Introduction
While Social Security provides those ages 62 and older with a predictable stream of income, most households need other resources as well for a secure retirement. The bulk of these other resources come from employer-sponsored retirement plans, although more affluent households may save additional amounts on their own. With the shift from traditional defined benefit (DB) plans – where employers make the contributions and bear the risk – to 401(k)-type plans – where households are responsible – market risk has become a major concern for many households. The determinants of households’ market risk exposure – i.e., their asset allocations – have been studied extensively. But, insights gleaned by studying actual asset allocations may not reflect retirement investors’ true preferences due to the public’s general inertia when managing their money, the minor hassles involved in signing up for a plan and choosing investments, and defaults built into the retirement system, such as target date funds (TDFs). Indeed, research has increasingly recognized the impact of these factors on observed allocations, and very little is known about what asset allocations might be if retirement investors were unencumbered by these influences. In other words, how much do retirement investors’ desired asset allocation differ from their actual allocations?
This paper reports the results from a new survey on how retirement investors ages 48-78 perceive market risk and its impact on their desired allocation. The analysis compares the desired stock holdings reported in the new survey to actual holdings reported in two major household surveys, and explores the relative importance of individual characteristics versus institutional arrangements – namely, the target date funds that are often the default investment option in 401(k) plans.
The paper proceeds as follows. The first section briefly describes the population for whom market risk is important, illustrates the role of market risk in their wealth accumulation, and summarizes the literature on household portfolio choice in the context of retirement. The second section describes the main data sources for the analysis: the new retirement investor survey, the Health and Retirement Study (HRS), and the Survey of Consumer Finances (SCF). The third section describes the methodology by which desired and actual allocations are compared. The fifth section presents the results, documenting and exploring the difference between households’ desired and actual holdings of risky assets.
The final section concludes that – on average – retirement investors’ desired allocation to risky assets tends to be lower than their actual allocation. This result is likely due to desired allocations that reflect overly pessimistic expectations for equity returns and generally conservative risk preferences compared to actual allocations that are often driven by target date fund defaults based on historical returns and average risk preferences. So, although many retirement investors may be holding more equities than they want due to defaults in the retirement system, to the extent that the defaults correct for investors’ misperceptions of equity returns, it is probably a good thing.
