Investing
In Brief: Risk Preferences And The Investment Decisions Of Older Americans
Research Report
Vickie L. Bajtelsmit, Colorado State University
Alexandra Bernasek, Colorado State University
June 2001
Learn More
- Report Home
- Full Report (PDF)
- In Brief (HTML)
Given the trend in private and public pensions to allow more participant choice in retirement saving asset allocation, how much risk individuals take when they invest their assets has important implications for retirement income security. The “three-legged stool” of retirement income policy assumes that, in addition to relying on future Social Security benefits, individuals and households will participate in private pensions and will save on their own to provide an adequate retirement standard of living.
Examination of individual investment portfolios reveals that for many households, accumulated wealth may be an insufficient component of future retirement income relative to their needs. Although part of the problem is that households simply have lower savings rates than previous generations, also of concern are the high levels of personal debt, the high proportion of household portfolios invested in housing, and the fact that many households are risk averse, i.e., they tend to make substantial investment in low-risk assets with low returns. Since riskier portfolios generally result in higher returns for long-term investors, overly conservative investment behavior may result in inadequate retirement wealth. The role of risk aversion, i.e., attitudes toward risk, in the portfolio allocation and investment decisions of individuals and households – is the focus of a study by Vickie Bajtelsmit and Alexandra Bernasek of Colorado State University, entitled “Risk Preferences and the Investment Decisions of Older Americans.”
The researchers use the 1994 Health and Retirement Study (HRS)1, to estimate risk preferences and evaluate the factors that impact asset allocation in individual and household portfolios for a sample of older Americans. Of particular interest are differences in the willingness of women and men to make risky investment allocation decisions. They also examine other important factors that impact investment decisions including age, race, wealth, marital status, and education.
The first part of the report reviews the economics of risk aversion and the relevant literature on investment allocation. Most studies have concluded that individuals, when left to their own devices, tend to invest much more conservatively than professional asset managers would recommend. The implication of this is that, on average, individual portfolios will have lower return on investment and consequently will produce lower retirement accumulations than would otherwise be the case. Studies have fairly consistently found that women are more conservative investors than men, older people have lower-risk portfolios than younger people, and higher-income people are less risk averse than lower-income people.
In the second part of the report, relative risk aversion is estimated using the 1994 wave of the HRS. Examination of the household portfolios of HRS respondents reveals that low levels of savings and high levels of personal and real estate debt are serious problems for many households in this age category. The net effect is that many households have relatively little wealth to rely on for retirement income. Fifty percent of all households have less than $200,000 in net wealth. The lowest 25 percent have less than $77,000 in net wealth. This distribution of wealth persists across all the age groups with those at or near retirement age having only slightly more wealth on average than those in their 50s. The study measures net present wealth only and excludes the value of future inheritances, defined benefit plan benefits, and Social Security benefits.
The determinants of portfolio allocation to risky assets are examined with and without including housing wealth, since housing is purchased for both consumption and investment. Whether or not housing is included, net wealth is found to be a significant determinant of the willingness of households to take risk in investing their portfolio. The results of this research are strongly consistent with the hypothesis that the wealthier the household, the greater is their portfolio allocation to risky assets, even after controlling for age, income, dependents, and other demographic characteristics. The percentage allocation to risky assets is found to be lower for those with less education and for those with higher defined contribution pension balances. When housing is not included in the definitions of wealth and risky assets, blacks have lower-risk portfolios and single women have higher-risk portfolios. When housing is included in both the definitions of wealth and risky assets, blacks tend to have a larger allocation to risky assets – an indication they allocate more of wealth to housing than non-blacks.
Marital status and gender differences are investigated by comparing portfolio allocations of single women to single men and married couple households, not including housing equity in the definition of wealth. Although all three groups are found to invest more in riskier assets the more wealth they have, single women are found to be relatively less risk averse than married couples and single men because they allocate proportionally more to risky assets than men or couples. Single men and women tend to have less risky portfolios when they are homeowners and more risky portfolios as the number of children increases. Unlike the results for the entire sample, having lower education and being black is not a significant factor in single women's risky allocation.
These findings raise policy questions about how to improve the retirement income prospects for men, women, and minorities. Encouraging saving and increasing wealth among segments of the population with especially low saving rates is an important retirement income policy concern in itself. However, increasing wealth may lead to greater investment in risky assets, which can lead to higher long-term investment returns. The findings also raise the question whether some Social Security reform proposals that involve individual accounts will produce sufficient returns to risk averse low-income households to offset reductions in benefits.
Top of PageFootnote
1The HRS is a nationally representative panel study of 6,979 households nearing retirement (respondents age 51 to 61 in 1992 and their spouses). This survey includes detailed information on financial, health, experiences, and attitudes.
Written by Vickie L. Bajtelsmit and Alexandra Bernasek,
Jules H. Lichtenstein, Project Manager, AARP Public Policy
Institute
June 2001
©2001 AARP
May be copied only for noncommercial purposes and with
attribution; permission required for all other purposes.
Public Policy Institute, AARP, 601 E Street, NW, Washington, DC
20049
Pub ID: INB44