Finkelstein and Poterba (2004) Flashcards
(19 cards)
What are the two main predictions tested in the paper regarding asymmetric information?
- Higher risk individuals self-select into insurance contracts that offer features more valuable to them than to lower-risk individuals.
- Eqm pricing of insurance policies reflects the variations in the risk pool across different policies
Why is the annuity market particularly good for testing adverse selection?
The annuity market has minimal moral hazard (individuals are unlikely to significantly extend their lives due to annuity ownership), so tests for asymmetric information directly measure adverse selection rather than a mix of selection and moral hazard effects.
What is an annuity?
A contract that pays its beneficiary (the annuitant) a pre-specified amount for as long as they are alive, insuring against the risk of outliving accumulated resources.
Why did Yaari (1965) find annuity markets puzzling?
Yaari documented the welfare-improving role of annuities for individuals facing uncertain mortality, making the small size of voluntary annuity markets puzzling.
What is the size difference between compulsory and voluntary annuity markets in the UK
The compulsory market is approximately 10 times larger than the voluntary market, both for the sample firm and for the industry as a whole
What did Finkelstein and Poterba find about adverse selection on the extensive margin between the two markets?
Adverse selection on the extensive margin (average price of annuity contracts) is roughly half as great in the compulsory market as in the voluntary one.
What three annuity features were examined for potential selection effects?
- Annual annuity payment amount
- Degree of backloading (payment profile timing)
- Payments to the estate (through guarantees or capital protection)
How does backloading relate to mortality risk?
Backloaded annuities (with payment profiles providing a greater share of payments in later years) are more valuable to individuals with longer life expectancy.
What is the “single crossing property” in the context of annuity features?
The property where, at a given price, the marginal value of each annuity product feature varies monotonically with risk type (meaning preferences shift only once as life expectancy increases).
What does the hazard function represent in this study?
The probability that an annuitant with characteristics x dies at time t, conditional on surviving up to time t.
What was the predicted coefficient relationship for annuities with payments to estates in the hazard model?
Positive coefficients were predicted for annuities with payments to estates (guaranteed or capital-protected), indicating higher mortality hazard and shorter life expectancy.
What were the findings regarding backloaded annuities and mortality?
In both compulsory and voluntary markets, individuals who bought more backloaded annuities were longer-lived, with coefficients on indicator variables being negative and statistically significant at the 1% level (-0.185, -0.152, LPM)
How did guarantee periods relate to mortality hazard?
The hazard increased monotonically from no guarantee period to 5 to 10 years, with statistically significant mortality differences between individuals with 5-year and 10-year guarantee periods.
What is the “money’s worth” concept?
The expected present discounted value of annuity payments divided by the initial premium. A money’s worth of unity represents an actuarially fair annuity.
How were prices affected by annuities making payments to estates?
Annuities that made payments to estates had significantly lower prices in both markets than annuities that did not make such payments (used a hedonoic regression model via OLS)
How were index-linked annuities priced compared to nominal annuities?
Index-linked annuities were priced significantly higher than nominal annuities in both compulsory and voluntary markets.
This shows that the typical annuitant who purchases these are longer lived than the typical annuitant who purchases a nominal annuity.
What was found regarding the pricing of larger annuity payments?
The positive coefficient on the square of the initial annual payment was consistent with Rothschild and Stiglitz’s (1976) prediction of a higher marginal price for larger quantities of insurance.
Why did the authors discount moral hazard as an explanation?
They suspected that households were unlikely to substantially modify their behavior in response to annuity income, citing Banks and Emmerson (1999) who showed annuity income represented less than one-fifth of annual income.
What data did they use?
Administrative data fom a large UK annuity company over a 17 year period. Trends from the sample firm were sure to mirror industry wide trends reported in the Insurance statistics year book.
