Minimax Policies for Selling a Nondivisible Asset

Abstract

In this paper we have obtained a simple expression for the minimax regret policy for selling an asset. The policy is very simple to implement. As opposed to most optimal stopping rules, it does not involve any iterative computation of the stopping boundary. Also in using this result, only the maximum possible increase and decrease in prices in each period have to be known, as opposed to the entire distribution function of price changes in each period.