Updating the Annuity Puzzle

Economists and many financial advisers believe that most people should annuitize a portion of their nest eggs. That is, once they enter retirement they should buy an immediate annuity paying a fixed amount annually no matter how long they live. (Or you can buy an inflation-indexed annuity so the annual payment increases with inflation, but still is guaranteed for life.) By doing so you shift the risk of a long life and low investment returns to the insurer, who is better able to handle it. Relatively few people actually do this. Economists refer to this as The Annuity Puzzle, because they don’t understand why few people take an action that seems to logical to them. Economists also offer a number of reasons why people might not buy annuities. The most common reasons are that people don’t want to give up control of part of their nest eggs and don’t want to give up the potential of higher returns.

A recent paper reaches a different conclusion. Two economists did their own calculations and concluded that only a minority of people would benefit from annuitizing part of their nest eggs. Their point is that traditional analysis focuses only on the potential for a long life and outliving your nest egg. But many people have the opposite problem. They might have health problems and need extra cash to pay for care. Or they might die before life expectancy, in which case the insurance company benefits from their money.

I’m going to study this more, and I’m sure others also will analyze and comment on it. But it’s an interesting idea and deserves consideration. Perhaps the experts have been wrong all along and the intuition of regular people makes more sense. It wouldn’t be the first time.

Jason Scott, head of Financial Engines FNGN -0.74%’ Retirement Research Center, neatly sums up the new research this way: “In the simpler (old)  model, when you get hit by the bus you die and you don’t have any demand for money. This new model says you get hit by the bus and you can’t work and you have a greater demand for money.”

In an interview, Smetters emphasized that the paper is still a work in progress and that it doesn’t suggest behavioral theories about why people don’t buy annuities are wrong— just that purchasing annuities may make economic sense for a whole lot fewer people than economists previously believed. The paper calculates that at age 65, when loss of income from work is no longer a factor, some (but not total) annuitization makes sense for 27% of households after fees and desired wealth transfers are taken into account. But those households who should annuitize tend to be the wealthier ones —rich enough to generate an annuity stream that can cover extended long-term care, which isn’t paid for by Medicare.