Retirement planning is based on a significant number of assumptions. And from market movements to investor preferences to individual life spans, uncertainty reigns.
It's the job of financial advisers and other investment professionals to bring some order to the chaos. We can't predict the market, but we can help ensure that our investments have the right mix of forecast risk and return. We can measure investors' risk preferences, even as they change over time. And we can shed some light on when a person might be expected to die. It's a delicate matter, but one with major consequences on retirement planning, one in which we need to be as accurate as possible.
(More: Finding solutions to the key challenges of modern retirement)
Here are a few guidelines when estimating a life span for yourself or a client.
First, life span is personal, but the estimation tools are not. Second, conservatism rules: It's far better to die with money in the bank than to live one's later years in poverty. And third, probability is the antidote for uncertainty.
Let's take those in order.
Generic tools have value. We know from actuarial tables, for example, that people are living longer on average. For example, in 1980, the life expectancy for a 65-year-old man was
14 years; it's now closer to
18 years, according to
Social Security Administration data. That means the duration of retirement is increasing.
Also, older people are dying later and fewer people are dying early — both of which raise the average life span. We expect a baby born today to live an average of 79 years, while the average 65-year-old man is expected to live to 83.
PERSONAL DETAILS
The problem, of course, is that very few people are average. If your father died at 50 and your mother lived to be 100, should you expect to live to be 75?
When estimating an individual's life span, it is important to consider personal details, such as family longevity, medical history, relevant habits (smoking, diet, exercise) and environment (stress, marital status, etc.). This is inexact, but if all these factors point to a longer life, this should inform your estimate.
It's also useful to point out that richer people tend to live longer. While not every
financial planning client is a “one percenter,”
recent research suggests that wealthier individuals as a group are generally well above average from a longevity perspective. This life expectancy gap for wealthier people is widening, particularly among women.
All of that argues for taking a more conservative approach to estimating life expectancy and the length of retirement.
In my research, I gravitate toward annuity mortality tables, which usually are more conservative than regular mortality tables because of adverse selection. People who buy annuities are likely to be healthier than average and therefore live longer. A mortality table I've been using recently is the
2012 Individual Annuity Mortality created by the Society of Actuaries.
PROBABILITIES RULE
These tables help make our estimates more precise, as they provide more accurate longevity probabilities than a traditional mortality table, such as the Social Security Administration mortality table.
(More: Declining marriage rates present a planning puzzle for long-term couples)
For example, the odds that either member of a couple, male and female 65-year-olds, will live past 95 is only about 20%, which makes a 30-year retirement period seem reasonable. Based on the annuity mortality table, however, this probability increases to about 43% when both spouses are 65 today, and is projected to rise to over 50% in about 15 years.
In other words, there's a 50/50 chance that either member of a couple of 50-year-olds planning to retire in 15 years will live more than 30 years after retirement. Therefore, the safety of a 30-year retirement (or assuming death at 95) varies significantly based on your assumption.
There are no easy answers when predicting retirement periods. But make it personal, stay conservative and lean on probability.
David Blanchett is head of retirement research at Morningstar Investment Management.
