$N$-player and Mean-field Games in It\^{o}-diffusion Markets with Competitive or Homophilous Interaction

TL;DR

This paper develops and solves models of N-player and mean-field games in Itô-diffusion markets, analyzing how players' interactions via competition or homophily influence optimal portfolio strategies under various market completeness and risk tolerance scenarios.

Contribution

It introduces explicit solutions for equilibrium strategies in N-player and mean-field portfolio games with competitive or homophilous interactions, including cases with random risk tolerances.

Findings

Explicit equilibrium solutions derived for various market settings.

Analysis of the impact of interaction types on optimal strategies.

Extension to models with random risk tolerances.

Abstract

In It\^{o}-diffusion environments, we introduce and analyze $N$-player and common-noise mean-field games in the context of optimal portfolio choice in a common market. The players invest in a finite horizon and also interact, driven either by competition or homophily. We study an incomplete market model in which the players have constant individual risk tolerance coefficients (CARA utilities). We also consider the general case of random individual risk tolerances and analyze the related games in a complete market setting. This randomness makes the problem substantially more complex as it leads to ($N$ or a continuum of) auxiliary ''individual'' It\^{o}-diffusion markets. For all cases, we derive explicit or closed-form solutions for the equilibrium stochastic processes, the optimal state processes, and the values of the games.