Bold Forecasts but Timid Choices

Therefore, when it comes time for us to act upon the facts and data we have gathered and the analysis we have undertaken about them, these biases — unjustified optimism and unreasonable risk aversion — conflict. As a consequence, we tend to make bold forecasts but timid choices.

Dealing with these conflicting biases is never easy. But we can start by recognizing the nature and extent of the problem.

These biases go a long way toward explaining what academics call the "annuity puzzle" — the idea that people buy income annuities to guarantee retirement income far less often than good sense says they should. For example, we tend to overestimate the likelihood of good future portfolio returns while underestimating the likelihood of major portfolio drawdowns such that we take no action to deal with longevity risk by purchasing an income annuity. In short, we tend to accept risks when we don't think we will have to bear them.

This reluctance to take clear and explicit responsibility for potential losses is both costly and powerful. As a result, as shown by political scientist Aaron Wildavsky in his famous book Searching for Safety, we are typically better off engaging in an active search for safety rather than relying upon a passive prevention program. An appropriately active response to longevity risk is to provide a guaranteed income source so that retirees can rest assured that they will not run out of money.

One way to overcome these biases is by framing the decision to be made differently. The "certainty effect" is a sharp discrepancy between the weights that are attached to sure gains and to highly probable gains in the evaluation of prospects. Thus, when an income annuity is positioned as a certain gain of guaranteed income for life rather than as a payment resulting in a loss of control of one's nest egg in exchange for future payments, it is more likely to be well-received.

We also tend to consider problems one at a time in isolation instead of aggregating the problems and the risks and looking at them together. If the benefits of an income annuity are examined in conjunction with longevity risk, market risk, sequence risk and health risk, it becomes a much more attractive proposition. As Kahneman and Lovallo affirm, "evaluations of single risky prospects neglect the possibilities of pooling risks and are therefore overly timid."

Similarly, we will often reject a single gamble but will accept multiple plays of that same gamble (for example, a gamble where we will either gain $200 or lose $100). Thus an income annuity framed as a one-time purchase pitting a fixed result against unlimited future possibilities is unlikely to be seen as desirable. Instead, prospect theory recommends broadening the decision frame. Outcomes should be evaluated in terms of final assets rather than gains and losses associated with each choice. For example, research by John Ameriks, Robert Veres, and Mark Warshawsky shows how income annuities need not be an all-or-nothing proposition (ideally they should not be). The authors do not try to find an optimal annuity amount, but they do show how annuitizing a portion of one's savings can help improve success rates for a 4.5 percent withdrawal rate.

The "annuity puzzle" remains a vexing problem for advisors who are intent upon the best interests of their clients. But if dealt with carefully and communicated fairly and well, the value of guaranteed income can be demonstrated to many consumers.