Moshe Milevsky has popularized a concept he calls Product Allocation in his book Pensionize Your Nest Egg. No, this is not what your local whole life insurance salesman is trying to do when he suggests you invest a big chunk of your savings into cash value life insurance. This is what a retiree may wish to do in order to deal with the unique risks a retiree faces.

According to Milevsky, the three big risks a retiree faces are:

  1. Inflation Risk
  2. Longevity Risk
  3. Sequence of Returns Risk

According to Milevsky, each of these risks has a solution, which can be incorporated into a portfolio using a specific product. The first product is a Systematic Withdrawal Plan (SWP.) This is the classic “safe withdrawal rate” style retirement plan where you invest in a stock and bond portfolio (hopefully using low-cost index funds) and take out ~ 4% a year of your portfolio each year to spend. There are lots of variations, of course, but the point of these SWPs is that your money will keep growing allowing you to keep up with inflation throughout your retirement. This portion of your portfolio also provides for any legacy you wish to leave behind, as well as any large, unforeseen expenses. But mainly, it is to combat inflation.

The second product is a pension. This may be Social Security, an employer provided pension, or a purchased pension, such as a Single Premium Immediate Annuity (SPIA.) A SPIA is a lump sum of money given to an insurance company in exchange for a monthly payment every month for as long as you live. Since all of these pensions will pay until the day you die, they protect against longevity. Sometimes these pensions have an inflation-adjustment, and so also help combat inflation along with longevity. Pensions solve many retirement income issues (including declining cognitive abilities) with the main downside being that there is no lump sum of money for you or your heirs to access.

The third product is a variable annuity (VA) with a living benefit rider. Although I have a healthy skepticism of these complex, often overly expensive products, Milevsky considers them a solution for Sequence of Returns risk. Due to their high expenses, they will underperform a traditional portfolio when economic times are good. However, when times are bad, you can activate the rider and basically turn it into a pension. The idea here is that if sequence of returns risk rears its ugly head in your first few years of retirement, the bear market basically “pensionizes” the VA thanks to the living benefit rider. If it doesn't, then you get market returns minus the cost of the VA and its rider. This essentially moderates what you would have gotten from a traditional portfolio- not quite as good in the best case scenario, but not quite as bad as in the worst case scenario.

Milevsky is so fond of these VAs that in his mathematical model (which relies heavily on assumptions which may or may not be valid) he suggests a retiree allocate 1/3 of his portfolio to each of these products.

One other reason he advocates for these VAs is that it allows you to take more equity risk, and he suggests you take it in the VA. For example, he might say that instead of a 50/50 portfolio, you could use a 60/40 portfolio if you had a VA to moderate the risk. He might also say you fill up the VA preferentially with equities, that way if sequence of returns risk shows up, your investment portfolio isn't hurt as much (since it is mostly fixed income) while the VA becomes a pension.

I'm not entirely convinced the VA is an essential part of this equation. VA subaccounts often have hefty expense ratios of 1-2%. Add on the costs of the VA (which at-cost Vanguard says are 0.19% in its VAs) plus another 0.75% a year for the living benefit rider, and now we're underperforming a standard portfolio by 2%+ a year. 30 years of 5% growth instead of 7% growth on a lump sum reduces the end portfolio size by 43%. Besides, simply purchasing inflation-indexed SPIAs with a larger portion of the nest egg by itself helps protect against longevity, inflation, and sequence of returns risks and avoids even having to wrap your head around how these complicated products work. Michael Kitces seems to agree that these living benefits riders aren't all they're cracked up to be, especially given recent changes in this marketplace.

I emailed both Mike Piper and Wade Pfau, both of whom have spent a great deal of time thinking about retirement income issues like these to get their thoughts on allocating a significant portion of their portfolio to VAs. Here are their thoughts:

Mike Piper:

Doing the actual analysis on annuities with living benefit riders is beyond my skill set, so I rely on the opinions of other people whom I trust. The following articles from Wade Pfau (including the linked-to articles) may be of interest:

Wade Pfau:

Milevsky's book is great, but I do share your concerns about VAs with guarantees. My efforts to create efficient retirement frontiers with product allocation always point to keeping upside and downside separated by using stocks and SPIAs.  For me, allocations to VAs are 0%. I'm not sure what exactly Moshe is assuming about the VAs to get them on the frontier with a sizeable allocation.

So there you have it. Everyone still agrees that purchasing a SPIA, especially an inflation-adjusted one, is a good idea for most retirees, especially those without a pension or those with barely enough to support their desired lifestyle. But purchasing VAs for their living benefits riders is still controversial, at best. And even Milevsky isn't suggesting you go out and purchase cash value life insurance for retirement income.

What do you think? What role do you see for “product allocation” in your life? Do you already own or plan to buy a SPIA? Why or why not? Do you already own or plan to buy a VA with living benefits? Why or why not? Comment below!