One Step Back, Two Steps Forward - a Machine Learning-Powered Options Trading Strategy

Abstract

Stock options are an often-traded financial derivative which allow traders to insure themselves against making bad trades in exchange for a small premium. A certain combination of stock options, known as a straddle, will be profitable as long as the price of the underlying asset moves more than the cost of this option. This work explores the possibility of predicting whether or not such an even will occur, using machine learning techniques, and applies a custom trading strategy to these predictions in order to generate earnings in any market direction.