Millions of American and European workers invest some of their payroll taxes in selfdirected retirement accounts of the defined contribution variety. While it is generally agreed that such accounts can augment defined benefit pension plans, some policymakers worry that financially inexperienced individuals may fare poorly if they do not understand risks associated with investments and payout options. Recently the US Treasury Department has undertaken an initiative “to give employees and employers more options for putting the pension back” in private retirement systems (T.D. 9673, RIN 1545-BK23; Iwry 07/2014). Comparable to German Riester plans, this initiative provides the possibility to convert accumulated assets into a longevity income annuity, which starts lifelong payments at a maximum age of 85. While traditional life annuities offer households an effective way to hedge longevity risk, the illiquidity of these instruments is a major disadvantage since premiums are not refundable (e.g. in case of health shocks or early death). It is an open question whether longevity income annuities should provide some liquidity (withdrawal) options, and how the insurer should price such features. This subproject quantifies the welfare implications of incorporating life annuities (with liquidity options) in privateretirement accounts. We study the impact of households’ demographic characteristics (gender, education), changing family status (marriage, divorce, death), unexpected deterioration of health, and tax-requirements (like the RMD-rules for retirement accounts). This project contributes to the literature on life cycle consumption-investment choice, and our findings are of interest to policymakers seeking to reform private pension account systems to enhance retirement income security in an aging society.