price, whereas aput option gives the holder the right to sell it at an agreed exercise price.

OTC (over-the-counter): bilateral transactions not conducted on a formal exchange.

Passive management: a style of investment management, which seeks to replicate theperformance of a market index. Passive management is also called index management.

Pfandbriefe: German mortgage bonds.

Primary dealer: selected credit institution authorised to buy and sell original issuance ofgovernment securities in direct dealing with the Treasury.

Primary market: market for new issues of securities.

Quote-driven market: usually a decentralised market where a class of participants, possiblymarket makers, post bid and offer quotes, often indicative, with prices being determined throughbilateral negotiation.

R-squared: measure of quality to fit in econometrics. Ranging from 0% to 100%, R-squaredshows what percentage of a variable’s movement is explained by the applied model.

Real-time gross settlement (RTGS) system: a settlement system in which processing andsettlement take place on an order-by-order basis (without netting) in real time (continuously).

Remote access: access to a system granted to a participant which has neither its head office norany of its branches located in the country where the system is based.

Repo/repurchase agreement: an agreement to sell an asset and to repurchase it at a specifiedprice on a predetermined future date or on demand. Such an agreement is similar to collateralisedborrowing, although it differs in that the seller does not retain ownership of the assets. Sale andrepurchase agreements are also termed repo transactions and are traded on the repo market.

S&P 500: Standard and Poor’s stock index of the 500 leading US companies.

Secondary market: exchanges and over-the-counter markets where securities are bought and soldsubsequent to the original issuance, which took place in the primary market.

Settlement: completion of a transaction by exchange of instruments and funds.

Special collateral: collateral other than general collateral.

Standard & Poor’s (S&P): index provider.

Stationarity: in econometrics a time series is said to be (weakly) stationary if the mean andauto-covariances of the series do not depend on time. Any series that is not stationary is said to be


non-stationary. Standard inference procedures do not apply to regressions which contain a non-stationary dependent variable or integrated regressors.

Swap: an agreement on an exchange of payments between two counterparties at some point(s) inthe future and according to a specified formula.

Swap spread: difference between the fixed rate of an interest rate swap and the government bondrate.

Synthetic collateralised debt obligation: redistributes the risk inherent in a portfolio of CDSacross a number of tranches that have a strict seniority ordering.

TARGET (Trans-European Automated Real-time Gross settlement Express Transfersystem): the RTGS system for the euro. It is a decentralised system consisting of 15 nationalRTGS systems, the ECB payment mechanism (EPM) and the interlinking mechanism.

Treasury bill: short-term government debt instrument issued at a discount with a maturity of oneyear or less.

Treaty: refers to the Treaty establishing the European Community. The Treaty was signed inRome on 25 March 1957 and entered into force on 1 January 1958. It established the EuropeanEconomic Community (EEC), which is now the European Community (EC), and is often referredto as the “Treaty of Rome”. The Treaty on European Union (which is often referred to as the“Maastricht Treaty”) was signed on 7 February 1992 and entered into force on 1 November 1993.The Treaty on European Union amended the Treaty establishing the European Community andestablished the European Union. The “Treaty of Amsterdam”, which was signed in Amsterdam on2 October 1997 and entered into force on 1 May 1999, and most recently the “Treaty of Nice”,which was signed on 26 February 2001 and entered into force on 1 February 2003, amended boththe Treaty establishing the European Community and the Treaty on European Union.

UCITS: undertakings for collective investment in transferable securities.

Vector autoregression (VAR): in econometrics an approach in which every endogenous variablein the system is treated as a function of the lagged values of all of the endogenous variables in thesystem.

Vector error correction (VEC): a VAR that restricts the long-run behaviour of the endogenousvariables to converge to their cointegrating relationships while allowing for short-run adjustmentdynamics.

Warrant: a security that entitles the holder to buy a proportionate amount of common stock at aspecified price, usually higher than the price at the time of issuance, for a period of years or inperpetuity.

Zero coupon bond: a security issued at discount or one which delivers a single coupon atmaturity.

The euro bond market study, December 2004

Documents

Transcript of The euro bond market study, December 2004