Periodic evaluation of defined-contribution pension fund: A dynamic risk measure approach

TL;DR

This paper proposes a dynamic risk measure framework for periodic evaluation of defined-contribution pension funds, utilizing reinforcement learning to optimize strategies amid stochastic environments and mortality uncertainties.

Contribution

It introduces a novel dynamic risk measure approach combined with a model-free reinforcement learning algorithm for pension fund management, considering interim evaluations and mortality improvements.

Findings

Periodic evaluations promote risk-averse strategies.

Mortality improvements lead to risk-seeking behaviors.

Reinforcement learning effectively optimizes investment strategies.

Abstract

This paper introduces an innovative framework for the periodic evaluation of defined-contribution pension funds. The performance of the pension fund is evaluated not only at retirement, but also within the interim periods. In contrast to the traditional literature, we set the dynamic risk measure as the criterion and manage the tail risk of the pension fund dynamically. To effectively interact with the stochastic environment, a model-free reinforcement learning algorithm is proposed to search for optimal investment and insurance strategies. Using U.S. data, we calibrate pension members' mortality rates and enhance mortality projections through a Lee-Carter model. Our numerical results indicate that periodic evaluations lead to more risk-averse strategies, while mortality improvements encourage more risk-seeking behaviors.