Archives for October 2018

A reader writes in, asking:

“Can you please write an article about SPIAs with guarantees of a minimum payout period?”

For those who are unfamiliar, a single premium immediate lifetime annuity (sometimes referred to as a SPIA) is an insurance product where you give the insurance company a lump sum of money (which you cannot get back) and in exchange the insurance company promises to pay you a certain amount of money every month for the rest of your life. In short, it’s a pension that you purchase from an insurance company.

Such annuities are useful because they protect against longevity risk (i.e., the financial risk that comes from living very long and therefore having to pay for a very long retirement).

One thing that stops many people from buying such annuities, however, is the fear that they’ll die soon after purchasing the annuity. For example, if you spend $100,000 on a SPIA that pays you $6,000 per year for the rest of your life, but the rest of your life only turns out to be a couple of years, you will have had a net loss of $88,000.

And that’s why insurance companies offer the option to purchase a “period certain,” whereby the insurance company promises to pay out for at least a given period of time. For example, for a lifetime annuity with a 10-year period certain, the insurance company promises to pay out for the rest of your life but no less than 10 years. (So if you died after two years, the insurance company would continue to make payments for another 8 years to your named beneficiary.)

Of course, because of this guarantee, a lifetime annuity with a period certain will cost more (i.e., will require a higher premium) for a given level of income than you would have to pay for a lifetime annuity without a period certain.

Why a Period Certain Is a Bad Deal

The whole point of insurance is risk pooling. For example, consider 1,000 people who purchase homeowners insurance from a given insurance company. Most of those people will not have their homes destroyed by a fire or a tornado. And that fact — the fact that the insurance company is going to collect money from all of those people without ultimately having to make a huge payout to them — is how the insurance company can afford to make a huge payout to the person whose home is destroyed by a fire.

A key point, however, is that for every dollar that an insurance company receives in premiums, they keep some part of it to cover their administrative costs and to provide profit to shareholders. So only some of the money spent on premiums ultimately goes to pay for claims to people purchasing the insurance product in question. So in general it is unwise to purchase an insurance product unless:

1. There is risk pooling going on (i.e., many people are going to ultimately get a bad deal so that some people can get a very good deal), and
2. You need such risk pooling (i.e., you cannot reasonably afford to cover this risk out of pocket on your own).

With a lifetime annuity, risk pooling occurs because some annuitants will die prior to reaching their life expectancy (i.e., the insurance company will pay less than the “expected” amount to those people — which is how it can afford to pay more than the “expected” amount to the people who live beyond their life expectancy).

But if the insurance company is providing a period certain, it knows it must pay out for that entire period. In other words, the annuity then offers no risk pooling for that period. Instead, what’s occurring for that period is just the insurance company gradually paying your money back to you — after taking a piece off the top for profit and expenses — without any actual net insurance effect.

In most cases you would be better off investing the money yourself for the period certain, then buying the annuity at the end of that period. (Of note: if you would be considering a 10-year period certain, don’t buy 10-year bonds. Instead buy longer-term bonds to offset the interest rate risk that you face with the annuity purchase. See this prior article and this Bogleheads discussion for a more thorough explanation.)

David Blanchett of Morningstar recently released a piece of research discussing the uncertainty of retirement age: The Retirement Mirage (pdf). I linked to it in the most recent roundup, but I wanted to highlight its findings, as I know that any single article in a roundup can be easy to miss.

Blanchett looked at 12 years of data from the University of Michigan’s Health and Retirement Study (HRS). The HRS is interesting because it tracks a large group of people (approximately 20,000) over a period of time, so you can see how people’s circumstances and views change over time.

Blanchett learned two interesting things from the HRS data.

First, he found that you’re likely to retire closer to age 61 than you think. In Blanchett’s words:

“According to the Health and Retirement Study data, planned and actual retirement ages align at 61, with those planning to retire earlier than that tending to retire later than expected, and those planning to retire after 61 tending to retire earlier than expected. In other words, actual retirement ages pull toward 61, with each retirement year planned before or after age 61 resulting in a half-year’s difference in actual retirement age. For example, someone who plans to retire at age 69 will likely retire at age 65 (69 – 61 = 8 × 0.5 = 4; 69 – 4 = 65)”

Of those planning an early retirement, many are ultimately unable to retire as early as planned.

And of those planning on working well into their 60s, many are unable to do so for one reason or another. (Alternatively, some people are probably retiring earlier than anticipated because they’re finding that they do not need to work as far into their 60s as planned.)

So, what might help us to predict who will be in the “retiring earlier than planned” or “retiring later than planned” groups? That leads us to Blanchett’s second noteworthy finding:

“The rich HRS data set allowed us to test more than a dozen factors, including general personal characteristics such as gender, marital status, and education, along with factors that might be expected to lead to retiring early, such as job stress level, how physical a job is, and whether health problems limit someone’s work. However, these factors had little or no predictive power on retiring early. The only factor that appeared to tell us much about when someone might retire was their planned retirement age and its distance from the previously noted ‘magic’ retirement age of 61. […] Not only do many people retire earlier than expected, but it’s nearly impossible to predict who will be part of this group.”

This obviously presents some challenges as far as retirement planning. But it also suggests a few financial/life strategies that are likely to be worthwhile.

If you’re planning an early retirement, keep your professional skills sharp, as there’s a good chance you’ll be working longer than you anticipate. Also, if you’re currently in the position of “grinding it out” at a job that you hate, with the hope of being happy once you achieve an early retirement, you may want to consider a different approach. “Just 2 more years” could well turn out to be 3, 4, or 5 more years. Focusing instead on making changes that allow you to be happy while still working is likely to be a good idea.

At the other end of the spectrum, if for example you are currently age 60 and planning to work until 68, socking away that last chunk of necessary retirement savings may be more urgent than you think. Retirement may ultimately be something that happens to you, rather than a decision you make.