Answers are grounded in this paper and cite their sources.
Consumption and saving models typically suggest accumulation in life-only annuities that is at odds with actual investments. Milevsky and Salisbury (2022) add another dimension to this annuity puzzle by documenting a shift from life-only annuities to those offering money-back guarantees to the heirs of the annuitant. We provide further evidence for a balanced coexistence of both annuity types, however, standard utility frameworks that rely solely on consumption and bequest utility fail to reproduce this relationship. To overcome the dominance of life-only annuities, we incorporate loss aversion into individual preferences. Specifically, we assume aversion to dying before the total payouts exceed the premiums paid, creating a natural obstacle to life-only annuities. Using realistically calibrated life cycle models with access to liquid savings and both life-only and refundable annuities, we show that loss aversion can flexibly control the demand for annuities, thereby enabling a balanced relationship between the annuity types. While our findings shed further light on behavioral drivers behind the annuity puzzle, our utility framework does not simultaneously establish the empirical relationship across annuity types as well as the overall demand for annuities.
No takes yet. Share an insight, caveat, or question.
Hallstein et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: