Assessing Default Investment Strategies in Defined Contribution Pension Plans

However, the specific allocation to risky assets (such as equities) at different ages is a matter of
much debate, both in academic and policy circles. There is also an on-going debate on the relative merits of deterministic investment strategies with a fixed glide-path over the life-cycle and dynamic strategies that take into account some supposedly long-term features of asset returns, such as mean-reversion.

The main goal of this paper is therefore to assess the relative performance of different investment strategies. This is also done for different structures of the payout phase. In particular, it looks at whether the specific glide-path of life-cycle investment strategies and the introduction of dynamic features in the design of default investment strategies affect significantly retirement income outcomes.

The paper combines a stochastic analysis of the performance of different investment strategies for
different payout options with a historical analysis to test the findings of the stochastic simulation with actual market data from Japan and the United States. The stochastic model using simulations of returns of the different asset classes (cash, bonds and equities) generates, depending on the form of the payout phase, stochastic simulations of income at retirement. In the historical analysis, all the strategies examined have the same average allocation to equities during the accumulation period. Additionally, two contribution periods are considered: 40 and 20 years.

The main conclusions and policy recommendations of this analysis are as follows:
• The relative performance of investment strategies depends on the type of benefits during the payout phase. Life cycle-strategies do best when benefits are paid as life annuities and are less
valuable when benefits are paid as programmed withdrawals. Dynamic strategies seem to work
better with programmed withdrawals. A mixed of life-cycle and dynamic strategies may be required when benefits are paid combining programmed withdrawals and deferred life annuities
bought at the time of retirement.

• There is no “one-size-fits-all” default investment strategy. None of the investment strategies
dominate in all simulations. Some, such as those with low exposure to equities (less than 10%)
and those with overly high exposure to equities (more than 80%), generally proved inefficient.
• Life cycle strategies that maintain a constant exposure to risky assets during most of the
accumulation period, switching swiftly to bonds in the last decade before retirement seem to
produce adequate results. They provide higher expected benefits for a given level of risk than
other life cycle strategies.

• The introduction of dynamic management strategies can provide somewhat higher replacement
rates for a given level of risk than the more deterministic strategies, at least in the case of payouts
in the form of variable withdrawals.

• The length of the contribution period affects the ranking of the different investment strategies.
Life-cycle strategies perform better than fixed portfolio strategies when the contribution period
is short, for example 20 years. Longer contribution periods reduce the benefit impact of lifecycle
strategies.