Part Two of the Interview with Wharton Professor David Babbel

This is the second part of an interview with Professor David Babbel.

Part one can be found here.

Annuity Digest: Which study results were most surprising to you?  I assume you went into the study with firm views on the non-normality of asset returns.  Given that, was it: a) short-run comparative results, b) long-run results, c) valuation results, or d) performance in light of the risk tolerance overlay?

Professor Babbel: I was not surprised by the positive stochastic dominance results or the mean-variance results, both of which showed that the returns on FIAs historically, when adjusted for risk, performed quite well relative to alternatives. Indeed, for most levels of risk aversion, they have dominated the alternatives. However, what surprised me was that the returns on FIAs outperformed the alternatives over the lifetime of their existence (since 1995) for every year that they have been issued. This relates to 9-year FIAs and 14-year FIAs, held through maturity. This superior performance prevailed EVEN BEFORE there was any adjustment for risk, which adjustments typically reduce the risk-adjusted returns of the alternatives to FIAs, thereby boosting even more the FIA relative returns.  The second finding that surprised me was that when the observed non-Normality of asset returns was taken into account, it was about twice as likely to have FIAs crediting medium and high rates than what would occur under an assumed Normal distribution.  I suspected that non-Normality might have some effect, but never thought that it would be as dramatic as it is.

Annuity Digest: In a world of high volatility and asset returns that are not normally distributed, isn’t there a very strong case to be made for a product that captures part of the market upside while having a floor of protection for large downside possibilities?

Professor Babbel: Yes, as long as the pricing is appropriate and not gouging.

Annuity Digest: Is the case for the FIA over any time period pretty much made by high volatility and asymmetric asset returns?  Would this case be undermined in a low volatility environment?

Professor Babbel: Yes, but it would again depend on appropriate (competitive) pricing.  The case is not necessarily undermined in a low volatility environment. The reason is that the crediting rate caps and participation rates depend on volatility, because higher volatility makes it more expensive for the insurer to hedge the promises and guarantees in the contract. Indeed, the most generous contractual provisions can occur when interest rates are high and volatility is low. Of course, more generous crediting rate caps and participation rates do not imply a better product, because it would then be less likely to make high FIA returns due to the lower volatility.

Annuity Digest: The sequence of returns effect is starkly evident.  Can you talk a bit about this in light of some of the highly volatile periods we have experienced in the capital markets over the past 15 years or so—and also possibly on a forward looking basis in light of asymmetric asset returns?

Professor Babbel: No, because I have not studied this.

Annuity Digest: Can you capture the essence of the valuation argument for our consumer audience?

Professor Babbel: The FIA can be a product particularly suited to people with moderate to high risk aversion, yet who still wish to have some exposure to equity returns.

Annuity Digest: Can you talk a bit—again, if possible, for a consumer audience—about what goes into expected utility and also the risk aversion coefficient?

Professor Babbel: Utility analysis is at the heart of all economics.  It involves models of human behavior and happiness/satisfaction. In all economic models, “utility” or happiness/satisfaction increases as wealth/consumption increases. Conversely, “utility” falls when wealth/consumption falls. If a $10,000 increase in wealth gives you additional utility, then a $10,000 loss in wealth would create in the risk-averse consumer a decrease in utility of an even greater amount.  For example, if an increase in your wealth of $10,000 raises your utility by, say 10 “utils” (units of happiness/satisfaction), than a similar decrease of $10,000 would injure the risk-averse consumer by, say 12 “utils” or 30 “utils” or even more, depending on how risk averse the individual is. Many studies have been done that show this pattern among most consumers. It is not uncommon for an individual to require a much higher possible payoff on an investment that has a chance of losing money.  The index of risk aversion that is typically used in economics studies ranges from 0 for no risk aversion to 90 for extreme risk aversion. Once risk aversion goes above a 2 or 3, the FIAs begin to shine relative to their investment alternatives. Recent studies show that most consumers have risk aversion levels well above those levels, although those who invest directly in stocks often show risk aversion levels in the 0 to 2.5 range.

Annuity Digest: Can you provide some clarification—ideally for a consumer audience—on the implications of actual/historical crediting rates versus simulated crediting rates?  Is it the asymmetric aspect of the historical crediting rates that work in favor of annuity comparative performance?

Professor Babbel: The main advice would be that the consumer (1) is comfortable with the quality of the company standing behind the contract, examine carefully the structure of the FIA and see whether it (2) provides them with satisfactory downside protection and (3) an upside potential that is sufficiently attractive. In looking at historical crediting rates, keep in mind that they will depend on two things beyond the contract’s structure and initial parameters: what actually happened to the index to which the interest credits are linked, and what changes did the company make over time in the caps or participation rates. The company cannot control what happens to the index but can control the caps or participation rates that it periodically revises. It is not generally an indication of profiteering on the company’s part to reduce the caps or participation rates – it may be occasioned simply by the cost of hedging having increased due to higher market volatility. There is likely an annual budget available embedded within the contract for the hedging costs, and the budget will stretch further when the hedges are cheap, which typically happens when volatility is reduced. The company cannot control market volatility as it evolves over the life of a contract, so some resetting of contract levers is prudent and to be expected. Nonetheless, some useful information can be gleaned from a historical review of crediting rates.