Supervised Machine Learning Classification for Short Straddles on the S&P500

Abstract

In this paper, we apply machine learning models to execute certain short-option strategies on the S&P500. In particular, we formulate and focus on a supervised classification task which decides if a plain short straddle on the S&P500 should be executed or not on a daily basis. We describe our used framework and present an overview of our evaluation metrics for different classification models. Using standard machine learning techniques and systematic hyperparameter search, we find statistically significant advantages if the gradient tree boosting algorithm is used, compared to a simple “trade always” strategy. On the basis of this work, we have laid the foundations for the application of supervised classification methods to more general derivative trading strategies.