Execution game in a Markovian environment (Financial Modeling and Analysis)

Abstract

This paper examines an execution game model in a Markovian environment. We focus on how two risk-averse large traders execute a large volume of a risky asset to maximize the expected utility of each large trader from the terminal wealth over a finite horizon. The price impact caused by each large trader and the Markovian environment are assumed to affect the market and execution price. A formulation as a Markov game model enables us to solve this problem. We obtain an equilibrium execution strategy and its associated value function under a Markov perfect equilibrium via the backward induction method of dynamic programming.