Black-Scholes option pricing via genetic algorithms

Abstract

Investors use the Black-Scholes option pricing model to find theoretically correct option prices. These prices are then compared to market prices to discover mispriced options. However, a difficulty arises in the use of the model, because one variable in the model must be estimated: the instantaneous variance of the underlying security's returns. This is usually proxied by the implied volatility of the security's returns. The more accurately investors are able to estimate this value, the more accurate their estimates of theoretical option values will be. It is argued that investors can find more precise theoretical values of options by using genetic algorithms than by using traditional calculus-based search techniques to find estimates of the implied volatility.