On loss-avoiding lump-sum pension optimization with contingent targets

Abstract

Consider a lump-sum pension fund problem, in which an agent deposits an amount with a fund manager up front and is later repaid a lump sum x(T) after time T. The fund manager may be both cautious in seeking a payoff x(T) meeting a certain target, but relaxed toward the possibility of exceeding this target. We use a computational method in stochastic optimal control (“SOCSol”) to find approximately-optimal decision rules for such “cautious-relaxed” fund managers. In particular, we examine fund optimisation problems in which the target is contingent upon market conditions such as inflation.