![Sample our Health and Social Care journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5938/TOC-Subject-Sample-2021-Health-Social-Care.jpg"})
Abstract
Financial markets have been characterized by boom and bust cycles since the 1980s, while the responsibility for managing retirement wealth has increasingly shifted onto individual households at the same time. Policy makers and experts have expressed concern over rising risk exposure among older householders, who appear to be increasingly exposed to the growing financial risks just as they near retirement. We consider household data from the Federal Reserve’s Survey of Consumer Finances from 1989 to 2010 to analyze the correlation between age and risk exposure. We test whether older householders’ risk exposure has indeed grown over time, whether it has increased more than that of younger householders, whether changes in the demographic composition of older householders have contributed to older households’ rising risk exposure, and the degree to which increases in risk exposure can be traced to a growing concentration of household assets held in stocks and housing and to rising householder indebtedness. Our results indicate that risk exposure has grown more for older householders than for younger ones, that demographic changes among older householders have contributed to additional increases in older householders’ risk exposure, and that the growth of older householders’ risk exposure is driven more by rising risky asset concentration and less by greater indebtedness.
Notes
1. This assumes that DB pensions expose households to less market risk than individual savings do, since DB pensions can smooth asset market fluctuations over time.
2. See Weller & Bernardo (Citation2014) for a detailed discussion of housing as a risky asset.
3. See Weller & Bernardo (Citation2014) for a more detailed discussion of financial risk exposure.
4. Some studies examining the trends between risk tolerance and age rely on self-reported risk tolerance measures (Grable & Lytton, Citation1998; Sung & Hanna, Citation1996; Yao, Gutter, & Hanna, Citation2005; Yao et al., Citation2011). We also include a self-reported measure of risk tolerance in our model, but we are primarily interested in the relationship between age and risk exposure. We thus focus on risky asset concentration and leverage as indicators of total risk tolerance. See Riley & Chow (Citation1992) for further discussion on the merits of using self-reported risk tolerance and risky asset concentration as household risk exposure measures.
5. Stocks include all directly held stocks and indirectly held stocks, e.g., in 401(k) plans, in individual retirement accounts, and in other managed accounts. The data do not allow for a further disaggregation of stocks, but this is of limited concern since diversification between asset classes not within asset classes largely determines risk exposure, as discussed earlier. Houses include the self-reported gross value of all residential real estate; that is, it is not the net of mortgages since we measure leverage separately.
6. This ratio implies relative risk aversion because we define risk exposure relative to total assets rather than absolute dollar amounts. We select these particular cutoffs to ensure reasonably large sample sizes for our analysis. The conclusions of this discussion remain robust with changing cutoff points for very high risk exposure.
7. See Weller & Bernardo (Citation2014) for detailed discussions of summary data.
8. We separately estimate our regression equation pooling all observations, adding a dummy for later years and interaction terms between the dummy for later years with age and age squared. These results are shown in Weller & Bernardo (Citation2014). The correlation between age and risk exposure do not change materially in this specification.