Assessing the Labour, Financial and Demographic Risks to Retirement Income from Defined-Contribution Pensions

This article relies on a stochastic model that takes into account the uncertainty inherent in returns on investment, inflation, discount rates, life expectancy, employment prospects and real wages, coupled with their correlations, in order to examine the extent of uncertainty of retirement income from DC plans. The main conclusions drawn from examining these risks when assessing retirement income can be summarised as follows:

• The impact on retirement income of varying employment and real-wage career paths, as well as uncertainty about investment returns, inflation, discount rates, and life expectancy is far from negligible. The risk of a shortfall in retirement income is well above 50%, with replacement rates quite dispersed based on any target or median replacement rate considered. Moreover, replacement rates can be quite low when considering worst-case scenarios.

• Labour-market risk, as well as uncertainty about returns on investment and inflation, have the largest impact on retirement income from DC pension plans.

• When assessing the impact of labour-market risk, there is a need to use an absolute standard for the amount of retirement income workers are entitled to receive, in addition to the replacement rate. The replacement rate may be misleading, since individuals who suffer spells of unemployment may have higher replacement rates than those with uninterrupted careers, even though the absolute pension benefits of the former may be lower. The same may happen when there is a shift to different real-wage career paths.

• Regulators and policy makers should seriously consider implementing life-cycle strategies, at least as defaults. Life-cycle investment strategies that reduce exposure to risky assets in the decade before retirement are quite helpful in reducing the risk of sharp reductions in retirement income, in particular when a negative shock to equity markets occurs in the years just prior to retirement.

• The length of the contribution period also matters. The shorter the contribution period, the stronger is the positive effect of life-cycle strategies on retirement income in the event of negative shocks.

• Employment policies may need to focus on younger workers, as workers who suffer spells of unemployment early in their careers will have lower pension benefits than those who suffer otherwise similar bouts of unemployment late in their careers.

If stagflation or deflation were to occur, policies may also need to focus on older workers as either scenario will have a bigger impact on people reaching retirement age within the next decade than on people who are just beginning their careers. Both scenarios are characterised by lower growth, which leads to higher unemployment and lower returns on equities, but the portfolio-size effect dominates.

Based on the analysis, the main recommendation for regulators and policy makers is:

• First establish a target replacement rate for DC pension plans taking into account the overall structure of the pension system;

• Then, set contributions and the length of the contribution period accordingly, keeping in mind that to achieve adequate retirement income people need to “contribute and contribute for long periods, and that it is preferable to increase contribution periods by postponing retirement;

• And, afterwards, focus on asset-allocation strategies, and establish default life-cycle investment strategies that reduce exposure to equities in the last decade before retirement. This is particularly important if contribution periods are short or intermittent, or when a main policy issue is how to address sharply lower replacement rates for those near retirement due to a negative shock to equity markets.