An Improved Particle Swarm Optimization Algorithm for Option Pricing

Abstract

To prevent premature convergence, an improved particle swarm optimization algorithm (IPSO) is proposed, in which the mean value of global optimal positions in iteration process and the mutation operation are considered. As an application, we apply the IPSO to estimate the implied volatility for European option which is a critical parameter in option pricing. Compared with existed binary particle swarm optimization algorithms, IPSO displays simpler implementation and faster convergence. Further, by using the IPSO, the estimated value of European call option gets more close to its actual value with fewer fluctuation.