Where Clients Go Wrong in Predicting Life Expectancy

The length of retirement is one of the most important assumptions in a financial plan. Despite its uncertainty, care must be taken to ensure mortality assumptions are reasonable because errors in estimates can significantly impact the plan.

Recently, I explored the efficacy of subjective mortality estimates, or in less fancy terms, how accurate people are at predicting their life expectancy, in research published in the Journal of Retirement.

While there is lots of research on objective mortality factors (i.e., the actual drivers of life expectancy), such as smoking, there isn't as much on subjective estimates.

The analysis primarily leverages data from the Health and Retirement Study (HRS), in particular a question in the original survey (in 1992) that asked respondents "What do you think are the chances that you will live to be 75 or more?" With an average respondent age of 58, it is possible to observe the accuracy of these estimates with respect to whether the respondent actually survived to age 75.

A benefit of using the HRS is that it tracks the same households over time and includes other information that is useful when exploring mortality-related topics, like health, smoking status, income, education, etc. The dataset includes 5,499 primary respondents, which is a good sample size.

Notable Gaps

This analysis suggests that while individuals appear to have some sense about their likelihood of survival (i.e., their subjective mortality), there are notable gaps in these estimates that are consistent with past research.

For example, respondents in the first wave of the HSA who said they had a 0% probability of surviving to age 75 actually had about a 50% chance, and those who said they had a 100% probability actually had about an 80% chance.

Objective factors appear to influence subjective mortality estimates differently. For example, individuals do a relatively good job incorporating their health status into projections, however, they do a relatively poor job incorporating things like smoking and income (and wealth). In other words, while smoking is negatively related to life expectancy, people who smoke don't fully take that into account.

Income Impact

An especially important consideration for advisors is the growing gap in life expectancy estimates based on income, where households with higher levels of income are living significantly longer. Because financial planners tend to work with households in the highest two income deciles, their clients need to be aware that their retirement periods are likely to be longer than the average American and they need to plan accordingly.

When it comes to financial planning assumptions, a model based on a few key inputs, especially including self-assessed health status, is likely to be much more accurate than asking a client a mortality-focused question.

These estimates can be improved with other basic (easily available) information, such as age and gender, as well as other factors, like income and smoker status.

There are lots of tools out there that can be used to estimate a retirement period. I'm a fan of the Longevity Illustrator from the American Academy of Actuaries and the Society of Actuaries.

When determining the assumption about the length of a retirement for a financial plan, don't ask individuals how long they think they will live; instead ask more objective questions to drive the assumption. Of course, there likely will be gaps in the resulting value and client estimates, which will be important to address when developing the optimal strategy.

David Blanchett is managing director and head of retirement research at PGIM DC Solutions. PGIM is the $1.5 trillion global investment management business of Prudential Financial. He can be reached at [email protected].