It's RILA time: An introduction to registered index‐linked annuities

Abstract

Registered index-linked annuities (RILAs) are increasingly popular equity-based retirement savings products offered by US life insurance companies. They combine features of fixed-index annuities and traditional variable annuities (TVAs), offering investors equity exposure with downside protection in a tax-deferred setting. This article introduces RILAs to the academic literature by describing the products' key features, developing a general pricing model, and deriving the providers' hedging strategy by decomposing their liabilities into short-term European options. Numerical illustrations show that RILAs offer investors similar risk profiles (in the long run) as TVAs with maturity guarantees, and that many products currently sold appear to be priced quite favorably for investors. For providers, RILAs may be a preferable alternative or complement to TVAs as they greatly simplify the management of the embedded equity risk and can naturally reduce the TVA capital requirements. These features position RILAs as a viable long-term solution for this product space.

The Appeal of Registered Index-Linked Annuities for Myopic Prospect Theory Investors

“…However, we argue that our model provides a good starting point for our analysis, as it provides a good proxy for policyholder cap rates a policyholder might expect in the future. Furthermore, the model coincides with Moenig (2021) under an equal set of assumptions.…”

“…Furthermore, we see that for investors bracketing over five years the optimal choice would be the five-year RILA II for equally reasonable (albeit somewhat higher) threshold values, thereby explaining the existence of both products in the market (and adding support to the hypothesis that the framing of payoffs in marketing materials matters). In practice, insurers usually offer higher than fair-value cap rates on RILAs with longer crediting periods, as they expect to earn a spread above the risk-free rate for investments with longer maturities, which they then partially pass on to the policyholder (Moenig (2021)). We do not capture this effect in our model, but note that our results should therefore be interpreted as conservative, as a RILA with an equal buffer but a higher cap could only be deemed as more attractive by any policyholder (simulations including higher caps and the investment spreads necessary to generate them can be found in the Appendix).…”

“…At contract maturity, which usually occurs once a policyholder retires, the account value is paid out to the policyholder as a lump sum. 3 The combining feature of all RILA guarantees is some type of downside protection against negative stock index returns financed by some limitation on positive returns (see Moenig (2021) for a detailed market overview). In a single crediting period the protection on the downside can either come in form of a buffer, covering the first tranche of losses, or a floor, covering all losses after a certain percentage.…”

“…Following Moenig (2021), we see that the buffer/cap crediting rate RILA CB t (C, B, τ ) can also be expressed as…”

“…In practice, a buffer/cap RILA has a fixed buffer and crediting period length, but varying caps over crediting periods to adjust for changes in the market (such as higher volatility or changing interest rates). As we model the stochastic process behind S P R A more flexible pricing approach based the perspective of an insurer hedging a RILA can be found in Moenig (2021), which is model-independent by incorporating empirical option prices, fees, surrender charges and credit/illiquidity spreads an insurer may expect to earn over the risk-free rate. However, we argue that our model provides a good starting point for our analysis, as it provides a good proxy for policyholder cap rates a policyholder might expect in the future.…”