Saving for retirement and other factors lead to a distinct age profile with regard to wealth holdings and personal debt. The links between wealth and age have important implications for global ageing in future, increasing demand for financial assets, for example, relative to real assets like housing. A key issue is the strain that population ageing will place on public pension systems and other supports for the elderly, which is likely to prompt greater private provision of resources for retirement.
It has long been recognised that the level of personal wealth is related to age. One obvious reason is that the passage of time provides more opportunity for households to acquire wealth via savings out of income, successful investments, capital appreciation and inheritances. Later in life, wealth may be drawn down in order to maintain a customary level of consumption in the face of lower income, or to make gifts to children and grandchildren. So it is natural to expect wealth to peak at some point after the end of working life. However, the rate at which wealth rises and falls over time, and the age at which wealth peaks, vary considerably across individuals, between countries, and over time. Much depends on the motives for saving or – to be more precise – the strength of desire to acquire, accumulate and retain wealth.
Motives for saving
There are three main reasons for saving. The precautionary motive aims to provide “rainy day” insurance against events like unemployment or ill health. Precautionary saving results in a buffer stock of fairly liquid assets that can be used to survive adverse unanticipated events without taking on excessive debt.
A second reason for accumulating wealth is life-cycle saving, undertaken to smooth consumption in response to anticipated future periods when income may be lower than normal, mainly retirement years, but also possibly child-rearing spells. Consumption smoothing for retirement results in a hump-shaped age-wealth pattern, made familiar by the life-cycle model developed by Franco Modigliani and others a half century ago.
The third motive for saving is to accumulate assets to pass on to one’s heirs. The strength of the bequest motive tends to rise with income, and is very important for some HNWIs; but saving by people at all income levels is often influenced by a desire to leave something for future generations.
The Life-Cycle Model
The link between wealth and age is most prominent in the context of life-cycle saving. To understand the link, imagine a simplified world in which an individual contemplates the future at age 20, expecting to work until age 60 and to die at age 80. Assume further that income will be USD 30,000 per year, desired consumption is constant, the interest rate is zero, and no bequests are planned. Then the individual should spend USD 20,000 per year, resulting in constant annual saving of USD 10,000 while working, and constant annual dissaving of USD 20,000 during retirement. Figure 1 illustrates the tent-shaped age-wealth profile, with peak wealth of USD 400,000 at age 60.
Varying the parameters in this simple model yields clear results. For example, if longevity is fixed and the working life is longer than the period of retirement, a decrease in retirement age increases peak wealth. And if longevity is rising, peak wealth will rise too, as long as neither the working lifetime nor the period of retirement becomes shorter. Overall it is safe to say that peak wealth tends to rise with increases either in the length of retirement, or in longevity, in the life-cycle model.
Another implication of the simple life-cycle model is that the ratio of aggregate wealth to income is positively correlated with the length of retirement and with longevity. If the population is equally distributed by age, mean wealth will equal half of peak wealth. This means that in the world of Figure 5, average wealth is USD 200,000. Averaged over the entire adult lifetime, annual earnings are USD 20,000 per year, so that the wealth to income ratio is 10. Empirical work has confirmed the positive impact of increasing longevity and longer retirement on the wealth-income ratio. The upward trend in wealth-income ratios seen in the G7 countries over the last 25 years may in part reflect these effects.
Intergenerational transfers: From the old to the young
If wealth accumulation is driven by precautionary motives rather than lifecycle considerations, then the level of desired wealth would not be expected to show great variations by age. If both motives are present, it is likely that precautionary needs will be satisfied by wealth accumulated for lifecycle purposes, except perhaps for the youngest and oldest members of society, so the impact on the age-wealth pattern should be minimal.
Bequests and gifts may be intentional or else “accidental” – arising because individuals are usually unable to guarantee leaving precisely zero wealth at the time of death. If bequests are purely accidental and recipients consume them gradually, average wealth will rise for older age groups, and wealth could peak at a higher age. Planned gifts and bequests may reduce saving and wealth holdings in early years by those anticipating receipt of bequests, but increase wealth later on in the lifecycle by those planning bequests. The net result again is higher average household wealth, and a higher peak level of wealth, with the peak possibly occurring later in life.
Intergenerational transfers: From the young to the old
Retirement pensions of all types – state pensions and private pensions of either defined benefit (DB) or defined contribution (DC) form – raise particular difficulties when attempting to confront the life-cycle model with data on household wealth. All types of retirement pensions offer the prospect of future payments in old age, thus tending to substitute for private saving and lower non-pension wealth, although evidence suggests that individuals do not view pension wealth as a perfect substitute for private saving.
The capital value associated with entitlements to DC or DB pensions can be estimated in principle, but the required information is not usually available to those assembling estimates of the wealth holdings of individuals. Most household surveys therefore omit private pensions, but there are important exceptions, for example in Canada, the UK and the USA, as discussed in the next section.
Public pensions raise deeper issues
The capitalised values of the associated income streams can be estimated, but while it may be possible to mobilise private pension wealth through early withdrawals in some cases, or by borrowing against future pension income, this is less likely to be true with public pensions, so it becomes more problematic to regard the pension as part of an individual’s wealth. More importantly, since public pensions are mostly on an unfunded, pay-as-you-go basis, they do not appear as an item in national household balance sheets: the positive wealth values for older persons are exactly offset by the negative values of (unsuspecting) younger individuals when future public pension liabilities are taken into account.
Our wealth estimates in this report try to follow the standard approach to national balance sheet accounting, by including private pensions in household wealth, but not public pensions. For countries with relatively larger public pensions compared to private pensions, this should lead to both a lower hump in the age-wealth profile and a lower ratio of aggregate wealth to income.
The empirical evidence
Empirical evidence on the age-wealth relationship is available for a limited number of countries. This section considers age-wealth relationships in four high income countries (Canada, Italy, the UK and the USA), and two emerging market countries (Chile and Indonesia), based on data from household surveys that provide differing coverage of private pensions. The wealth estimates in this report adjust for these differences, so that the figures fully reflect private pension wealth; but here the results of the raw household surveys are reported without adjustment. The surveys in Canada and the UK cover private pensions fully. The US survey includes DC but not DB pensions. In Italy, Chile and Indonesia, private pensions are completely excluded.
Figure 2 shows the mean wealth of various age groups expressed as a multiple of overall disposable income. Despite the differences in pension coverage, a hump-shaped age-wealth profile is observed in each case, with a peak in two countries at ages 55–64 and in four countries at ages 65–74. Peak wealth is about 60% above overall average wealth in Canada, Chile, the UK and the USA, but only about 30% above the average in Italy and Indonesia.
In Italy and Indonesia the age profile of wealth is flatter than in the other countries. This is likely to be partly a reflection of the omission of pension wealth, but also reflects lower debt among the young in these two countries compared with the others.
The pattern is broadly in line with the predictions of the lifecycle savings model. The early peak observed for Canada and the UK could be due to the fact that the graphs refer to cross-section data for different age groups at a given time, rather than longitudinal data for a cohort of individuals as they age. The decline in wealth above age 65, which averages about 2% per year across the countries shown, seems in line with expectations, but this level of dissaving could simply reflect the lower lifetime earnings of older cohorts in growing economies.
Asset allocation and ageing
Figure 3 shows that the share of financial assets in total assets (excluding debts) varies widely across countries. The high levels in the English-speaking countries are partly explained by the full inclusion of private pensions in Canada and the UK, and partial inclusion in the USA. The low levels observed in Italy, Chile and Indonesia may reflect not only the omission of pension wealth, but also data problems and other factors. Historic episodes of inflation in these countries may have made households shy of financial assets, so real assets may be preferred for consumption smoothing: in Indonesia, for example, gold plays an important role as a store of emergency funds.
In five of the countries, the share of financial assets rises at least up to retirement, but in Indonesia the importance of financial assets declines uniformly with age, perhaps reflecting a transition to greater use of financial assets by younger people for consumption smoothing. The other five countries provide an interesting contrast in what happens after retirement. In Canada and the USA, financial assets continue to increase in importance, perhaps reflecting a downsizing in home ownership. In Chile, Italy and the UK, the importance of financial assets peaks around retirement age and then gradually declines – as expected if retirement savings are being spent and people remain in the same home.
Debt shows more dramatic changes with age than assets (Figure 4). For people aged 25–34 in the English-speaking countries, debts – largely mortgages – are equal to about 40% of total assets. As mortgages are gradually paid off and financial assets are accumulated, the debt ratio declines fairly rapidly, almost vanishing by the age of retirement, except in the USA, where debt still averages 6.4% of total assets at ages 65–74. The debt ratio in Chile is similar when averaged across age groups, but starts at a lower level (25%) for the 25–34 year old group. Italy and Indonesia have much lower overall debt levels than the other countries, and the debts of the young are very modest in comparison to those in the other four countries. In Italy, this pattern is related to the difficulty of obtaining mortgages.
Implications of ageing
The theoretical predictions and empirical evidence on the relationship between wealth and age have important implications for global wealth holdings in the context of the major demographic and social changes likely to take place in the future. Some of the future trends directly affect the age-wealth pattern: this category includes pension changes, longer life expectancy, and – particularly in developing countries – social trends such as increased education, smaller families, and higher female labor force participation. Other future trends may have important social or economic consequences because of age-wealth relationships: the major factor in this second category is population aging.
Changes to public and private pensions
A number of high income countries have raised, or plan to raise, the minimum age for receiving public pensions, which is expected to lead to lower levels of private wealth since it will likely lead to later retirement. An increase in pension contributions will reduce net incomes and reinforce the tendency toward lower saving. Reducing pension benefits, on the other hand, increases the need for private retirement saving and hence raises wealth. Thus different wealth impacts can be predicted for countries that pursue different solutions to public pension deficits.
Regarding private sector employer-based pensions, there has been a significant shift toward defined contribution pensions in high-income countries, which is expected to increase life-cycle saving and wealth. Low current rates of return on most financial assets and increased uncertainty about future income streams resulting from the trend to defined contribution schemes will tend to raise the need to accumulate wealth. Widespread confusion about the expected future benefits from both public and private pensions could, on the other hand, lead some people to abandon their saving efforts if unsure whether their private saving will make any difference.
Changing family values
Reduced fertility and rising female labor force participation – trends seen all around the world, although not in every single country – raise real incomes per head and therefore also tend to increase both life-cycle saving and household wealth. Moreover, the wealth-income ratio will increase further because a family’s relative consumption needs fall in the years when children are at home and shift to the retirement period. If there is also a trend toward retired people living separately, rather than with their offspring, which may occur in developing and emerging countries in the future as it did in high income countries in the past, the need for retirement savings will be increased, again leading to greater wealth.
Most of the above factors tend to increase life-cycle saving and wealth. Rising education is likely to reinforce this trend, since its main effect is to raise lifetime earnings, which will lead to higher saving for retirement.
Savings and wealth in an ageing world
The most important future consequences of the age-wealth relationship are likely to arise because of the global demographic trends expected to lead to population aging in all regions of the world over the next 50 years. Overall, higher age groups are wealthier than average, so population aging increases wealth through a simple compositional effect. There will also be compositional effects on portfolio patterns. The relative demand for financial assets, and especially for risky financial assets, will likely increase since the importance of these assets tends to increase with age. The converse is a relative decline in the demand for housing. Household debt should also decrease in relative terms since older groups are less indebted.
Population aging will have many other effects, of course. It is expected to increase real wages, but reduce returns to capital, with opposing effects on household saving and wealth. It is also expected to increase public expenditures on health care and other services to the elderly, leading to higher taxes on working age people, which may reduce their saving ability. On the other hand, there may be cutbacks in public services for the elderly due to their increased cost, leading to greater need for private preparation for old age. The net effect of all these factors is unclear.
Wealth accumulation occurs for three principal reasons: for precautionary reasons, in anticipation of retirement needs, and to make intergenerational transfers. Saving for life-cycle purposes, in particular, means that age is an important determinant of household wealth levels. Economic theory predicts a “hump shape” for the age-wealth profile, and cross section data confirms this pattern in all countries. Empirical evidence also shows that – expressed as fractions of total assets – debt drops sharply with age, and financial assets tend to rise fairly strongly with age, at least up to retirement.
The life-cycle saving model suggests that, in future, the age-wealth profile will be more peaked, and average wealth will be higher if life expectancy is longer, retirement is earlier, or public pensions are lower. At present, life expectancy continues to increase in many countries and public pensions are tending to become less generous, pointing to a rise in household wealth. On the other hand, retirement ages are trending upward in the developed world, with the opposite effect. So, although it seems likely that wealth/income ratios will rise in developing and emerging market countries in the future, they may not continue to rise everywhere.
In the coming years, the relationship between wealth and age is likely to have greatest significance in the context of global demographic trends leading to population aging. Average wealth and wealth-income ratios will rise as the population weighting shifts to older groups. Aging is also expected to increase the demand for financial assets relative to real assets like housing. More important, perhaps, is the strain that population aging places on public pension systems and other supports for the elderly, which is likely to further strengthen calls for private responsibility in preparing for retirement.
The open question is whether population aging will precipitate a crisis due to insufficient private saving for retirement. While low saving for retirement remains a concern in several countries, and while some people are woefully unprepared for retirement financially, overall household wealth accumulation has increased and wealth/income ratios have risen. Lower estimates of the level of income needed to maintain pre-retirement standards of living, provide additional grounds for optimism that most countries will avoid a fiscal crisis due to an aging population.