Derivative Trade Optimizing Model Utilizing GP Based on Behavioral Finance Theory

摘要

This paper proposes a new technique which creates strategy trees for derivative (option) trading investment decisions based on behavioral finance theory and optimizes them by evolutionary computation in order to achieve high profitability. The strategy tree uses technical analysis based on a statistical, experienced technique for investment decisions. The trading model is represented by various technical indexes, and the strategy tree is optimized by genetic programming (GP), a form of evolutionary computation. This paper also proposes a method using the prospect theory based on behavioral finance theory to set the psychological bias for profit and deficit and attempts to select the appropriate strike price of options for higher investment efficiency. This technique was found to produce good results and the effectiveness of this trading model by the optimized dealings strategy was demonstrated.