Abstract
Longevity insurance annuities are deferred annuities that begin payment at advanced ages, such as age 82. They provide insurance against running out of money in old age. They simplify the retirement spend-down calculations in that they can be used to convert a problem with an unknown end date (date of death) to one with a fixed end date, which is the start of the longevity insurance annuity. They may allow retirees to have riskier portfolios because they are a steady source of income. In this study, we compared the provision of longevity insurance annuities in the private and public sectors. We analyzed the need for longevity insurance annuities using Monte Carlo simulations that demonstrate the risk of running out of money in a 401(k) account and modeled how longevity annuities might work in practice.
Disclosure: The authors report no conflicts of interest with respect to this article. The research for this article was performed while Dale Kintzel was employed as an economist at the US Social Security Administration.
Editor’s Note
Submitted 17 January 2020
Accepted 5 August 2020 by Stephen J. Brown.
This article was externally reviewed using our double-blind peer-review process. When the article was accepted for publication, the authors thanked the reviewers in their acknowledgments. Moshe Arye Milevsky, Raul Leote de Carvalho, and Don Ezra were the reviewers for this article.
Acknowledgments
We have received valuable comments from Stephen J. Brown, Luis Garcia-Feijóo, CFA, CIPM, participants at the 2015 Pension Policy Research Group conference in Dublin, Jason Brown, Irena Dushi, and Gayle Reznik. We have also benefited from collaboration on earlier papers on longevity insurance annuities with David Blake, David McCarthy, David Rajnes, Gerard Hughes, Mark Iwry, Agnieszka Chłoń-Domińczak, and Tianhong Chen.
Notes
5 In practice, there is very little difference in outcomes when lower rebalancing frequencies are used. See Vanguard (2015) for more details.
6 Bootstrapping is done by random sampling, with replacement, of the returns for each of the three investment options in the dataset. The computer randomly samples and records monthly returns from the dataset, puts that return series back in the dataset, and repeats the same sampling process to construct 100,000 blocks of 40-year investment horizons to correspond to the length of the life expectancy table.
8
and
, respectively.