Population aging, financial markets and monetary policy
Speech by Lucas Papademos, Vice-President of the European Central Bank at the conference “Exploring the Future of Pension Finance and the Dynamics of Institutional Pension Reform”,Amsterdam, 23 March 2007
The magnitude of the demographic changes unfolding in Europe poses substantial challenges for policy makers in virtually all areas and at both micro and macro levels. While this contribution cannot address all pertinent issues, it will focus on the key implications of population aging for the conduct of monetary policy and the development and functioning of financial markets. However, even this more focused subject requires the examination and assessment of an extensive set of potential effects and pertinent policy issues.
2 Population aging and monetary policy: an overview
Population aging may have implications for monetary policy as a consequence of its possible effects on the economic and financial environment within which monetary policy is conducted. It can affect key features and the functioning of the real economy, as well as the structure and development of financial markets. More specifically, population aging can have a significant effect on the following:
a. long-term economic growth, by inﬂuencing the economy’s potential growth rate and the equilibrium real rate of interest;
b. the dynamics of aggregate demand and the monetary policy transmission mechanism, by inﬂuencing the intertemporal allocation of saving and consumption, as well as the determination of asset prices and portfolio diversification;
c. the structure of financial markets and the role of different financial institutions in the intermediation process;
d. public finances, by the increasing expenditure related to public pension systems, most of which are based on the pay-as-you-go principle, and to public health and long-term care systems; and
e. international capital ﬂows, to the extent that the pace of the aging process and the evolution of available investment opportunities differ significantly across countries.
Thus the potential impact of population aging on the structure and functioning of the economy and on the conduct of monetary policy could be substantial and far-reaching. At the same time, this impact is very gradual, fairly complex and highly uncertain. It is likely to become visible not within years, but rather over decades. The ‘glacial’ nature of these effects – to use a term coined by Charles Bean  – makes it difficult to link the long-term implications of population aging for the economy directly to the decisions taken by a central bank over a medium-term horizon. Moreover, the ultimate impact of population aging is conditional on the response of public authorities, the financial markets and the economic agents themselves.
In my presentation, I will examine in greater detail some of the potential effects of population aging on key macroeconomic variables and the monetary policy transmission mechanism and assess the implications of such effects for the conduct of monetary policy. I will also give an indication of the expected order of magnitude of some of these effects, while always keeping in mind the uncertainty and the conditionality surrounding any empirical estimates and the associated policy conclusions.
3 The potential impact of population aging on long-term growth and the equilibrium real interest rate
Let me first consider the implications of demographic change for two variables which are of particular relevance for policy-making, namely the economy’s potential growth rate and the equilibrium ‘risk-free’ real interest rate. To explain why we should expect that exogenous demographic factors will, over time, exert considerable ‘pressure’ on key endogenous variables such as the potential growth rate and the equilibrium interest rate, I will point to three stylized demographic developments for the euro area derived from recent projections.  First, we are witnessing a marked slowdown in the growth rate of the working age population in Europe. In fact, from 2010, the working age population is expected to shrink. Secondly, life expectancy at birth is predicted to increase further in the future, rising from the current level of 78 years to around 83 years in 2050. Thirdly, the combination of these two developments (the reduced inﬂows into the labour force and the increased life expectancy) will result, ceteris paribus, in a steady increase in the old-age dependency ratio (which will rise from 26 per cent in 2006 to around 55 per cent in 2050). These developments will, indeed, result in significant changes to the macroeconomic landscape.
The effects of the expected demographic change on potential output growth over the long term cannot be predicted with precision. However, the direction of the impact is fairly clear (from first principles) and an initial estimate of its likely magnitude can be obtained if we assume that other determinants of potential growth remain unchanged. Let me illustrate this by reviewing a number of findings from a recently completed study by ECB colleagues, which is based on a comprehensive growth accounting framework. Real GDP growth can be decomposed into three determining factors: the rates of change of labour productivity, labour utilization and the working age population.  To generate concrete forward-looking scenarios, the study takes as a benchmark the average growth rates of these variables over the period 1995–2005. If we combine, in a first scenario, the projected likely evolution of working age population growth in the coming decades with the benchmark values for the growth rates of labour productivity and labour utilization, we find that potential output growth gradually declines, falling by nearly 1 percentage point by the year 2050.
Of course, it can – and, indeed, should – be argued that the trend growth rates of labour productivity and labour utilization are not going to remain unchanged, and that there is substantial scope for improvement in these two variables, which can be fostered through structural reforms. That said, to imagine that increases in the growth rates of labour productivity and labour utilization could entirely offset the negative contribution to potential growth by the projected decline in the working age population requires a high degree of optimism. To understand what this means in numerical terms, consider the following: in order to maintain potential output growth until 2050 at the current level of around 2 per cent, average labour productivity growth in the euro area would, ceteris paribus, have almost to double by the year 2030 from the benchmark value of 1 per cent to a level close to that recorded in the United States in the period 1995–2005, which was around 1.8 per cent. And such a significant average annual improvement in productivity would have to be achieved by an aging society, meaning that enhancing economic dynamism and innovation would represent a formidable challenge. Alternatively, if labour productivity growth were to remain at the benchmark value of 1 per cent, labour utilization growth would have to increase substantially from its benchmark value of 0.8 per cent by 2050. Although further significant improvements in labour utilization are certainly possible and desirable, there are natural limits which would be reached well before 2050 under such an (admittedly ambitious) scenario. 
Forward-looking growth projection exercises of this type are insightful and thought-provoking. However, I should stress that they do not capture adequately all general equilibrium interactions between the factors of production. In this respect, it is important to look at the interactions between these factors not only in the context of a closed economy, but also from an open economy perspective; that is, by examining the possible interactions between the euro area economy and the rest of the world. To illustrate this, let me review and assess some findings on the likely future path of the riskfree equilibrium interest rate, following a two-step approach which first focuses on a closed economy and then considers the issue in an open economy context.
Theoretical analyses of the likely impact of population aging on the real interest rate are often based on overlapping generations models, which have the advantage of being able to capture the effects of demographic changes and their impact on life-cycle savings in a concrete and transparent way. In the context of a closed economy, these models typically predict that over the next 25–50 years the expected demographic trend will result in a reduction of 50–100 basis points in the real interest rate.  This is a relatively broad range, which reﬂects the different assumptions made in the various studies regarding the likely effects that the slowdown in population growth and the increase in life expectancy will have on the capital–labour ratio and the saving rate, on the one hand, and on the potential shift toward funded pension systems, on the other hand.
Let me brieﬂy elaborate on these consequences. First, a slowdown in working age population growth implies, all other things equal, that labour becomes scarcer than capital. This calls for a change in factor prices, as fewer people need to be ‘equipped’ with capital, and leads to an increase in the capital–labour ratio, resulting in a decrease in the real interest rate.  Secondly, an increase in life expectancy can reasonably be expected to increase the incentive to save, and this, ceteris paribus, also places downward pressure on the real interest rate. Thus the fact that different assumptions are made concerning the size of the likely impact of these two demographic trends on the real interest rate explains the wide range of predictions as regards the magnitude of that decline. Moreover, alternative hypotheses about the features and the design of future pension systems has significant implications for the predicted effects. The likely decline in the real interest rate will be larger in scenarios which assume, ceteris paribus, that reforms will result in a substantial shift towards funded pension systems, which will make the motive for additional savings particularly strong. On the other hand, scenarios based on the assumption of an increase in the effective retirement age imply that the effects on the interest rate are mitigated because the profile of savings is less steep over the life cycle.
These type of models suggests that the long end of the yield curve could also experience a decline as a result of expectations of lower short-term interest rates in the future that are priced in the yields of long-term bonds. The magnitude of that impact will depend on the size of the expected decline in the equilibrium short-term real interest rate and on the extent to which financial markets form expectations about the effects of other factors and possible future policies which could moderate or even offset a decline in the equilibrium real rate stemming from the aging process. There is no clear-cut evidence as regards the magnitude of this impact, which can be presumed, at present, to be small and surrounded by a substantial degree of uncertainty.
This reasoning on the likely effects of aging on the equilibrium interest rate requires further qualification if one considers the issue from an open economy perspective. Demographic developments are not identical across major regions of the world. Two features are especially relevant in this respect. First, within the group of OECD countries, the effects of the aging process in Europe and, in particular, Japan are mitigated by a more benign aging outlook for the United States. Secondly, in most emerging market economies the phenomenon we refer to as ‘aging’ will not become an issue until some time in the future. These features should, in principle, mitigate the downward pressure on the real interest rate in Europe. In particular, open economy models suggest that capital should be ﬂowing from developed economies (which have older populations and richer capital endowments, and where investment opportunities are relatively limited) to emerging market economies (where the population is younger, capital is scarcer and investment opportunities are relatively more abundant).
So much for the theory. What evidence is available from the world economy? It appears that capital is not exactly moving in the expected direction: we are currently observing a ‘savings glut’, which is originating mainly in Asia.  There are of course a number of factors that explain why this is the case. However, when we examine the demographic patterns, we find that some important emerging market economies, notably China, are not only growing very rapidly, but are also aging relatively rapidly, while their public pension systems are very inadequate or even non-existent. From this point of view, the pattern of international capital ﬂows observed over the past few years could be partly explained by demographic developments, although specific exchange rate and trade policies have certainly affected these ﬂows. Looking forward, while there is substantial scope for increased international factor mobility, I share the view that factor price differentials between the euro area and the rest of the world will not be entirely arbitraged away, either because of political barriers (which can restrict, in particular, the mobility of labour) or because of market-based risk concerns (which can impose limits on international capital ﬂows across countries and regions). In conclusion, open economy considerations tend to mitigate, but do not entirely eliminate, the downward pressure of the aging process on the real interest rate discussed earlier.
The potential implications of aging for the economy’s long-term growth and the equilibrium real interest rate that I have examined so far will of course affect the environment within which monetary policy operates. One important issue for policy makers is whether these effects, which imply gradual, but fundamental, changes to key aspects of the economy, will require any substantive change to the monetary policy strategy. I will address this issue with regard to the two key features of the ECB’s policy strategy: its objective and analytical framework.
It is self-evident that the primary objective of the ECB’s monetary policy to maintain price stability should not be affected by population aging. Indeed, I would argue that the pre-eminence of price stability as a policy goal of the central bank would be reinforced in an aging society because there is a greater need to ensure that the value of a growing proportion of wealth invested in nominal assets whose return is not likely to be indexed to inﬂation is not eroded by inﬂation.
Another issue that has been raised is whether the expected increase in the old-age dependency ratio and the wealth-to-income ratio might strengthen the case for the central bank to ‘lean against the wind’ in the face of asset price misalignments, because the welfare losses from boom-and-bust cycles in asset prices would presumably be larger in an aging society. If so, would this effectively imply a broadening of the central bank’s objective? This is clearly not the case. However, this argument underscores the usefulness of the ECB’s analytical framework, in which the results of the economic analysis, which focuses on identifying risks to price stability over the short to medium term, are cross-checked with the signals derived from the monetary analysis, which seeks to identify risks to price stability from a medium to longer-term perspective. This analytical framework allows the ECB to detect early warning signals about emerging asset price misalignments. More generally, it is also crucial that we continue to monitor and analyse carefully the way in which demographic forces, associated financial market developments and pertinent reforms could affect the monetary transmission mechanism over time, and in particular impact of aging on the functioning of financial markets. This is the subject on which I will now focus.
4 Population aging and financial markets
In an aging society the relative importance of young and old people, in terms of size of those groups, changes in favour of the latter. Financial markets are affected by this shift, because these two groups of people have different saving and investment preferences and different financing needs. Consequently, they can be expected to make different use of financial markets and instruments. In economic theory, the well-known life-cycle hypothesis of saving suggests that individuals seek to smooth out consumption relative to their income over their lifetime. This implies that people accumulate savings during their working life and acquire assets so that, after they have retired, they can sell them, and thus dissave, in order to fund consumption. Moreover, older people can be expected to become more risk-averse as the time horizon of their investment shortens. These considerations raise three pertinent questions:
a. First, is the life-cycle hypothesis of saving supported by the empirical evidence in the euro area?
b. Secondly, if so, how would the implied changes in the aggregate saving pattern affect asset prices (such as bond, equity and real estate prices) in the context of an aging society?
c. Thirdly, what kind of changes to the financial intermediation process could we expect as a result of the differing financial requirements of a ‘greyer’ population?
With regard to the first question, a look at the evidence from households in the euro area shows that people do not actually seem to behave as postulated by the simple, stylized life-cycle hypothesis: in a number of euro area countries, old people tend to be net savers and even spend significantly less than working households.  That said, if we consider a more general specification of the life-cycle hypothesis, which takes into account the bequest motive, the existence of public social security systems and the fact that people live out of their pension after retirement, the implications of the generalized life-cycle theory are consistent with the facts. Pensions are, after all, equivalent to annuities stemming from rights accrued in the past, and in that sense, they are a form of wealth.  However, if the benefits of public pension systems were to be significantly reduced, people’s behaviour might change, as they would have more of an incentive to increase their private savings during their working years in order to use them when they retire.
The life-cycle hypothesis of saving implies that an economy with an aging population will be characterized by a relatively high wealth-to-income ratio. A greater reliance on private savings and accumulation of wealth, following a pension reform, in order to sustain a certain consumption level and a longer life expectancy after retirement, would reinforce the effect of wealth on consumption. Therefore, the transmission of the effects of monetary policy on aggregate output and the price level via the wealth channel can be expected to become progressively more powerful as the baby-boom generation approaches retirement. The increasing significance of this channel will be gradual and its magnitude can only be ascertained empirically. At the same time, the bank-lending channel may become relatively less significant, as it tends to reﬂect the demand for credit by agents that are younger and with a lower level of wealth.
4.1 The potential impact of aging on asset prices
To what extent and in what way do the increased saving of a diminishing group of younger people, and the greater dissaving of a growing group of pensioners) inﬂuence asset prices? Two aspects are relevant and deserve particular attention in this respect: (a) whether, and to what extent, retired people can actually sell or mobilize the assets accumulated during their working years to finance their consumption during old age; and (b) whether, and to what extent, the inﬂow of a significant amount of funds into the pension fund industry will affect the prices of particular financial instruments, notably long-term bonds.
Are the assets accumulated by people of working age actually tradable after those people retire? This is a highly pertinent question, as the largest share of households’ wealth in the euro area – around 60 per cent – is real estate. Of course, we should keep in mind that, in most cases, houses are often used to live in, which in turn makes the households’ consumption pattern less dependent on the ‘monetary value’ of residential investment. Nevertheless, some recent studies suggest that investment risk in residential real estate is likely to increase over the coming 30 years as a consequence of population aging in the EU.  As a result of increased regional and crossborder movements of people, regional discrepancies in the real estate markets are correspondingly growing, among urban centres, rural areas and coastal areas. Against this background, large price swings cannot be ruled out in the longer term, in the event that a large number of retirees would like to sell their real estate property in a certain region.
But a potential mismatch of this type between buyers and sellers of assets in a domestic market could be mitigated by foreign investors, provided that the relevant assets were in fact internationally tradable. More generally, the more assets are internationally tradable, the smaller the impact of aging on asset prices will be, since demographic patterns and the age structures of populations differ across countries. Incidentally, it should be noted that there has been much attention to the possible impact of aging on the portfolio reallocations between bonds and stocks and very little on the related consequences for property prices. 
The second aspect I would like to address with regard to the impact of aging on asset prices concerns the way in which pension funds and institutional investors allocate the growing inﬂow of funds within their financial portfolios. Given the size of the funds under management, even small changes in the way they invest their portfolio and manage their liabilities could potentially have a significant impact on the price of assets in the financial markets. In particular, the price of long-term bonds, which determines the level of long-term interest rates, may increase because the potential demand for long-term bonds by institutional investors – and pension funds in particular – may be significantly higher than the outstanding supply of such bonds. This proposition has been broadly confirmed by recent analyses.
That said, the possible impact on bond prices of this excess demand for long-term bonds remains very hard to disentangle quantitatively from other factors.  Up until now, there is very limited empirical evidence that changes in institutional investors’ asset allocation have exerted a significant impact on bond yields. However, in some cases – notably in the United Kingdom – there are some indications that long-term yields have declined partly as a consequence of regulatory developments which have affected the portfolio allocation of institutional investors.  In the short to medium run, it is difficult to foresee a sizable increase in the supply of long-term bonds. Over a longer horizon, however, governments, international institutions and other highly-rated entities could increase the issuance of long-term bonds. The net impact on long-term yields of these factors would be expected to become greater in the future, as the effects of regulatory policy reforms unfold in other countries and the importance of institutional investors increases. Moreover, this impact might be augmented by the additional and more general downward pressure on the long end of the yield curve stemming from the possible general decline in interest rates owing to population aging.
4.2 Aging and the financial intermediation process
Let me now turn to some of the consequences of population aging for the financial intermediation process and the structure of financial systems more generally. We are already observing important changes in the role and functioning of the retirement savings industry. Confronted with population aging, governments are finding it increasingly difficult to maintain the entitlements of the existing public pension schemes. Many governments are moving in the direction of reducing benefits relative to contributions to public pension schemes, reducing the ratio of the post-retirement income to the net pre-retirement income. Faced with these changes or prospects, individuals will be increasingly supporting their old-age consumption via alternative pension plans. Thus it can be expected that a larger proportion of household savings will be placed in privately funded pension schemes and life insurance policies.  Accordingly, over the coming decades, a significant increase can be expected in the value of assets managed by the retirement savings industry, particularly in countries where this industry is less developed and there are underfunded public pension systems with defined benefits. The Netherlands is an exceptional case among the euro area countries in this respect, since in this country the pension fund industry is of a significant size.
As key financial intermediaries, banks will, of course, also be very much affected by population aging. What are the likely effects? Given the long-term nature of the effects of population aging and the wide range of possible counterbalancing forces, the answer to this question is highly tentative and should not be interpreted as a prediction. Demographic changes – through their impact on GDP and population growth rates – may result in a downward pressure on banks’ intermediation ratios and to reduced demand for consumer credit and mortgages, lowering interest income. Banks will face incentives to move away from activities linked to liquidity and duration transformation and to increase revenue and value added per customer, for example by cross-selling non-bank financial products and focusing on services with higher added value. At the same time, banks face increased competition from other intermediaries and they might partly respond by offering new products tailored to ‘senior citizens’, for instance consumer loans, in order to finance increasing health and long-term personal care costs, and reverse mortgages, aimed at extracting value from real estate. Banks will also respond by searching for new markets. Such prospects are likely to foster the formation of large financial groups and the strengthening of cooperation between banks, insurance companies and mutual funds.
The blurring of boundaries between segments of the financial sectors (banks and other intermediaries) also has a number of regulatory implications because of the potential increase in cross-sector contagion. A number of regulatory policy reforms have been launched recently or are forthcoming in the pension and insurance sectors, partly triggered by the problems of underfunding in defined-benefit pension plans which emerged in recent years as a result of demographic changes.
5 Implications for central bank policy
What are the consequences of these developments for central bank policy? A first consequence is an increasing need to monitor carefully possible changes in the transmission of the effects of monetary policy particularly through the wealth channel, but also through the bank channel, in the light of the increasing importance of funded pension systems and private savings arrangements. As I indicated earlier, portfolio reallocations on the part of institutional investors could, at times, exert downward pressure on long-term yields that is not directly related to macroeconomic fundamentals or other structural factors. This, in turn, may lead to ‘excessive’ borrowing by the private sector and to temporary distortions in the allocation of resources. Another important policy issue is related to the increasing use by pension funds and insurance companies of novel, sophisticated and complex financial instruments to transfer risks on their portfolios to other market participants. Although this risk transfer entails benefits, it tends to complicate the assessment, by central banks and bank supervisors, of the outlook for financial stability, since it is becoming more difficult to keep track of the way the risks are actually distributed and the extent to which those ultimately bearing the risks are actually capable of doing so in situations of stress. How to address these financial stability concerns effectively is an important policy issue for central banks.
At the same time, taking into account the current and anticipated need for instruments capable of managing interest-rate, inﬂation and longevity risks, and possible demand–supply imbalances in financial markets other than the long-term bond market, policy makers have an important role in promoting the development and the efficient functioning of financial markets for the pooling and the transfer of risks. This task may also involve a more active role for government debt management offices which may promote the issuance of long-dated and inﬂation-linked bonds while taking decisions related to the management of the debt maturity of such bonds.
Lastly, I would also like to point to another issue from the perspective of the individual. The amount of risk borne by individual households depends on the design of pension schemes. Households’ wealth is becoming more exposed to financial markets, and retirement income is subject to greater investment risk than before. This underscores the importance of fostering efforts to increase financial literacy among households, which – as survey evidence suggests – is limited, and it seems to be overestimated by the individuals themselves. 
6 Concluding remarks
I have examined a fairly long list of issues pertaining to the potential effects of population aging on the real economy and the financial markets, and the possible implications of such effects for the conduct of monetary policy and the performance of other central bank tasks. I will not attempt to summarize the specific conclusions which have emerged from this examination. I would like, however, to stress a few general points.
Firstly, population aging could have implications for the monetary policy as a consequence of its possible effects both on the economy’s aggregate supply and demand, and on the channels and dynamics of the monetary transmission mechanism, as also determined by the evolving structure of financial markets and the role of financial intermediaries. The aging process can affect both the long-term equilibrium values of key macroeconomic variables and important factors determining the dynamic response of the economy to shocks and policy actions.
Secondly, over an extended period of time, the cumulative effects of population aging could be significant and far-reaching, but the materialization of these effects is very gradual and their magnitude is surrounded by a high degree of uncertainty. Moreover, policy action (notably reforms in labour and product markets and in pension and health care systems) and the economic agents’ responses to the expected demographic change and to the implementation of reforms will determine the ultimate impact of population aging on the real economy and financial markets.
Thirdly, another general conclusion is that the potential effects of population aging on the European economy should be mitigated as a result of the globalization of financial markets and the fact that demographic developments are not identical across major regions of the world.
Fourthly, with regard to the implications of population aging for the conduct of monetary policy, three points should be made:
The effects of population aging on the real economy and the financial markets do not require any changes to the central bank’s strategy monetary policy.
The ECB’s policy objective remains unaffected and, indeed, the primacy of price stability would be reinforced in an aging society to ensure that the value of a growing proportion of wealth invested in nominal assets whose return is not indexed to inﬂation is not eroded by inﬂation. It has been argued that the expected increase in the old-age dependency ratio and the wealth-to-income ratio could strengthen the case for ‘leaning against asset price misalignments’, as the welfare losses from boom-and-bust cycles in asset prices would presumably be larger in an aging society. Such an argument does not imply the adoption of an additional policy objective, but rather underscores the usefulness of the ECB’s strategy and analytical framework, which can help provide early warning signals about emerging asset–price misalignments.
It is essential to monitor and analyse carefully the way in which demographic forces, associated financial market developments and pertinent reforms could affect the monetary transmission mechanism over time, as this will have implications for the conduct of monetary policy. More theoretical and empirical research is needed in order to enhance our understanding of these effects, to estimate their quantitative significance and to facilitate their monitoring.
My final point is more general but not less relevant for the conduct of monetary policy in an aging economy. There is an urgent need to continue to implement the appropriate reforms in labour and product markets as well as in pension and healthcare systems so as to counteract possible adverse effects of population aging on Europe’s future growth and prosperity. Even though these effects will start to emerge over the next few decades, the reforms should be implemented sooner rather than later, because it will take a certain period of time before they achieve their full impact and because the political support for these reforms may diminish as the majority of voters approach the retirement age. Such reforms would also facilitate the conduct and enhance its effectiveness in preserving price stability and thus sustaining economic growth.
Alexopoulou, I., F. Drudi and J. Scheithauer (2006), ‘What accounts for the low level of the term premium?’, background paper for the BIS report Institutional investors, global savings and asset allocation, December.
Ang, A. and A. Maddaloni (2005), ‘Do demographic changes affect risk premiums? Evidence from international data’, Journal of Business, 78(1), 341–79.
Antolin, P., F. Gonand, C. de la Maisonneuve, J. Oliveira and K.Y. Yoo (2005), ‘The impact of aging on demand, factor markets and growth’, OECD Economics Department Working Paper No. 420.
Batini, N., T. Callen and W. McKibbin (2006), ‘The global impact of demographic change’, IMF Working Paper No. 06/9.
Bean, C. (2004), ‘Global demographic change: some implications for central banks’, Overview Panel, Federal Reserve Bank of Kansas City Annual Symposium, Jackson Hole, Wyoming.
Bernanke, B.S. (2005), ‘The global savings glut and the U.S. current account deficit’, remarks at the Homer Jones Lecture, St Louis, Missouri, April.
Börsch-Supan, A. and J.K. Winter (2001), ‘Population aging, savings behaviour and capital markets’, NBER Working Paper No. 8561.
Börsch-Supan, A., A. Ludwig and J.K. Winter (2004), ‘Aging, pension reform, and capital ﬂows: a multi-country simulation model’, University of Mannheim Discussion Paper No. 04-61.
Börsch-Supan, A., F. Heiss, A. Ludwig and J.K. Winter (2004), ‘Pension reform, capital markets and the rate of return’, German Economic Review, 4(2), 151–81.
Canton, E., C. van Ewijk and P.J.G. Tang (2003), ‘Population aging and international capital ﬂows’, European Network of Economic Policy Research Institutes, Occasional Paper No. 4.
De Serres, A., C. Giorno, P. Richardson, D. Turner and A. Vourc’h (1998), ‘The macroeconomic implications of aging in a global context’, OECD Economics Department Working Paper No. 193.
ECB (2006a), ‘Demographic change in the euro area: projections and consequences’, Monthly Bulletin, October.
ECB (2006b), ‘The impact of aging on EU banks’, EU banking structures, October.
Economist, The (2006), ‘Caveat investor’, 12 January.
Kara, E. and L.v. Thadden (2006), ‘Monetary policy aspects of demographic changes’, European Central Bank, mimeo.
Krueger, D. and A. Ludwig (2007), ‘On the consequences of demographic change for rates of return to capital, and the distribution of wealth and welfare’, Journal of Monetary Economics, 54(1), 49–87.
Maddaloni, A., A. Musso, P. Rother, M. Ward-Warmedinger and T. Westermann (2006), ‘Macroeconomic implications of demographic developments in the euro area’, ECB Occasional Paper Series, No. 51, August.
Miles, D. (1999), ‘Modelling the impact of demographic change upon the economy’, Economic Journal, 109(452), 1–36.
Poterba, J.M. (2004), ‘Population aging and financial markets’, paper presented at the conference Global Demographic Change: Economic Impacts and Policy Challenges organised by the Federal Reserve Bank of Kansas, Jackson Hole.
 See C. Bean (2004).
 See A. Maddaloni, A. Musso, P. Rother, M. Ward-Warmedinger and T. Westermann (2006).
 See A. Maddaloni et al. (2006).
 However, it is worth emphasizing that the required improvements are smaller if one conducts the thought experiment of keeping the trend growth of output per capita in line with the benchmark values. That variable is also arguably more relevant from a welfare perspective.
 For further details on the reasoning in this and the next paragraph, see, among others, D. Miles (1999), N. Batini, T. Callen and W. McKibbin (2006), P. Antolin, F. Gonand, C. de la Maisonneuve, J. Oliveira and K.Y. Yoo (2005), A. Börsch-Supan, A. Ludwig and J. K. Winter (2004), A. Börsch-Supan, F. Heiss, A. Ludwig and J. Winter (2004), E. Canton, C. van Ewijk and P.J.G. Tang (2003), A. De Serres, C. Giorno, P. Richardson, D. Turner and A. Vourc’h (1998), E. Kara and L.v. Thadden (2006), D. Krueger and A. Ludwig (2007).
 The impact can be ambiguous, but most studies agree that capital will fall by less than labour and that, consequently, the capital–labour ratio, the crucial determinant of the interest rate in neoclassical growth models, will increase.
 See Ben S. Bernanke (2005).
 Incidentally, I would like to stress the need for more accurate and comparable data on household ﬁnance and consumption, especially in the euro area. The Eurosystem is currently evaluating the possibility of conducting a household ﬁnance and consumption survey in the euro area, and a network of experts has already been established in order to prepare a full proposal for such a survey.
 See, for example, A. Börsch-Supan and J.K. Winter (2001).
 See chapter 2 in the ECB report EU banking structures, October 2006.
 There is an extensive literature on the impact of demographic changes on stock prices. See, for example, J.M. Poterba (2004). For an empirical analysis of the impact in stock markets around the world, see A. Ang and A. Maddaloni (2005).
 For an overview of pertinent studies, see the BIS report, ‘Institutional investors, global savings and asset allocation’, prepared by a working group established by the Committee on the Global Financial System, December 2006, page 24.
 See I. Alexopoulou, F. Drudi and J. Scheithauer (2006).
 For developments in the euro area, see ECB (2006a).
 See ‘Caveat investor’, in The Economist, 12 January 2006.