The attributes of people who purchase a given insurance product (e.g., an annuity) can significantly affect the expected benefits of that product. Specifically, differences in demand (i.e., total sales) and life expectancies (i.e., mortality experience) for different types of annuities can potentially affect expected income levels.
While this point is relatively obvious, I'm not sure it is fully appreciated by everyone in the decumulation space when estimating the efficiency of different annuities.
The "optimal" annuity cannot be determined in isolation, and strategies that balance behavioral and economic efficiencies are likely going to end up being the ones that provide the greatest benefit to retirees.
A Hard Sell
The lack of annuitization among Americans is well-known and contrary to the consensus among economists. Many note that holding retirement assets in an annuity-type structure is more efficient than withdrawing from an investment portfolio during retirement to fund consumption given an uncertain lifespan (i.e., self-annuitization).
The lack of demand for annuities is due to a variety of factors, one of which is product design. Early annuities, such as single premium immediate annuities (SPIAs), which are especially common in retirement research, require an irrevocable transfer of the insurance premium for the promise of lifetime income.
This is a difficult "trade" from a behavioral perspective. Deferred income annuities (DIAs) — in particular, qualified longevity annuity contracts (QLACs) — represent an especially difficult purchase, given the delay from purchase to income commencement. Though widely touted among retirement academics, they remain relatively muted.
Not surprisingly, combined sales of SPIAs and DIAs are relatively low in terms of total annuity sales, representing 3% of the total in 2021 based on an analysis by LIMRA.
While total sales for SPIAs and DIAs in 2021 would still be relatively larger given the absolute value (at $6 billion and $2 billion, respectively), there are clearly other annuities with higher demand among households (e.g., variable annuity strategies with some living benefit) that could potentially improve pricing based on increased market competition.
Mortality Experience
Mortality experience also differs across annuities. This effect is demonstrated in the graph below, which includes mortality experience data from the Society of Actuaries 2009-2013 Individual Payout Annuity Mortality Experience Report for deferred and immediate annuities and the Ruark 2018 Variable Annuity Industry Mortality Experience Study for variable annuities.
For context, the lower the ratio, the healthier the respective annuitants (and vice versa).
Annuities with lower "commitment" levels tend to have annuitants that are less healthy, on average. For example, individuals who purchased variable annuities, which are typically fully liquid (and revocable) had mortality rates that were consistent with annuity mortality tables, while mortality rates among annuitants in DIAs were less than half of expected mortality rates.
