Evolutionary simulation of hedging pressure in futures markets

We present a real world application that models a financial futures market. The agent-based simulation includes speculator agents each of which uses a Genetic Algorithm to improve its profitability in the market. This is a realistic simulation whose rates-of-return distribution is similar to those of real futures markets such as corn and FTSE100 futures. The futures markets have never before been simulated to this level of detail, and the simulation is used to test the long-held belief that speculators are more profitable if they incorporate "hedging pressure" into their price calculations - essentially, the use of market knowledge about supply and demand. Surprisingly, we show that hedging pressure cannot be used to improve profits for speculators.