Use of distributed computing in derivative pricing

Monte Carlo option pricing algorithms are well suited to distributed computing because simulations can be run on different computational units with no need for communication between these tasks. In this paper we investigate and compare the use of two distributed computing environments for such computation: a PC grid that exploits the spare computing capacity of up to 470 computing cores in 300 office and teaching lab PCs scattered on a university campus, and a scientific computing cluster of 120 computing cores in 32 rack-mounted servers. We outline the process of adapting a Monte Carlo algorithm for computing prices for a set of 100 arithmetic Asian options with stochastic volatility to run on these environments, and investigate the performance for different distributing strategies. The paper closes with a discussion of the opportunities and challenges of distributed computing in computational finance.