Traditionally, the pension systems of most Western European countries were textbook examples for the dominance of public pay-as-you-go pensions. This has changed.
2. Imprint
Publisher: Allianz Global Investors AG, Seidlstraße 24-24a, D-80335 Munich | http://www.allianzglobalinvestors.com | Editor: Dr. Alexander Börsch,
Senior Pensions Analyst, Allianz Global Investors AG, E-mail: alexander.boersch@allianzgi.com, Phone: +49 (0) 89 1220 7472 | Contributors: Dr. Alexander Börsch,
Allianz Global Investors AG; Dr. Renate Finke, Allianz Dresdner Economic Research; Dr. Martin Gasche, Allianz Dresdner Economic Research; Jordy Peek, risklab;
Dr. Jürgen Stanowsky, Allianz Dresdner Economic Research | Layout: volkart:51 GmbH, Munich | Printing: repromüller, Übersee | Closing date: December 5, 2008 |
This study was conducted in cooperation with Allianz Dresdner Economic Research. The OECD provided information on the second and third pillars.
The entire content of this publication is protected by copyright with all rights reserved to Allianz Global Investors AG. Any copying, modifying, distributing
or other use of the content for any purpose without the prior written consent of Allianz Global Investors AG is prohibited. The information contained in this
publication has been carefully verified by the time of release, however Allianz Global Investors AG does not warrant the accuracy, reliability or completeness
of any information contained in this publication. Neither Allianz Global Investors AG nor its employees and deputies will take legal responsibility for any
errors or omissions therein.
This publication is intended for general information purposes only. None of the information should be interpreted as a solicitation, offer or recommendation
of any kind. Certain of the statements contained herein may be statements of future expectations and involve known and unknown risks and uncertainties
which may cause actual results, performance or events to differ materially from those expressed or implied in such statements.
2
3. International Pension Studies Western Europe
Preface
T raditionally, the pension systems of most
Western European countries were text-
book examples for the dominance of public
the financial assets of European households
in international comparison and includes
forecasts on the future development of finan-
pay-as-you-go pensions. This has changed. cial assets. The third article addresses trends
More and more European countries are try- in Western European pension markets and
ing to spread the retirement income of their pension asset projections, while the conclud-
citizens across a wider base. They have intro- ing article of the first part is concerned with
duced new funded occupational and private the effects of a possible application of Solven-
pension schemes with the goal to diversify cy II on defined benefit pension funds. The
retirement income for future pensioners. second part analyses the pension systems in
each Western European country; for the pur-
Despite different starting points, almost poses of this study we defined Western Europe
all Western European countries have fol- as the EU-15 plus Switzerland and Norway.
lowed the trend towards funded pensions.
This has taken several forms. Besides the in- The evolution of funded pensions into
troduction of new schemes, many countries a crucial element of retirement income in
decided to introduce state pension reserve Western Europe has important ramifications
funds to back public pension systems and for public policy and the financial industry.
strengthen their sustainability. These funds For example, the regulation of funded pen-
have grown considerably and are now crucial sions, the effectiveness of plan design, the
players in the financial markets. The world- risk management of investments and the
wide shift from defined benefit to defined quality of financial products become ques-
contribution plans in occupational pension tions that will shape the financial security
provision is also taking place in Europe, how- of many future retirees. By creating trans-
ever, with a different speed and depth in the parency about the pension system designs
respective countries. in Western Europe, this study aims to con-
tribute to the discussion on the future of
This study is divided into two main parts. European pensions. We strongly believe
The first part is comprised of four articles, in- that transparency and comparability are
cluding one on the economic impact of ageing the foundations for mutual learning and
populations. The second article focuses on best practice sharing.
Brigitte Miksa,
Head of International Pensions
Allianz Global Investors AG
3
6. Introduction International Pension Studies Western Europe
Western Europe –
The Economic Impact
of Ageing Populations
Demographic development Chart 1 Western European population [million]
in Europe 415
While Western European countries vary con- 410
siderably from one another, they also have
a number of commonalities. For instance, 405
in most of Western Europe, the population
400
is expected to stop growing within the next
twenty years. It will not be the first time that 395
European societies see a decline in their pop-
ulations. The Black Death and both World 390
Wars are two dramatic examples. Indeed, in
385
the past, wars and pandemics led to quickly
2005 2009 2013 2017 2021 2025 2029 2033 2037 2041 2045 2049
shrinking populations.
Source: Eurostat
In the coming decades, the decline in
Western Europe’s population will certainly be at the big picture to make rough estimates of
more gradual. Overall population will peak in future GDP growth within a growth account-
20 years. It will then start to decrease, slowly ing framework. Finally, we assess the impact
at first and then increasingly faster. The age- of ageing on the Western European pension
ing process of Europe’s populations will have business.
a significant impact on the makeup of its so-
cieties. The proportion of elderly people (i.e.
Longer lives, fewer children
people aged 65 and over) in the total popula-
tion will increase from about 17% today to 30% Across Western Europe, the basic demograph-
in 2050. In this publication, we have defined ic trends of ageing and shrinking populations
Western Europe as the former EU-15 coun- are to blame for many political reforms that
tries, plus Norway and Switzerland. have seen benefits cut – particularly pension
benefits – and extended working lives. The un-
The present article examines the impact derlying causes of low or falling fertility and
of ageing and declining populations on GDP increasing life expectancy are well known.
growth. We discuss how the shrinking labour However, the reasons and mechanisms be-
force has a negative effect on growth prospects hind these factors are not as well understood.
and also address the impact of ageing. First, A number of questions need to be answered,
we look at the impact of population decline on among them why women do not have more
the labour market. Next, we discuss the quali- children and what can be done to increase
ty of labour, the role of education, and the im- fertility. Other open questions include how
pact of ageing on productivity. Third, we look fast life expectancy will increase, or whether
at whether capital input can offset shrinking it will decrease due to less healthy lifestyles.
labour input and shed some light on the ques-
tion of how ageing affects an economy’s ability Since there are many possible answers to
to innovate. Once we have done this, we look these questions, it has thus far been impos-
6
7. Introduction International Pension Studies Western Europe
sible to make any sound scientific conclu- Can ageing societies remain
sions. Population development forecasts
must therefore rely on assumptions about competitive?
future developments.1 For the EU-15 coun- To answer this question, the effects of ageing 1 We have used Eurostat
tries, fertility is expected to increase slightly on the economy and its growth prospects must figures for all countries
except in France and Ireland, where current be examined. The following paragraphs shed except Norway and
high fertility levels are forecast to decrease. some light on this issue. We will identify the Switzerland. For each of
Given the lack of scientific evidence on the relationship between demographic change these two countries, we
drivers of change in fertility in industrialised and economic growth and discuss its poten- have used the figures
countries, it is possible that actual develop- tial impact on Western Europe. and forecasts of national
ments will deviate from these assumptions. statistical offices.
Demographic development affects eco-
Even if fertility in France keeps rising as nomic activity in many areas. Gross domes-
it has in recent years and current low levels tic product (GDP), i.e. the sum of all goods
remain constant in other countries, the im- and services produced within a given period,
pact would barely be noticeable in the near is the result of an application of capital and
future. Indeed, it will take many years for labour combined with the effects of techni-
such small developments to alter the age cal progress. Clearly, labour force develop-
composition of populations. After several ment has an influence on output. It should
decades of low fertility, the number of poten- be noted, however, that not only the number
tial mothers has significantly decreased,
so that an increase in fertility will not have
Chart 2 Fertility rates in Western Europe [children per woman]
a noticeable impact in the short term.
2.5
A similar argument can be made about 1990
life expectancy. An increase has been fore-
2 2006
cast for all countries in our sample, though
it will be higher in some than in others. The
highest life expectancy increase for men be- 1.5
tween 2006 and 2050 will presumably occur
in Austria, with 7.2 years. For women, life ex-
1
pectancy will increase the most in Belgium,
by 6.4 years. We have assumed the lowest
increase in the Netherlands, with 3.8 years 0.5
for men and 2.7 years for women. Current
national assumptions may deviate from the 0
estimates used by Eurostat. However, even Germany Italy France Spain UK The Nether- Sweden Switzer-
lands land
if deviating national assumptions were used,
changes to the overall picture would be
negligible. Chart 3: Old-age dependency ratios [65+-year-olds/15–64-year-olds]
70
The continuous ageing of societies is one 2008
of the consequences of these developments, 60
2050
and is reflected in the old age dependency
50
ratio. It depicts the number of people aged
65 and older per 100 people of working age
40
(15–64 years) and demonstrates that the
ageing process varies in different Western 30
European countries. For instance, Spain, Italy
and Germany are much more affected than 20
countries like the UK, France or Sweden.
10
0
Germany Italy France Spain UK The Nether- Sweden Switzer- Western
lands land Europe
Source: Eurostat
7
8. Introduction International Pension Studies Western Europe
of workers matter, but also their quality. of the 15-64 age group will differ considera-
Education, i.e. the quality of human capital, bly across Europe. Germany and Italy will see
is an important determinant of productivity. the sharpest declines. Labour force potential
in these two countries will shrink by 25% by
Since population development and the 2050. In contrast, Ireland will see an increase
age structure of a population affect the use of 15%, the increase in Luxembourg will be
and quality of production factors in a number almost 30%, and Sweden’s labour force po-
of ways, the impact of demographic change on tential will grow by 3%. All other countries
output is difficult to determine. Some simple will have to cope with a declining potential
economic models describe these relation- labour force.
ships. In this basic accounting exercise, GDP
is the product of the number of employed A decline in the potential labour force
persons (working population) times the does not mean that the size of the labour
output in goods and services per employed force automatically shrinks. This is because
person (labour productivity): many people aged 15-64 are not active in
the labour market. A substantial number of
people are pursuing their education, some
GDP =
have retired early and others are currently at
labour productivity x working population
home to raise children or for other reasons.
In 2007, the labour force participation rate
The working population is a subset of the (i.e. people aged 15-64 who were either em-
people aged 15-64 who could potentially work ployed or looking for a job) stood at 74% for
(potential labour force) and do participate in men and almost 60% for women. The figures
the labour market. for Norway and Switzerland are well above
the EU-15 average.
Working population =
These participation rates can change for
participation rate x potential labour force
a number of reasons. Developments in the
economic cycle are an important influence,
This leaves us with: but changes in the legal environment also play
a role. Since 1995, the labour force participa-
tion rate in the EU-15 has increased by 3.5
GDP = labour productivity x
percentage points for men and 10 percentage
participation rate x potential labour force
points for women. Legislation to increase
Labour productivity depends on the avail-
ability of capital (i.e. tools and machinery), Chart 4 Development of the working age population aged 15–64
the quality of human capital and technologi- [2005=100]
cal progress. All of these items are influenced
105
by demographic development – some more
directly than others. Let us begin our analysis
100
with the most obvious variable, the potential
labour force.
95
Germany
Italy
Demographic impact 90
France
on the labour market 85
Spain
UK
The Netherlands
How do demographics affect the labour 80
Sweden
market? If we assume that the potential
Switzerland
labour force comprises all people between 75 Western Europe
15 and 64 years of age, we will see some
changes throughout Western Europe. 70
According to current population forecasts, 2005 2009 2013 2017 2021 2025 2029 2033 2037 2041 2045 2049
labour force potential will decline by 15%
Source: Eurostat, national statistical offices
between 2008 and 2050. The development
8
9. Introduction International Pension Studies Western Europe
Chart 5 Labour force participation rates of 55–64-year-olds, 2007 [%]
men 55–64
90
women 55–64
80
70
60
50
40
30
20
10
0
EU-15 Denmark Ireland Spain Italy The Netherlands Portugal Sweden Norway
Belgium Germany Greece France Luxembourg Austria Finland UK Switzerland
Source: Eurostat
retirement ages and an improved economy policies can certainly alleviate its most
have contributed to that development. dramatic effects.
This demonstrates that a declining potential
labour force does not automatically mean For instance, extending working life and
that fewer people are available on the labour closing loopholes that lead to early retire-
market. Participation rates could increase, ment are two possible measures that could
just as unemployment could decrease. be implemented to protect the labour mar-
ket from a declining population. Another is
About 13 million EU-15 citizens are current- increasing the willingness to join the work-
ly unemployed, which represents the entire force by offering childcare facilities for young
combined potential labour force of Belgium parents and/or more flexible working hours.
and Greece. With unemployment rates be- Taxes play an important role, too. If taxes
tween 2.5% and 4%, Norway and Switzerland on a couple’s secondary income are too high,
have reached almost full employment. Given the incentive to work is low, as experience in
the high participation rates in these two coun- the Netherlands has shown.
tries, changes in the potential labour force will
have a more direct impact on the availability Even if the total number of workers does
of labour. In other EU-15 countries, there is still not drop by as much as some may fear, there
some room to manoeuvre. For instance, la- is no denying that the workforce is ageing.
bour force participation rates for those aged In Western Europe, the current median age
55 to 64 can increase substantially in most is about 40 years; it is expected to increase
EU countries. Currently, only Sweden comes by roughly 7 years by 2050, as will the medi-
close to Switzerland and Norway. an age of the labour force. Until now, there
has not been any empirical evidence on how
If it were possible to increase labour force such a process will affect productivity.
participation across EU-15 states to rates as
high as Sweden’s, the demographic impact on
The role of education
the labour market could be cushioned con-
siderably. In fact, the expected labour force As already mentioned, the quality of the
decline of 15% could be reduced by half. Even labour force plays an important role – espe-
if some aspects of demographic development cially in countries with decreasing popula-
cannot be changed in the short term, smart tions. In Western Europe, Germany and Italy
9
10. Introduction International Pension Studies Western Europe
in particular must cope with pronounced pop- To assess the development of productivity
ulation decline. In these countries, and also in in an economy, the weight of different jobs
Portugal, Greece and Spain, efforts must be in the overall workforce is important. Mental
made in the realm of education to counter abilities, such as problem-solving skills when
the effects of ageing and population decline. faced with new challenges, logic and the
The decline in the absolute number of po- ability to understand complex topics, seem
tential workers must be countered by an in- to decrease with age. Verbal abilities and
crease in the productivity of each individual. communication skills, on the other hand,
tend to improve with age, as experience plays
To achieve this, national school systems are a more important role. The importance of
one of many factors that must be examined. different combinations of these abilities vary,
Certainly, reducing the number of school depending on the industry. For older em-
dropouts is critical. However, the education ployees, changes that are the result of rapid
of the workforce as a whole must also be technological development pose a particular
considered. Lifelong learning should not challenge with regard to productivity, as they
only be a catch phrase, it must become an make experience acquired over decades
embedded concept in companies’ personnel irrelevant. For this reason, industries that
departments. Giving the increase in retire- are characterized by such change will suffer
ment age, workers aged 55 and over will have more from an ageing workforce than others.
to spend a decade or more in employment
before they can access their pensions. Hence, There are very few empirical studies that
even for these experienced workers, continu- demonstrate these dynamics, and those that
ing education will become increasingly com- are currently available do not provide clear
mon. The depreciation of human capital in answers to the critical questions. Most stud-
a shrinking labour force must be prevented ies conclude that productivity declines with
(Ludwig, Schelkle, 2007). age. However, it is unclear at what age the de-
cline sets in and whether it is significant. This
Age and productivity leaves us with the following results: While a
decline in productivity due to ageing is pos-
The development of labour productivity is sible, it appears that it will not be substantial.
extremely important for an economy’s long- We must also consider the possibility that the
term growth prospects. Ageing populations deterioration of productivity can be countered
and workforces require that answers be by further education and training. There is
found to the question of how productivity ample evidence that employees can acquire
is affected by ageing. In rapidly ageing coun- new skills at any age. If companies keep this
tries like Germany and Italy, this question in mind, they may be able to maintain pro-
is high on the agenda. The fact that certain ductivity despite ageing societies. This is par-
mental abilities decrease after early adult- ticularly true in sectors where experience is
hood is one of several aspects that are un- important.
disputed.
However, this does not mean that produc- Demography and the use
tivity decreases as workers age. Experience, of capital
which increases with age, is also important.
A good balance between mental ability and When the supply of labour declines for demo-
experience leads to the best results. If we look graphic reasons, the importance of capital in
at purely physical work, the picture is slightly maintaining economic performance grows.
different, as productivity undoubtedly decreas- Capital must be used to substitute labour
es with age. In Western Europe, however, ever and increase the productivity of the remain-
fewer jobs rely exclusively on physical abili- ing jobs. In other words, investment in real
ties. Even in the automotive industry, it has capital will gain importance. In ageing West-
not been proven that younger assembly line ern Europe, capital investment is influenced
teams are more productive than their older by the need to substitute labour, which will
counterparts. While younger teams tend to become more expensive. Production will be-
be faster, they also make more mistakes. come even more capital intensive. The neces-
sary increases in productivity will be achieved
10
11. Introduction International Pension Studies Western Europe
by investing more in human and physical cap- the propensity to innovate seems to in-
ital – increasing capital demand even more. crease until age 40 a positive correlation
between age and innovations can be found
in younger populations. Chart 6 shows that
Demography and techno such a relationship is not contradicted by
logical progress the data, even though it does not prove it. At
least an increase in the share of elderly, here
Technological progress is an important factor the number of people aged 45 to 64, corre-
for economic growth. It results in new products lates positively with the number of patents.
or new, more efficient ways of producing exist-
ing products. Innovations are frequently the
Ageing and economic growth
result of technological progress. For Western
European economies, which trade a great deal Having addressed all of these considerations
with the rest of the world (and of course with regarding future labour force development,
each other), technological progress is one of productivity and technological progress, we
the pillars of growth. Western Europe can only can now have a closer look at Western Europe’s
successfully compete on global markets with growth prospects. The frame of reference for
very advanced products. If ageing were to in- this task is the simple growth accounting
terfere with countries’ ability to innovate, their framework introduced above. GDP growth
international competitiveness would face can be broken down into productivity growth
a direct threat. Countries such as Germany, and employment growth. The latter depends
which is a very strong exporter of highly so- on the growth of the working age population
phisticated machinery and cars, would suf- and changes in labour utilization (i.e. labour
fer if it could not maintain its competitive force participation). For example, the Euro-
advantage over foreign competitors. Assess- pean Commission and the European Central
ing the effect of ageing on technological Bank (ECB) conducted this exercise, and the
progress is therefore critical. result of our own research is in line with their
findings. It is very likely that GDP per capita
It is a commonly held thought that a soci- growth will decline over the next four decades.
ety’s ability to innovate suffers when it ages. However, this decline will not be dramatic
This rests on the assumption that creativity (Table 1).
decreases with age. In fact, recent empirical
research has shown that age tends to correlate
negatively with the propensity to innovate Chart 6 Share of people aged 45–64 and number of patents
(Schneider, Ragnitz, 2007). The likelihood of per 1 million inhabitants in the US
a successful innovation (new product, major
process or product improvement) peaks at an 350 0.26
average age close to 40. Afterwards, the likeli-
hood begins to decrease, dropping to lower 0.24
than at the age of 30 when the average age of 300
0.22
personnel is over 48. Hence, there is quite a Share of people aged 45-64,
long time span with fairly high innovation po- right scale [%]
0.20
tential. What is more, empirical results sug- 250
gest that a good balance of older and younger 0.18
employees must be reached to achieve high
innovation potential. 200
0.16
Number of patents per
US data on patent activity shows a rather 1 million inhabitants, 0.14
left scale
positive correlation between age and innova- 150
tions. This means that even if age affects the 0.12
likelihood to innovate negatively, this seems
to take effect only at higher ages – at least com- 100 0.10
pared to the average age of the workforce. 1960 1963 1966 1969 1972 1975 1978 1981 1984 1987 1990 1993 1996 1999 2002 2005
Given the projected increase in the age of the Note: Patents are defined as Utility Patents granted to US inhabitants
workforces in Western Europe, a strong neg-
Source: US Census Bureau, US Patent and Trademark Office
ative impact is unlikely to materialise. Since
11
12. Introduction International Pension Studies Western Europe
Table 1 Economic growth scenarios for Western Europe [average annual changes in %]
Labour Change in Working age Total GDP
Real GDP
productivity labour force population population per capita
Past: average
1 .1 0 .8 0 .4 0 .5 2 .3 1 .8
1995–2005
Base case
2011–2030 1.7 0.4 -0.3 0.1 1.8 1.7
2031–2050 1.5 0.2 -0.5 -0.2 1.2 1.4
Pessimistic
assumptions
2011–2030 1 .2 0 .1 0 .3 0 .1 1 .0 0 .9
2031–2050 1 .1 0 .0 0 .5 0 .2 0 .6 0 .8
EU commission
assumptions
2011–2030 1 .8 0 .2 0 .3 0 .1 1 .7 1 .6
2031–2050 1 .7 0 .1 0 .5 0 .2 1 .3 1 .5
Notes: Western Europe = EU 15, labour force participation: share of employed persons of working age population
Source: ECB, Eurostat, European Commission, own calculations
The scenarios have two main results: EU policies currently aim to achieve pre-
• As the table shows, growth will be affected cisely these goals, as summarised in the Lis-
dramatically only under very pessimistic bon agenda. Norway and Switzerland have
assumptions. In our base case scenario, similar goals with respect to growth. Indeed,
which is slightly less optimistic than the policies across Western Europe are geared
European Commission with regard to towards increasing labour force participa-
productivity developments, but more op- tion with higher retirement ages or better
timistic on labour force participation, in childcare facilities. They also aim to increase
the medium term GDP per capita growth labour productivity with efforts to raise edu-
will remain roughly at the level seen in the cational standards and foster research and
last decade. Towards the end of the fore- development. While progress towards reach-
cast period, growth will slow down as the ing these targets has not been consistent
change in the working age population in all countries, we expect all of Western
becomes more pronounced. Europe to make substantial progress within
the next decades. Although Europe is ageing,
• The table also demonstrates the impor- it will remain an economic force to reckon
tance of maintaining high labour produc- with in the future.
tivity. If the increase were to fall below the
past average of one percent, GDP growth
would be severely impeded. In fact, GDP Consequences for pensions
growth above one percent would be diffi- and pension investments
cult to obtain. An increase in labour force
participation is another important pre- What does all of this mean for pensions?
condition for economic growth to stay at First, the prospect of decent economic
levels close to those reached in the past. growth rates means that there is no need
12
13. Introduction International Pension Studies Western Europe
to fear the future. The consequences of rule out some major European countries such
ageing can be dealt with. Second, public, as Italy, Spain, and Germany, among many
pay-as-you-go financed pension systems others. However, as the above analysis has
will encounter difficulties as a result of the shown, an ageing society does not necessarily
changing ratio of people over 65 to the 15-64 face a bleak future.
age group, which finances pensions. This
ratio, which is commonly referred to as the The move towards a pension system
old age dependency ratio, will increase from that is funded to a much higher degree than
27 today to 52 in 2050. past systems can be seen as a reaction to the
changing needs of an ageing society. To suc-
This means that in the future, there will cessfully offset the decline in labour through
be 52 elderly people for every 100 people of investments in physical and human capital,
working age, compared with 27 today. Still, an economy needs a great deal of capital.
the outlook for public pensions would be Across Western Europe, the need to adapt
much worse if ageing automatically meant to demographic change varies significantly.
economic decline. This is not the case, as in- The situations in Sweden, France and the UK
comes are expected to keep growing. Wages, differ considerably from Germany or Italy.
which finance public pensions, will not be The need for reform is highest in the coun-
affected that much by ageing. However, age- tries that are ageing the most quickly and
ing will have a large impact on the viability have not yet sufficiently reacted by adapting
of public pension systems. their pension systems and economies to
the demographic conditions they will face
The most common reaction to this fore- in the future.
seeable development continues to be de-
creasing the generosity of public pension From an investment point of view, even
systems and strengthening private pension an ageing Western Europe seems like a
provision. The latter is usually of a funded promising place to invest. The demand for
type. People save money while working and capital to finance the investments required
receive the proceeds of their savings after to offset the decline in labour will grow. In
retirement. While this arrangement exposes some countries, notably Germany, this proc-
them to investment risk, pension asset man- ess has already begun or will start sometime
agers do their best to keep this risk manage- soon. In the end, return on investment and
able. They must continuously ask themselves the associated risks determine whether or
where the best possible investment can be not an investment is appropriate. Even if
made. If ageing societies were doomed to higher returns can be expected in emerging
slow growth, stagnation or eventual decline, economies, these also entail higher risk.
the only answer would be in young and/or Among others, exchange rates and political
fast growing economies. This concept would risks should be considered.
Dr. Jürgen Stanowsky,
Allianz Dresdner Economic Research
13
14. Introduction International Pension Studies Western Europe
The Financial Assets of
Private Households –
An International Comparison
Introduction above the level Japanese households had put
aside (EUR 9.7 trillion). However, because of
Over the past decade, Europe has seen many the difference in size of the regions, an anal-
pension reforms that have paved the way for ysis of the absolute values does not correctly
a shift from pay-as-you-go to funded pension reflect the ratios. Taking financial assets as a
systems. These reforms have led to a strength- percentage of gross domestic product (GDP),
ening of the second pension pillar in most we find that wealth in Western Europe, which
countries and highlighted the importance amounts to about 220% of GDP, is significant-
of third pillar provision, as replacement ly lower than in the other two regions. In 2007,
levels in the first pillar have been cut back. the United States’ monetary wealth of private
Especially in Continental Europe and EMU households was 328% of the country’s GDP.
(European Monetary Union) countries, the In Japan, it stood at 310%. In contrast, Eastern
introduction of tax-favoured savings prod- Europe lagged way behind, reaching only
ucts for retirement has triggered a build-up 73% of GDP.
process in this segment.
The new members of the EU in Central
The increasing importance of the pension and Eastern Europe (CEE) showed the most
segment is reflected in the financial assets dynamic development among the regions in
of private households. In many countries, the review. Over the past five years, they have seen 1 Former EU member
proportion of pension products has increased annual growth rates of almost 15%. Western states plus Norway and
in private household portfolios in recent years. Europe recorded annual growth rates of 7.2% Switzerland, not including
To a large extent, the differently structured on average after 2002. That year, an economic Luxembourg.
portfolios in European countries can be viewed upturn laid the foundation for resuming finan- 2 Conversion at 2007
as a result of differing pension systems and cial asset build-up in the household sector, rates.
saving behaviours. This study generally fo-
cuses on a narrow definition of retirement
Chart 1 Financial assets of private households [2000=100]
assets (i.e. pension funds and life insurance
260
reserves), as it is difficult to earmark other
financial assets as destined for retirement Western Europe
230
income. However, analysing the total finan- EU15
cial wealth of private households is a worth- EMU
200
while endeavour, as most financial products USA
can be used to top up retirement income. Japan
170
CEE
140
Trends in financial asset
development 110
At the end of 2007, financial assets in West- 80
2000 2001 2002 2003 2004 2005 2006 2007
ern Europe1 reached EUR 25.8 trillion2. This
was around 15% below the US level (EUR 30.8 Note: Europe excluding Luxembourg
Source: Central banks, statistical offices, Eurostat
trillion) and more than two and a half times
14
15. Introduction International Pension Studies Western Europe
which had ended abruptly with the stock
Chart 2 Private households’ financial asset structure, 2007 [%]
market crash at the beginning of the new
millennium. In the US, financial assets even 100 3.4 3.6 3.5 4.4
6.4
increased by 9.3% p.a. after the downturn Other
10.9
had torn an exceptionally large hole in house- 27.4 30.9 25.3 Insurance/
80 34.7
hold portfolios because of the high affinity pension funds
to capital market investments in the US. In 34.5 10.4
Shares/
mutual funds
contrast, Japanese households, with their low- 60
29.4 9.5 Bonds
risk and hence low-yield portfolios, recorded 26.0
2.4 40.6 Bank deposits
only a moderate increase of 2.7% per year. 40
7.9
6.1
However, the investment environment 45.8
50.5
changed in 2007. Growth rates dropped to 20 8.8
29.8 31.7
Note: Europe excluding
half the former average at the beginning of 16.3
Luxembourg
the downturn in housing markets and due to 0
economic uncertainties in many countries. Western EMU CEE USA Japan
Europe
The impact of the subprime crisis on financial
Source: Central banks, statistical offices, Eurostat
markets worldwide also played a major role.
In 2008, indications have thus far shown that
the growth of household wealth will slow down ments, but this is due to the privatisation
considerably. Assuming a stock market down- process in some Eastern European countries.
turn of 30 % year-end 2008 on year-end 2007
financial wealth in all major regions will even To a certain extent, bank investments in
decrease. both the US and Europe have seen a revival,
which is the result of a heightened desire for
In the US, the crisis will have an even larger safe investments. Nevertheless, US household
impact than in Western Europe, EMU coun- portfolios contain only 16% bank products.
tries and Japan. This is because the disparate In Western Europe, the bank share is almost
performance described above stems largely twice as high as in the US; in Japan, the figure
from differing household investment patterns is more than three times higher. In CEE coun-
in the respective regions. Compared with most tries, the share of bank products is almost
Europeans and the Japanese, American house- three times as high as in the US, which can
holds show a strong affinity to investments in be explained by a very low average income.
stocks and mutual funds, which bears both op- Usually, the portfolio mix initially comprises
portunities and risks for performance in the less risky and liquid assets (mainly bank de-
overall investment portfolio. posits) and shifts towards capital markets
and more sophisticated products as income
About 40% of US household portfolios is and wealth increase.
invested in equity and investment assets; in
1999, this share peaked at almost 50%. Direct
Chart 3 Share of bank products in household financial portfolios vs.
investment in corporate stocks is no longer
GDP per capita, 2007 [EUR]
as attractive as it was at that time. Indeed, on
balance, US households have reduced their 70%
exposure to equities. In Western Europe, the
60%
portion of equity and mutual fund shares Gr
CEE Jap A
currently stands at 26%, after peaking at 30% 50%
in 2000. 40%
Other Western
Eurpean countries
Nor
30%
As in the US, the increase in the overall fig-
ure in many European countries stems from 20%
valuation changes over the course of positive US
10%
stock market performance up to 2007. It is
not the result of new investment flows, as 0%
households began to invest more cautiously. 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000
[EUR]
The household portfolio mix in CEE contains
Source: Central banks, statistical offices, Eurostat
a surprisingly high portion of equity invest-
15
16. Introduction International Pension Studies Western Europe
Insurance and pension products also hampered insurance/pension asset build-up 3 In this context, the term
make up a major portion of household finan- in Greece, where only 3% of financial assets pension fund implies all
cial assets. These have benefited from strong- are allocated to this segment. This figure is types of investment in fully
er occupational and private pension provi- far lower than in all other countries. funded pension schemes.
sioning. In Western Europe, the share of these The investment vehicles
assets as part of household financial wealth Nevertheless, in most Western European have different features
has increased by 6.5 percentage points over countries, growth in the insurance/pension and names in some coun-
the last decade. Moreover, in most CEE coun- segment outperformed total financial asset tries. Wherever they are
tries, capital funded pension arrangements growth in previous years. Again, this demon- subject to regulation by
were introduced as mandatory plans. This strates the importance of insurance and pen- the insurance supervisory
gave an enormous boost to this investment sion products. Below average growth can be authorities, they are record-
segment in Europe’s emerging markets. Even seen in the mature pension markets of Swit- ed for statistical purposes
in the US, with its traditionally stronger capi- zerland and Denmark. It can also be observed in the “Insurances/Pension
tal funded pillar, improvements to the sys- in countries like Spain, where pension reforms Funds” category, in accord-
tem are an ongoing process. If IRA (Individual have yet failed to materialise. ance with ESA ’95.
Retirement Accounts) assets are also taken
into account, which are invested in other
Chart 4 Gross financial assets in Europe, 2007
product types such as mutual funds, bank
[% of GDP, private households]
accounts, insurance products or brokerage
accounts, the share of pension assets in the Switzerland 373
US further increased from an already high UK 295
level of 35% in 1997 to 40% ten years later. The Netherlands 280
Belgium 253
Italy 240
European countries Denmark
Portugal
235
223
Western Europe 218
The Western European figure for financial EMU 202
assets combines a variety of investment pat- Ireland 193
terns and levels of wealth. The four biggest France 189
Germany 188
economies combine two-thirds of total finan- Spain 182
cial assets, which amounted to EUR 25.8 trillion Sweden 165
in 2007. As the most populous country with Austria 152
Greece 139
the highest GDP, Germany ranked second be- Finland 112
hind the UK. In terms of gross financial assets Norway 103
relative to GDP, Germany is in a much lower
0 25 50 75 100 125 150 175 200 225 250 275 300 325 350 375
position (188%) than the UK, where 295% was
registered in 2007. With 373%, Swiss house-
Chart 5 Share of pension and insurance assets in the total financial assets
holds are the wealthiest in Europe, and are
of private households, 2007 [%]
also ahead of US or Japanese households.
The Netherlands 58
These states are followed by the Nether- UK 55
lands, Belgium and Denmark, all of which Denmark 43
Switzerland 42
have stronger funded pension systems than Ireland 41
countries with traditionally more extensive Sweden 40
pay-as-you-go pension systems such as Ger- France 38
Norway 36
many, Spain or Austria. The latter countries Western Europe 35
all rank in the lower part of the listing. Pre- EMU 27
sumably, the different pension systems have Germany 26
Belgium 24
an impact on asset formation and can par- Finland 21
tially explain the differing trends and finan- Portugal 18
cial asset levels in European countries. The Austria 18
Italy 17
share of insurance and pension fund assets3 Spain 14
as part of total financial assets reflects the Greece 3
importance of the segment in the respective
0 10 20 30 40 50 60
countries. The Netherlands, the UK, Denmark
Note: Europe excluding Luxembourg
and Switzerland are at the very top. Very high
Source: Central banks, statistical offices
replacement rates in the first pillar have
16
17. Introduction International Pension Studies Western Europe
Very high growth rates of pension / insur-
Chart 6 Insurance/pension asset growth vs. total financial asset growth,
ance assets compared to total financial asset
2001–2007
growth can be observed in Belgium. This has
been triggered by a new occupational pen- 16
sion scheme (“Vandenbroucke” law), estab-
14
Insurance/pension growth CAGR [%]
lished in 2004, but also life insurance products Irl
saw a strong development. Greece also saw 12
B Gr
high growth rates in pension/insurance as- N
10
I F
sets. This is likely to be the result of the gen- SP
8
erally low level reached up to now, which rep- NL
US S Dk
UK Fin
resents high potential to catch up. Ireland is 6
Ger A
another country where performance is high,
4
as it belongs to a group of countries that have CH
a very basic first pillar only. Low replacement 2
J
rates must be topped up with occupational 0
and private pension provisioning. The extreme- 0 4 6 8 10 12 14 16
ly dynamic growth of the overall Irish economy, Financial asset growth CAGR [%]
which has shown the highest average growth
Source: Allianz Dresdner Economic Research
rates in Western Europe in the last decade, has
given households the financial leeway they
need to save for retirement. The different behaviours and influences
from differing regulatory, statutory and fiscal
Other reasons for the varying development arrangements have led to strongly varying
of total financial wealth include the general trends. Depending on investment preferenc-
propensity to save and differences in invest- es, the stock market boom in the late 1990s
ment behaviour itself. As already shown for had a number of impacts on portfolios. In
the different regions of the world, risk aversion most European countries, financial assets
levels differ tremendously. Greek households grew rapidly between 1996 and 2000. This
hold the biggest share of bank products. This was related directly and indirectly to stock
risk aversion may be due to a lower level of market gains in particular, which affected
income and economic development. As one both the valuation of existing stockholdings
of the wealthier countries, portfolios in Aus- and attracted massive new inflows of funds.
tria are geared heavily towards security, with
almost 50% invested in bank deposits and The boom gripped the Finns, Swedes and
another 9% in debt securities. Spain and Greeks in particular, generating strong growth
Germany are also among the countries that between 1996 and 2000 – and losses in the
invest conservatively compared to other subsequent downturn. Still, together with
European countries. Spain and Norway, these countries experi-
enced the highest asset gains. At the end of
Savings ratio
Savings ratios are often used as an indicator of savings behaviour . This can be misleading for cross
country comparisons, as contributions to pension schemes are not counted as savings . In contrast,
the accumulated savings of pension assets do count as financial assets . The savings ratio is usually
calculated as the difference between income and consumption . Contributions to pension schemes
and investments in other financial assets, such as bank deposits, are treated differently . While the
former is recorded in national accounts as a household expense that reduces saving potential,
the latter is viewed as saving . However, since contributions to pension schemes increase financial
wealth, countries with large funded pension systems often show low savings ratios, but high finan
cial wealth (as share of GDP) . This puts the very low savings ratio of UK households compared with
the higher ratio in Germany, for instance, into a more realistic perspective .
See also: Eurostat, Savings rates in Europe, Statistics in Focus, 33/2002.
17
18. Introduction International Pension Studies Western Europe
2007, their wealth was about one and a half lead to higher asset valuations in the medi-
times higher than at the beginning of 1997. um and long term. In turn, this will provide a
In contrast, during this period Belgium and partial substitute for saving. This correlation
Germany registered respective increases of may be one of many reasons for the decreas-
only 50% and 60%. By 2005, all countries that ing savings ratios that were observed in most
had suffered absolute losses of wealth passed European countries until the beginning of the
their previous wealth peaks. And as a result new millennium. For the projection period,
of favourable capital market developments, we have assumed a share performance of 7%
they further increased their financial assets p.a. from 2009 onwards.
until 2007.
In 2008, the financial crisis that resulted
from the subprime crisis in the United
Financial asset projection States has put a burden on stock markets.
through 2020 The high 2007 year-end figures are unlikely
to be reached by the end of 2008. Our fore-
Moving forward, we expect the acquisition casts were undertaken in the third quarter
of financial assets to progress dynamically. of 2008 and we have assumed that stock
In many European countries, the develop- markets lose 30% in 2008. These perform-
ment of monetary wealth will be driven by ance figures also influence the assets under
efforts to provide for retirement. This will not management of investment funds.
only be the case in countries with already
strong funded pension systems (such as the Investments will flow into a wide variety of
UK, the Netherlands and Switzerland), but product categories. A precise product-specific
also in states where reforms to public pay- forecast or econometric estimate is virtually
as-you-go systems will lead to lower pension impossible for a longer period, as preferences
levels. Particularly in these countries (such as for particular investment vehicles may depend
Germany, Italy and Austria), people are grad- on factors that are difficult to predict, such as
ually learning to accept the need for supple- legal or tax circumstances and interest rate
mentary private provision and are building developments. For this reason, our assump-
up capital accordingly. Although this process tion on asset allocation is based on long-term
will develop differently across European coun- historical structural shifts. These are driven
tries and depends on awareness and the by rising incomes and personal wealth, which
progress of reform, sales of products for old- lead investors to higher performing, riskier
age provisioning are likely to be very suc- products. Although stock market volatility
cessful in the coming years. during this decade has reinstated the security
aspect, the basic change in investment be-
Moreover, the need for personal pension haviour with increasing income is likely to
provision will encourage broader sections of remain stable.
the population to save beyond state-incentiv-
ised pension schemes. In most countries, our In light of these patterns, countries with
projection therefore assumes a slightly high- higher portions of direct or indirect exposure
er savings rate than in the past. However, the to equity markets (through mutual or pension
ratios will barely get back to the levels of the funds), such as the UK, Sweden, Finland, Ire- 4 This growth rate differs
early 1990s. Additional retirement saving will land or Spain, will see higher growth in house- slightly from that in the
likely replace other precautionary savings hold financial assets than countries with a following article on pen-
efforts and the increasing number of elderly more conservative investment approach or sion assets. This is due to
people will reduce their personal savings smaller funded pension systems. On average, some classification dif-
efforts as they retire. we expect the financial wealth of Western ferences in the financial
European households to increase by 4.4% p.a. flow statistics of national
The increase in wealth is driven not only accounts and the specific
by savings, but also by asset valuations, as Total financial assets will therefore increase pension statistics we used
the development of financial assets has dem- by 75% until 2020, reaching almost EUR 45.3 for the market analysis.
onstrated in past decades. Stock market per- trillion. The strongest growth will be registered One major difference is
formance also represents an important part in the area of insurance and pensions (5.4%)4. that non-life insurance
of the increase in monetary wealth. Rising By 2020, the importance of these products in products are included in
investment in equities and equity funds will household portfolios will gain 3.5 percentage the financial accounts.
18
19. Introduction International Pension Studies Western Europe
points. Above average growth will also be ob-
Chart 7 The financial asset structure of private households, 2007–2020
served in the shares and mutual fund segment
(4.6%). This will extend their portfolio share
100 3 3
by roughly one percentage point. The main 3.4 3.6
Other
losers in this process will be bank products.
27.4 32 Insurance/
80 34.7 39 pension funds
In Continental Europe, particularly in coun-
Shares/
tries that belong to the monetary union, growth mutual funds
60
rates in the insurance and pension segment 29.4 Bonds
26.0 31
are slightly higher (5.6%) due to emerging pen- 27 Bank deposits
sion saving programs. With its above-average 40
7.9
6.1
growth rate, the UK market is not included. 5
7
In EMU countries, financial assets will grow 20
31.7
by only 4.3%, amounting to EUR 30.9 trillion 29.8 26 27
Note: Europe excluding
in 2020 from EUR 17.8 trillion in 2007. Luxembourg
0
2007 2020 2007 2020
Prospects for the long-term development Western Europe EMU
of financial assets are quite good, as growth
Source: Central banks, statistical offices, Eurostat, Allianz Dresdner Economic Research
rates are expected to be higher than GDP
growth. However, it should be noted that the
financial turbulence of 2007/08 is yet again At the same time, introducing fully-funded
putting individual investors and long-term pensions in industrialised nations has be-
savers in capital market products to the test. come more pressing than ever.
Explanatory notes
Financial assets
The household sector’s financial assets are calculated in the financial flow statistics provid-
ed by central banks or national statistical offices. The aggregate financial account shows
by whom, on what scale and in which form financial resources have been made available
in an economy. The system is aligned to the “European System of Accounts” (ESA), which
was made binding in the European Union in 1999. Its aim is to provide a consistent set of
criteria with which all economic sectors and activities can be defined.
Insurances and pension products comprise one product group in the financial accounts.
In this category, investment vehicles are recorded that are subject to regulation by insurance
supervisory authorities, in accordance with ESA ’95. These include all types of investment in
fully-funded pension schemes and insurance products. Prepayments of insurance premi-
ums and reserves for outstanding claims are also included in this segment. The values are
technical reserves. Insurance reserves and pension fund assets are not separated for all
countries. This means that non-life, life and pension assets are shown in an aggregate fig-
ure in the analysis of financial assets. This figure differs slightly from the definition in the
“retirement asset” part of the study (see next article).
Projections
While accruals in financial assets can be traced back to valuation changes and flows, flows
come from savings and contributions from other sources. As far as saving is concerned,
assumptions must be made on the development of disposable income and savings rates.
For disposable income, an increase in line with nominal GDP growth has been assumed.
The growth forecasts for real GDP and inflation in the individual countries up to 2020 are
based on Allianz Group Economic Research projections. The data on savings rates up to
2009 are sourced from OECD statistics; constant or slightly increasing savings rates have
been assumed for the respective countries.
19
20. Introduction International Pension Studies Western Europe
Particularly with regard to allocation, assumptions are difficult to make, as preferences
for special types of investment depend on interest rate trends or legal or tax conditions,
for instance. Inflow allocation to the various financial instruments has therefore been based
on the average behaviour of the past 10 years. Allowance has further been made for the
likelihood of additional funds being channelled into private retirement provision in any one
country, insofar as there are newly introduced private pension plans that are incentivised
by the government.
Dr. Renate Finke,
Allianz Dresdner Economic Research
20
21. Introduction International Pension Studies Western Europe
Trends and Asset
Development in European
Pension Markets
Reforming public pensions Nevertheless, almost all countries have
trimmed their public pension systems to
Western Europe’s pension systems have varying degrees and increased the retirement
changed continuously and considerably. age to strengthen the sustainability of their
For the most part, reforms of the last decade public pension systems. Countries like Swe-
were triggered by the insight that ageing den, Austria and Italy have established a very
Western European populations would place strong link between contributions and bene-
an unbearable burden on the public pension fits in the public pillar. They have introduced a
systems in place in the medium and long term. notional defined contribution system in which
This is because the ratio between contributing contributions are recorded in notional indi-
employees and retirees will worsen, in some vidual accounts and benefits depend on the
countries dramatically. The degree of change accumulated sum and cohort life expectancy.
and the depth of reform have differed across
Europe. This is hardly surprising, as the pen- Despite these reforms, Western Europe’s
sion policies of Western European countries public pensions are still very generous in
have been based on very different foundations. a worldwide comparison. Some European
countries show exceptionally high replace-
Western Europe’s public pension policies ment rates, which are defined as the ratio of
are largely based on two different systems. post- to pre-retirement income. In Italy and
Bismarckian systems comprise public pen- Portugal, the net replacement rate for aver-
sions that provide earnings-related benefits age earners is around 90% of pre-retirement
and aim at maintaining income in retirement. income. In Luxembourg, Spain and Greece,
In contrast, Beveridgean systems mainly aim it is 100% or more. This dominating public
to prevent poverty, often through flat-rate pillar will have a strong impact on public
systems. Austria, Belgium, France, Germany, finances in the future, which will influence
Italy and Spain belong to the former group, the need for pension system reform.
while Ireland, the Netherlands and the UK
are examples of the latter. Mixed models can The Allianz Dresdner Reform Pressure Gauge
be found in the Scandinavian countries and summarises the effects at work, compares and
Switzerland. Since Bismarckian systems are illustrates the sustainability of pension systems
associated with higher public pension expe- and the need to reform. It includes the likely
ditures, they have also faced a much greater effects of reforms already initiated. In so doing,
need for reform. Another factor that has de- it assesses the future sustainability of pension
termined the urgency of reform has been a systems and shows that reform pressure is
widely differing demographic outlook from highest in Greece. Generally, countries with
country to country. While Spain, Italy, Por- very high replacement rates and underdevel-
tugal, Greece and Germany must cope with oped funded systems are under the highest
strongly ageing societies, populations in reform pressure, as they will face dramatic
Switzerland, the Netherlands, the UK, Ireland increases in public pension expenditure in the
and Scandinavia are not ageing quite as future. Portugal, Spain and Luxembourg face
dramatically. this very situation. For Sweden, the UK, Den-
mark and the Netherlands, the Pressure Gauge
21