Date of this Version
We analyze the effect of health cost risk on optimal annuity demand and consumption/savings decisions. Many retirees are exposed to sizeable out-of-pocket medical expenses, while annuities potentially impair the ability to get liquidity to cover these costs and smooth consumption. We find that if out-of-pocket medical expenses can already be sizeable early in retirement, full annuitization is not optimal. In the other case of low health cost risk early in retirement, individuals should take advantage of the mortality credit that annuities provide and save out of the annuity income to build a buffer for health cost shocks at later ages. When comparing to empirically observed levels of annuitization, we find that high health cost risk early in retirement may resolve the annuity puzzle. Moreover, we explain the observed pattern of annuitization as a function of initial wealth at retirement. For personal financial planning purposes, we develop a simple rule of thumb for annuity demand, based on expected health cost risk early in retirement, wealth at retirement, and minimum consumption levels. We show that the welfare costs from using the rule compared to the full life cycle model are small.
Optimal life cycle portfolio choice, health cost risk, annuity, retirement
D14, D91, G11, I1
Working Paper Number
Opinions and conclusions are solely those of the author(s) and do not reflect views of the institutions supporting the research, with whom the authors are affiliated, or the Pension Research Council. Copyright 2010 © Pension Research Council of the Wharton School of the University of Pennsylvania. All rights reserved.
We thank Katie Carman, Norma Coe, Donna Gilleskie, Thijs van der Heijden, David Hollanders, John Bailey Jones, Ralph Koijen, David Love, Ronald Mahieu, Roel Mehlkopf, Olivia Mitchell, Lisanne Sanders, Kent Smetters, Ralph Stevens, Annie Yang, and seminar participants at the University at Albany, University of North Carolina, Tilburg University, The Wharton School, and the Netspar Pension Day for helpful comments and suggestions. Part of this research was conducted while Kim Peijnenburg was visiting the Wharton School of the University of Pennsylvania, and she gratefully acknowledges their hospitality. We acknowledge financial support by All Pensions Group (APG). We thank Eric French for sharing the parameter estimates from De Nardi, French, and Jones (2010a) and Stijn van Nieuwerburgh for sharing the estimates from Ameriks, Caplin, Laufer, and Van Nieuwerburgh (2009). The most recent version of this paper is available at www.kimpeijnenburg.com
Date Posted: 07 August 2019