Adverse Selection in Annuity Markets

Annuity markets around the world have not developed as well as other insurance products and one explanation is that adverse selection could limit the scope for its development. The Chilean case is interesting as the entire... [ view full abstract ]

Annuity markets around the world have not developed as well as other insurance products and one explanation is that adverse selection could limit the scope for its development. The Chilean case is interesting as the entire pension system relies on individual savings through compulsory individual retirement accounts. Upon retirement, workers must choose between programmed withdrawals or life annuities. We estimate survival models as a function of pension choice, providing direct evidence of adverse selection in the annuity market. Interestingly, we find that after a regulatory intervention in 2004 that brought both transparency and competition to the market (i.e., the introduction of SCOMP), the adverse selection effect became larger and/or more prevalent for all market segments. Women (men) who choose an annuity live on average five (two and a half) years longer than those choosing the phased withdrawal option. To control for potential unobserved characteristics that could be related to the choice of pension product, we use interest rate data at the time of retirement as an instrumental variable, within the framework of a parametric duration model. In this case, the marginal effect of annuitization is no longer significant, suggesting that there is no causal effect of annuities on longevity.