Why Some Income Guarantees Are a 'Good Deal'

Insurance, by definition, should not be wealth maximizing. People buy insurance to transfer risk, and insurance companies must make a profit or they would be unable to pay future claims. Given these relatively simple facts, though, not all insurance products are created equal.

One approach commonly used to determine the potential benefit of insurance is to estimate the economic value of the product, where the expected benefits are weighed against the costs (i.e., the premium).

Researchers have been providing some context on economic values for decades, especially for annuities that provide guaranteed lifetime income, such as single premium immediate annuities (SPIAs).

Payout rates for SPIAs and deferred income annuities (DIAs) are going to vary both over time and across insurers.  As we’ve explored in our two articles (both of which are available online), this variation is important because it suggests the economic value is likely to vary based on both when the individual purchases the annuity and from which provider.

In this piece, we summarize some key findings in a forthcoming white paper, where we use a dataset of historical life-only annuity quotes provided by CANNEX, from March 3, 2013, to July 10, 2022, when the income is either immediate (i.e., a SPIA) or delayed (i.e., a DIA), by two, five, 10 or 20 years.

For the analysis, we estimate the “actuarially fair” payout for the annuity based on the respective features where the discount rates are based on the Treasury High Quality Market (HQM) Corporate Bond Yield Curve and mortality is based on the Society of Actuaries Individual Annuity Mortality (2012 IAM) Table, incorporating forecasted improvement.

We use mortality rates based on those individuals who actually purchase an annuity (i.e., realized experience among actual policyholders) versus more general mortality tables (e.g., Social Security Administration tables) because unhealthy people don’t typically buy annuities.

There are obvious adverse selection issues at play that should be considered when estimating the economic value of an annuity (or any insurance contract) and to use general population mortality rates ignores this fact.

The exhibit below includes an estimate of the average economic value of annuities for a 65-year-old annuitant over the full period.  Each (bi-weekly) available quote is compared to the estimated economic value and the best, median, and worst available percentages are provided.