Portfolio Selection: A Game-Theoretic Approach

Abstract

Game-theoretic methods are not widespread in finance. One reason is that practitioners do not see straightforward applications of game theory to core investment problems. To that end, in this article the author describes two novel portfolio construction applications of the game-theoretic solution concept known as the Shapley value. In the first, the Shapley value is used as a type of shrinkage operator to blend the allocation weights generated by two different portfolio selection frameworks. In the second, the Shapley value is used to build a unique approach to portfolio selection using what the author calls an optimization game. In both applications, asset classes are assumed to be the players who are bargaining over representation in a portfolio. As the article shows, the Shapley value naturally facilitates the construction of well-diversified portfolios that can satisfy specific investor goals without the need for additional formal constraints.

TOPICS: Portfolio construction, portfolio theory, portfolio management/multi-asset allocation