Collective intelligence in option pricing: dftermining black-scholfs implied volatilities with generalized ant programming

In this paper we use Generalized Ant Programming to derive analytical approximations for determining implied volatilities of European-style options as well as American call options. Generalized Ant Programming is a new method inspired by Genetic Programming as well as Ant Systems and is designed to generate computer programs. Using experimental data sets we can show that the Generalized Ant Programming based formulas for calculating implied volatilities deliver accurate approximation results and outperform other approximations prevented in the literature.