A sparsity constrained dynamic trading framework to identify stochastic arbitrage opportunities with market index options

Abstract

This paper formulates a mixed-integer linear program that applies sequential stochastic dominance in a dynamic, daily-rebalanced trading framework to identify stochastic arbitrage opportunities with options written on the S&P 500 index. This paper finds that a dynamic trading framework succeeds at identifying stochastic arbitrage opportunities more frequently than static trading frameworks explored in recent research, but both strategies invariably return payoffs inferior to the market. This paper shows that the application of sparsity constraints to both static and dynamic trading strategies yields improved performance.