Deluge or dribble? EU pension reform and real estate investment flows

Deluge or dribble? EU pensionreform and real estateinvestment flowsPatrick McAllisterReceived (in revised form): 20th September, 2001

AbstractThis paper assesses the implications of increases in private pensionprovision within the European Union (EU) for the level anddestination of real estate investment. The first part discusses theincreasing ‘institutionalisation’ of the savings markets, current assetallocation patterns and demographic and fiscal pressures for reformof pensions markets. The projections for future growth in pensionsare outlined and implications for real estate investment assessed. Itis concluded that, although numerous imponderables render reliablequantitative projections problematic, growth and restructuring of thepensions market is likely to increase cross-border capital flows toreal estate markets. However, current projections imply limitedgrowth of direct real estate investment.

INTRODUCTIONAmong European economies there are striking disparities in thelevel and destination of pension fund savings. Furthermore, despiteongoing economic integration, there remains considerabledivergence between national states in terms of maturity of capitalmarkets, ‘institutionalisation’ of savings, regulation of investment,typical portfolio structures and taxation and legal provisions.Survey evidence from European investing institutions has found aconsensus that pension funds and pension-related life assurance willbe a major future growth market within the EU (EuropeanCommission, 1998a). In addition, the European pension fundmanagement industry has been experiencing change in terms ofconsolidation and internationalisation. This paper seeks to explorethe causes and potential consequences of anticipated growth in thepensions sector and to assess the implications for levels andpatterns of commercial real estate investment.The pattern of pension fund investment will be affected by the

level of investment and the configuration of the pension fundindustry. The ability of pension funds to invest and operateinternationally is influenced by a range of market and regulatoryfactors. However, the changes that are occurring in the organisationand regulation of the pension sector are only alluded to in thispaper where the main focus is the potential for growth in the levelof savings.1 The remainder of the paper is organised as follows. The

Patrick McAllister

is a lecturer in valuation and

property management in the

Department of Land

Management at the University

of Reading. He has published

widely on a range of topics

related to international

property investment and

valuation.

Keywords:European pension reform, realestate, institutional investment

Patrick McAllisterDepartment of Land Managementand DevelopmentFaculty of Urban and RegionalStudiesThe University of ReadingWhiteknights, PO Box 219Reading, RG6 6AW, UKTel: +44 (0)118 9316657E-mail:[email protected]

# HENRY S T EWART PUB L I C A T I ONS 1 4 7 3 - 1 8 9 4 B r i e f i n g s i n R e a l E s t a t e F i n a n c e VO L . 1 NO . 3 P P 2 4 8 – 2 5 72 4 8

first part outlines recent trends in institutional investment marketsand real estate allocation. This is followed by a discussion of thefiscal and demographic pressures for reform and the variations inthe savings ‘gap’ within the EU. Finally, projections for the growthof assets are critically assessed and potential implications for realestate investment are presented.

THE ‘INSTITUTIONALISATION’ OF SAVINGSAmong the most significant structural changes in global financialmarkets over the last two decades has been the rising concentrationof private savings in the control of investing institutions. The lasttwo decades have seen dramatic growth in the assets of pensionfunds, life assurance companies and, particularly, investmentcompanies (OECD, 1997). All can be categorised as fundmanagement service providers. Explanations of the growth of thefund management sector have focused on the consequences of fourmain, often interrelated, socio-economic trends: demographicchange, financial deregulation and liberalisation, technologicalinnovation and disintermediation in the banking sector. The latterthree developments have been significant drivers of the increasingblurring (where permitted) of functional boundaries in the financialservices sector as banks move into fund management, life assurancecompanies provide banking services (and vice versa) and pensionfunds and banks and life assurance companies enter theindependent fund management services sector.However, it is clear when the pattern of change is examined that

there are notable national variations in the size of, and the patternof change in, the private pension fund sector within the EU.Although the changes in the value of holdings can be distorted inthe short term by market and exchange rate performance, it is clearfrom Table 1 that in the last decade most of the growth in pensionfund assets within the EU is accounted for by the UK and theNetherlands. Given that long-term market growth will generallylead to growth in the value of holdings, the relative change in thevalue of pension fund holdings in Italy, Spain and Germany is

Table 1: European pension fund assets1 (billions of US dollars)

1990 1991 1992 1993 1994 1995 1996 1997 1998

Austria — 0.8 1.0 1.1 1.5 2.2 2.7 3.4 5.5

Belgium 3.9 5.6 5.7 6.2 6.8 10.1 11.0 11.6 —

Denmark 19.5 20.8 21.2 23.4 26.1 30.4 31.1 30.4 37.6

Germany 51.5 56.0 56.6 47.6 55.5 65.3 64.8 60.6 69.5

Italy 38.5 49.6 38.3 33.9 35.5 39.0 39.2 34.4 37.4

Netherlands 229.7 242.7 244.8 260.1 293.8 352.1 370.7 367.7 323.0

Portugal 1.1 2.0 2.8 4.2 5.8 8.9 10.2 10.5 12.7

Spain2 14.4 16.7 14.4 12.3 11.2 11.6 11.4 10.0 11.4

United Kingdom 536.6 599.4 552.4 683.2 660.5 759.7 893.2 1,067.0 1,137.0

(1) Autonomous and non-autonomous pension funds

(2) Autonomous pension funds only

Source: OECD (2000)

Growth of fundmanagement

Private pensionprovision in the EU

# H ENRY S T EWART PUB L I C A T I ON S 1 4 7 3 - 1 8 9 4 B r i e f i n g s i n R e a l E s t a t e F i n a n c e VOL . 1 NO . 3 P P 2 4 8 - 2 5 7 2 4 9

Deluge or dribble? EU pension reform and real estate investment flows

indicative of the limited contribution of these sectors to pensionprovision in their respective countries.Table 2 confirms the enormous variations in the relative size of

the long-term savings sectors within the EU and reinforces the viewthat this is mainly due to variations in the pensions sector relativeto other institutional savings markets.

CURRENT ASSET ALLOCATION PATTERNSFigure 1 displays the allocations of pension funds to what theOrganisation for Economic Cooperation and Development (OECD)categorises as ‘non-financial assets’. In this context, this is assumedto represent direct real estate investment. This assumption is basedon the reasoning that, although no separate classification for realestate is provided by OECD data, other studies have indicated asignificant allocation to this asset class (see European Commission,

Allocations to realestate

Table 2: EU pension fund and life assurance company assets

Financial assets of Financial assets of Financial assets ofinstitutional pension funds insurance cos.

investors% of GDP 1996 % of GDP 1996 % of GDP 1996

Austria 39.4 1.2 21.0

Belgium 63.0 4.1 30.9

Denmark 67.1 16.9 45.1

Finland 57.0 — 14.0

France (1) 90.6 5.6 (1996) 52.6

Germany (1) 57.5 5.8 (1996) 31.9

Greece 28.5 11.9 28.5

Italy (1) 53.2 2.9 53.2

Netherlands (1) 183.8 102.0 60.3

Portugal (1) 31.7 10.1 9.4

Spain (1) 56.1 2.0 20.4

Sweden 120.3 2.4 56.9

United Kingdom 193.1 77.5 88.6

Source: OECD 1999; BIS, 1998.

(1) 1997 figures

0.00%

10.00%

20.00%

30.00%

40.00%

50.00%

60.00%

Aus

tria

Belgium

Franc

e

Por

tuga

l

Unite

d Kingd

om

Den

mar

k

Ger

man

yIta

ly

% o

f to

tal assets

Figure 1: Allocation to non-financial assets by autonomous pension funds, 1996

Source: OECD (1999). The high allocation for Italian pension funds is very unusual. It reflects substantial investment

in the residential sector. However, the UK market is itself unusual in the almost complete aversion of British inves-

tors to housing investment.


McAllister

1997 and 1999). Furthermore, for the UK, the OECD data on non-financial assets corresponds to Office for National Statistics (ONS)figures for pension fund holdings of real estate in the UK. Apartfrom the major outlier of Italy, there is little dispersion concerningrelative investment in real estate and allocations are generally low.In addition, in common with UK and US experience, availableevidence suggests that the allocation to real estate by pension fundshas been declining in relative terms in the 1990s (see EuropeanCommission, 1998c). However, the evidence suggests that, incommon with the UK, there has been substantial investment innon-domestic equity and bond markets.

DEMOGRAPHIC AND FISCAL PRESSURES FOR PENSION MARKETREFORMAny review of the literature on European pension reform reveals abroad consensus among commentators that the current system offinancing pension provision within many EU countries isunsustainable and that demographic change and associated fiscalconsequences will lead to a restructuring of the funding ofretirement. Although the relative merits of different approaches tofunding retirement provision are politically contentious, it is widelyenvisaged that the solution to the ‘burden’ of unfunded, publicpension models will be a prefunded, private pension approach (seeEuropean Commission, 1998a). Within the EU, the central‘problem’ is that currently the majority of national pension systemsare unfunded. In 1998 over 80 per cent of total pension payoutswithin the EU were from unfunded schemes (EuropeanCommission, 1999). As reduced fertility and increased lifeexpectancy reduce the dependency ratio, if current trends continue,there will be a dramatic increase in the relative fiscal cost ofpension provision. Forecasts for the EU are that the dependencyratio will reduce from 4:1 currently to 2:1 in 2040 (EuropeanCommission, 1997).However, the implications are variable between countries. It has

been estimated that by 2030, without structural reform, pensionexpenditure will account for 15–20 per cent of gross domesticproduct (GDP) in Belgium, Finland, France, Germany and Italy.The comparable figures for the UK, Portugal and Ireland are under10 per cent (European Commission, 1997). The size of nationalpension fund assets relative to GDP provides a useful proxy for therelative size of the problem in individual economies. It is the clearview of the European Commission that reform of the funding ofpension provision is inevitable and that the reform will follow the‘Anglo-Saxon’ model of the growth of funded Second(occupational) and Third (individual) Pillar schemes.

INSTITUTIONAL PRESSURESFollowing important directives in the life assurance, investmentservices and banking sectors, the next major objective of the

Ageing populations



Internal Market Directorate General is to introduce a PensionServices Directive. Preparation for and consultation on such aDirective are well advanced. As indicated above, the EuropeanCommission has a relatively clear vision for the reform andintegration of pension service provision with the Union. The GreenPaper on supplementary pensions within the EU and the analysis ofresponses to this Green Paper suggest that they advocate a broadstrategy of growth in funded Second Pillar schemes and theintroduction of qualitative guidance on investment allocations. Themain aims and objectives are that there should be freedom ofinvestment and an ‘essential’ increase in investment in equities forthe longer term (see European Commission, 1999).

FORECASTING ASSET GROWTH AND DESTINATIONThere are major methodological difficulties in quantifying theconsequences of such changes for the private fund managementsector. In terms of market growth, the Economist Intelligence Unit(EIU) analysis below reflects the general expected repercussions ofincreased savings.

‘. . . the countries with the highest projected increases insavings will tend to be capital exporters to the rest of the euro-zone because they will want to build up pan-Europeaninvestment portfolios. Indeed the rate of growth of pensionfunds from currently low EU levels could be a significant driverof cross-border capital flows.’ (EIU 1999: 27)

For pension reform, although it is clear that there will be significantfiscal pressures on many EU countries, policy responses are likelyto be neither consistent nor comprehensive. Potential policyalternatives to private funded models are: the development ofpublic pre-funded schemes; adjustments of present entitlements suchas contributions levels, retirement age and indexation provisions;fiscal adjustments to expenditure or taxation levels; or macro-economic reform to influence demographic profiles, productivityand technological innovation. Manifestly, these are not mutuallyexclusive and a pragmatic blend of these policy alternatives is mostlikely and, indeed, can be observed in nascent form. Even if there isa shift towards a privately funded model, the consequences of thetransition will be influenced by uncertain factors such as marketreturns, growth in demand for lifestyle and income protection andthe speed of transition. As a result, forecasts of the growth ofpension fund savings markets have needed to make significant leapsof faith regarding future events and trends. Unfortunately, theseassumptions are not always transparent when their results andconclusions are reported.Indeed, many analyses simply seek to measure the size of the

‘savings’ gap rather than forecast market growth. Starting from anassumption of an instantaneous switch to private savings, Swiss Re

Pension reform

Policy responses

Forecasting problems


McAllister

(1998) estimated that annual private savings would need to expandin the region of $80bn in France, $130bn in Germany and $40bn inItaly in order to cover the financing gap in public provision.2

Alternatively for the same countries they suggest that privatesavings need to expand at annual rates of 3 per cent, 4.5 per centand 5.5 per cent, respectively. Swiss Re point out that the figuresare not forecasts of market growth but rather serve to illustrate thepotential for growth in the market. The Bank of England (1998)published data on the required increase in pension fund assetsnecessary to provide European economies with pre-fundingcomparable to US levels of pension savings. Although it is manifestthat many European economies will not be able or feel it necessaryto match a US benchmark, the figures displayed in Table 4 give anindication of the scale of the sums involved. Even if it is assumedthat a conservative proportion of the projected ‘shortfalls’ are metby private savings, this implies large capital flows of which aproportion will be invested in direct real estate. Research byPragma Consulting (for the European Commission) has producedbroad projections of the growth of assets of EU pension funds.Based on assumed nominal returns on portfolios of 6.4 per centp.a. and 2.6 per cent growth in net contributions, it is projectedthat by 2010 pension funds assets will grow nearly threefold withinthe EU (see Table 4).

Table 3: The funding ‘gap’ in EU pensions

Private pension fund Adjustment requiredassets 1996 to reach US level

$ bn $ bn

Austria 3 139

Belgium 11 148

Denmark 38 69

Finland 18 39

France 37 893

Germany 137 1,341

Greece 4 74

Ireland 32 14

Italy 32 777

Netherlands 349 -103

Portugal 30 50

Spain 22 317

Sweden 38 34

UK 966 -168

Total (excluding UK and Netherlands) 3,353

Source: Bank of England, 1998

Table 4: Projected asset growths of

EU pension funds

End 2000 C= 2,107.47

End 2005 C= 3,242.60

End 2010 C= 4,989.14

End 2015 C= 7,676.41

End 2020 C= 11,811.10

Source: European Commission, 1999

Scenarios



Analysis of the Pragma data suggests that they expect new fundflows in the region of C=60bn per annum. Although conservativecompared to Swiss Re and Bank of England figures, they wouldseem more realistic given the political complexities of and potentialalternatives to a swift switch to private pension schemes. Whateverthe quantity, it seems reasonably evident that there will be asignificant expansion of private pensions within the EU.Two notably pervasive features of the institutional investment in

real estate have been declining allocation and acute home countrybias. It is uncertain whether these trends will persist. Surveyevidence from the USA suggests that most investors are seeking toincrease their exposure to direct domestic and non-domestic realestate markets (Institutional Real Estate Inc., 1999). The keyquestions are:

. What proportion of new pension funds will be allocated to realestate?

. What proportion of new pension real estate funds will beallocated to cross-border real estate?

In order to arrive at estimates of net flows to direct real estate, the

Implied growth inassets

Issues for real estate

Table 5: Projected growth of EU pension fund assets 2000-20201 at current

prices

Year Projected Net Allocation to Of whichfund growth increase real estate Cross-border

C= bn C= bn C= bn C= bn

2000 2107.47

2001 2162.26 54.79 2.74 0.68

2002 2218.48 56.22 2.81 0.70

2003 2276.16 57.68 2.88 0.72

2004 2335.34 59.18 2.96 0.74

2005 2396.06 60.72 3.04 0.76

2006 2458.36 62.30 3.11 0.78

2007 2522.28 63.92 3.20 0.80

2008 2587.86 65.58 3.28 0.82

2009 2655.14 67.28 3.36 0.84

2010 2724.18 69.03 3.45 0.86

2011 2795.00 70.83 3.54 0.89

2012 2867.67 72.67 3.63 0.91

2013 2942.23 74.56 3.73 0.93

2014 3018.73 76.50 3.82 0.96

2015 3097.22 78.49 3.92 0.98

2016 3177.75 80.53 4.03 1.01

2017 3260.37 82.62 4.13 1.03

2018 3345.14 84.77 4.24 1.06

2019 3432.11 86.97 4.35 1.09

2020 3521.35 89.23 4.46 1.12

Assumptions:

Growth in contributions 2.60%

Allocation to real estate 5%

Cross-border 25%

1 Since the Pragma figures are five-yearly and include growth in asset values as

well as asset growth, the table represents the implied annual growth in pension

fund assets excluding growth in asset values.

Source: Adapted from European Commission, 1997


McAllister

first stage is to separate new inflows from growth in value ofexisting assets. The Pragma projections are that nearly 27 per centof the growth in assets will be due to new contributions. In Table 5,an average growth in contributions of 2.6 per cent per annum isassumed to be added to the existing level of assets. Next, theamount of new funds allocated to real estate is estimated. A figureof 5 per cent is used since this is broadly in line with currentEuropean and US allocations. Hence the allocation to real estatecolumn provides an estimate of the flows of ‘new money’ into realestate from pension-related savings within the eurozone over thenext 20 years. Assuming that 25 per cent3 of these funds areinvested outside the domestic market, then the allocation tointernational real estate column provides an estimate of the flow ofnew international real estate funds from increased pension fundswithin the EU.This produces estimates of new pension-related capital flows to

direct real estate investment of C=2.74bn per annum in 2001growing to C=3.5bn in 2010. Given assumptions about cross-borderflows, this reflects an increase in cross-border flows of C=0.7bn in2001. Manifestly, the reliability of such projections is dependentupon the reliability of the inputs and underlying assumptions. Table6 illustrates the sensitivity of the projections to relatively smallvariations in assumptions about savings expansion and allocationto real estate. It is apparent that relatively minor adjustments togrowth and allocation expectations can lead to large increases inthe projection of new funds.This analysis also raises questions about the route of investment

flows to the property market, probable destinations andimplications for returns. Although speculative, it is possible tomake a number of points. It is likely that a significant alternativechannel of pension funds to the property sector will be throughpooled vehicles. A proportion of the new funds will flow indirectlyto real estate with the expansion of equity markets and widelyacknowledged pension preferences for indirect vehicles. It isinstructive that the US pension approach to investment in EU realestate markets has mainly been through opportunist funds and realestate investment trusts (REITS). Indeed, Eichholtz and Lie (2000)identify significant variations in the level of securitisation amongthe real estate markets of the EU, estimate that only 1–2 per centof the total property stock of the EU is securitised and suggest that

Table 6: Sensitivity of 2001 projection allocation to real estate C= bn

6% 7% 8% 9% 10%

3.00% 3.79 4.43 5.06 5.69 6.32

3.50% 4.43 5.16 5.90 6.64 7.38

4.00% 5.06 5.90 6.74 7.59 8.43

4.50% 5.67 6.64 7.59 8.54 9.48

5.00% 6.32 7.38 8.43 9.48 10.54

Allocation

Indirect investment



there is considerable scope for expansion of this sector, particularlyin the Mediterranean economies. There has been conjecture that thehigh level of business owner-occupation in the EU will start toreduce as EU corporations start to liquidate their property assets inorder to raise capital (Carpenter, 2000). This is closely linked to anassessment of the implications for returns. Given the complexitiesinvolved in trying to identify the determinants of property returns,it is inappropriate in this context to try to make specific projections.However, key questions that need to be considered concern:

. To what extent will increased capital flows to European equityand (corporate) bond markets impact upon property returns?

. What proportion of additional investment will come fromchanges in ownership of existing stock relative to newconstruction?

. What proportion of additional investment will be channelledthrough securitised real estate vehicles?

. To what extent will the growth in investment demand exceed,match or fall below growth in supply of investment property?

. To what extent will there be significant geographical variations inthe allocation of new investment funds?

CONCLUSIONContemplating the questions above and the numerous otherimponderables associated with this discussion highlights thedifficulties of making any confident judgments about the potentialrepercussions of pension reform for EU markets. However, thisdoes not mean that the topic can be ignored. It is clear that there isa strong possibility that the next decade will witness dramaticdevelopments in the size and configuration of the pension fundsector within the EU. There is a broad consensus that the currentand latent fiscal pressures produced by ageing populations willproduce an increase in private savings within the EU, in particular,in France, Italy and Germany. While there is uncertainty aboutmagnitude and extent, the whole thrust of current developments isfor continuing growth and integration of the investment servicesmarket. The former implies increased demand for real estate as anasset class, while the latter implies increased cross-border flows ofreal estate investment. Although the main destination of thesesavings is likely to be in the equity and corporate bond markets ofthe eurozone, a proportion will go to real estate. However, thereare substantial barriers distinct to real estate which seem to inhibitcross-border real estate investment. Real estate is much more proneto home country bias than equity or bonds. The underlyingstructural trends seem to be towards lower allocations to direct realestate investment generally and relative to indirect real estateinvestment. Expansion and integration of investment servicesmarkets may simply act as a brake on these long-term trends.Given that recorded cross-border activity for office investment in

Other considerations


McAllister

1998 in Europe was approximately C=6.5bn, it seems reasonable toconclude that pension reform and associated deregulation will nothave dramatic widespread effects upon domestic and cross-borderreal estate investment within the EU (Europroperty, 1999). An areafor further research is whether there will be significant implicationsfor individual national markets and asset classes if investment isconcentrated into certain markets or investment vehicles.

ReferencesBank for International Settlements (1998) The Macroeconomic and Financial Implications of

Ageing Populations.

Bank of England (1998) Practical Issues Arising from the Introduction of the Euro, Issue 8,

June.

Carpenter, A. (2000) ‘Spin-offs’ dramatic effects’, Estates Gazette, 4th March.

Eichholtz, P. and Lie, R. (2000) Property Capital Flows: Moving the Frontiers, ING Real

Estate, The Hague.

EIU (Economist Intelligence Unit) (1999) EMU: A Catalyst for Change in the European

Financial Services Industry, The Merrill Lynch Forum.

European Commission (1997) Supplementary Pensions in the Single Market, Communication

of the Commission, com (97) 283.

European Commission (1998a) Overview of the Responses to the Green Paper on

Supplementary Pensions in the Single Market, COM (97) 283.

European Commission (1998b) Financial SC=rvices: Building a Framework for Action,

Communication of the Commission.

European Commission (1999) Rebuilding Pensions: Security, Efficiency and Affordability —

Recommendations for the Code of Best Practice for Second Pillar Pension Funds,

Communication of the Commission.

Europroperty (1999) ‘Direct office investment courts central Europe’, March, 1999.

Frampton, R. (1999) ‘Building pan-European structures’, Europroperty, October, 10–11.

Fund Managers’ Association (1999) Fund Management Survey 1999.

Institutional Real Estate Inc. (1999) Annual Plan Sponsor Survey: Tax Exempt Real Estate

Investment 1999, USA.

Mercer, W. M. (2001) European Pension Fund Managers’ Guide 2001/2002 — latest

developments in institutional fund management located at //www.wmmercer.com/global/

english/resource/index.html.

OECD (1997) ‘The impact of institutional investors on OECD financial markets’, Financial

Market Trends, 68: 5–50.

Pragma Consulting (for the European Commission), see European Commission (1999) above.

Swiss Re (1998) ‘Financial difficulties of public pension schemes: Market potential for life

insurers’, Sigma, No. 8/1998.

Notes(1) W. M. Mercer’s annual survey on trends in European pension fund management provides

a good source of information on how the structure of the industry is changing.

(2) The assumptions underlying the calculations are not clear. They appear to assume that

there is no relative expansion of public expenditure on pension provision and that

additional private savings fill the resultant shortfall.

(3) Obviously the figure of 25 per cent is rather arbitrary. It is much higher than current

allocations to international property for UK and US funds. It reflects the expected

impact of the introduction of the euro and the consequent elimination of currency

matching issues for eurozone pension funds.



Deluge or dribble? EU pension reform and real estate investment flows

Documents

Transcript of Deluge or dribble? EU pension reform and real estate investment flows