6 objections to the "myths" of DIAs

I’d like to start a movement to banish the term “life expectancy” from the retirement planning lexicon. It is one of the most abused and dangerously misunderstood concepts in retirement planning.  And this misunderstanding is holding us back. It’s holding us back from providing the middle-market consumer with real solutions to a serious problem – outliving one’s income in retirement. We need to do something to counter these misconceptions.

TREASURY RULING ON LONGEVITY ANNUITIES

As you may have heard, the Treasury has recently issued a ruling allowing longevity annuities, a.k.a. deferred income annuities (DIAs) to be sold as part of a 401K or IRA plan with favorable tax consequences. Simply put, a DIA allows you to purchase a guaranteed lifetime income stream now (say at age 65), that starts sometime in the distant future (say at age 85). Not to be confused with a typical deferred annuity, there is no cash value with a DIA (you can add a death benefit, but that’s not its “purest” form). It works exactly like a single premium immediate annuity (SPIA) except payments start further out into the future. This allows the DIA to provide a classic cost-effective insurance solution by deploying a modest percentage of your assets in a vehicle that spreads the risk of living a long time.  

BUT “FINANCIAL EXPERTS” THROW COLD WATER

This is an exciting opportunity to apply new solutions to a serious problem. Yet many “financial experts” are trying to throw cold water on the party. Here is just one recent example at Forbes.com. Now, if you review this author’s body of work, he seems biased against anything with the term “annuity” or “insurance company” in it, so we’d likely need to banish a lot more things to satisfy him. But in any case, his concerns are full of misconceptions. I will address each one later in this article, but want to focus first on what I think is the most important and common misconception. That’s the one surrounding “life expectancy”.

The argument goes something like this. DIAs pay you nothing if you die before the start date (say, age 85). And your life expectancy is around age 85. So you are “likely” to die before receiving any benefits, and so the evil insurance company is “likely” going to keep all of your money. The implication is that your life expectancy is some kind of virtual upper bound on your life, beyond which all but a few fortunate souls survive. That’s the misconception. This is not true or even close to what life expectancy means.

SO WHAT DOES LIFE EXPECTANCY MEAN AND HOW IS IT MISUNDERSTOOD?

As an actuary, I take some portion of the collective responsibility for the term life expectancy as it is somewhat of an actuarial concept. So bear with me as I use a little math (details relegated to this linked spreadsheet for those who want to check my work) to demonstrate the misconception.  

Let’s say you have 100 reasonably healthy age 65 male and female couples for clients. I’ve used the most recent Society of Actuaries individual annuitant mortality experience table to calculate their life expectancy. (You could justify other assumptions, but the story does not change that much for our purposes.) Life expectancy is the expected average age at death. So the spreadsheet simply calculates that by projecting out how many are expected to die each year and taking an average. For your 100 clients, the male life expectancy is 86 and the female life expectancy is 88.

But those numbers really don’t tell you anything useful. What you really want to know is how many of your 100 clients would be expected to live a long life? Under our assumptions, 56 out of the 100 males will live beyond age 85, 34 will live beyond age 90, and 13 beyond age 95. For the females, it’s 64 out of the 100 living beyond age 85, 42 living beyond age 90, and 21 living beyond age 95.

But that’s not even the whole picture. If you are advising these clients, you would be talking about providing income so long as either are alive. So you would want to use a joint and survivor DIA. Under those same life expectancy assumptions, you could expect that 84 out of 100 couples will have at least one person still alive at age 85. You’d expect that 62 would have one person alive beyond age 90 and 31 beyond age 95. This is all based upon actual industry experience for individual annuitants. It does not even assume any ongoing improvement in longevity, which shows little if any signs of slowing down.

As you can see, the life expectancy numbers of 86 and 88 really provide no useful information when considering how long your clients are likely to need income. It is in no way true to imply that the DIA is some kind of insurance company rip-off where you are likely to not receive any benefits. A DIA will provide a very good return for well over half of your clients, and precisely for all of those who will actually need it, because that’s what insurance is supposed to do. SO HOW DOES THIS MISCONCEPTION SHOW UP IN RETIREMENT PLANNING?

You typically see a few different approaches to retirement planning.

(1) Hope For the Best scenario (where “best” means you die young): Some advisors plan income spending to ensure assets last to your life expectancy age. We have shown that this is a disaster for 84 of your 100 client couples. Can you possibly feel like you’ve done a good job putting together a retirement plan that only works for 16 percent of your clients? And it only worked for them because they died young, and thus your plan did not actually protect anyone from outliving their assets?

To be fair, many advisors do realize that there is an income need beyond life expectancy. But they argue that clients’ retirement assets are so inadequate, that it is too overwhelming to discuss how to stretch those assets, so they resign themselves to “at least” provide income to their life expectancy age. But that makes no sense at all, as it just ignores the problem instead of solving it. It reminds me of the old joke about the drunk who lost his wallet in the dark alley. When asked why he is looking for his wallet under the street light, he says the light is better there.

That approach is not helping anybody find the keys to a retirement solution. The sad part is that solving the real problem costs only a little more than the one they are solving.

(2) Live Off the Interest scenario:  Finally, many advisors do realize that they need to provide a plan that will last for life. So they recommend a plan where you “live off the interest”. They claim it is safe to withdraw 4 percent to 5 percent of your retirement savings each year and that, over the long haul, interest will be sufficient to cover that. This plan has several problems. First, is it just me or does it seem odd that you save all this money for retirement and then are told by your advisor to never actually spend the money you saved for retirement? Second, it is inefficient to leave all your money on the table when you need income. $500K in retirement savings will only generate $20 to $25K in annual income under this approach. That’s not much and how many middle-market couples have saved as much as 500K? Third, this plan might not even work. In our current low interest environment, it is pretty hard to generate 4 percent to 5 percent investment income. So presumably, the plan includes the purchase of stocks, mutual funds, etc. Yet, it only takes two years of minus 30 percent returns to possibly make the plan fall apart. It happens. My spreadsheet shows just one such feasible scenario. Finally, a 4 to 5 percent withdrawal scenario likely runs afoul of Required Minimum Distribution (RMD) requirements assuming your money is in a tax qualified vehicle. So it can be inefficient from a tax standpoint, in that you’ll need to pay taxes earlier on more money than you are allowed to spend.