A machine learning approach to predict the S&P 500 absolute percent change

Abstract

Abstract Models of the stock market often focus on predicting the direction of the stock market. Instead of following this approach, we created a model to predict the daily absolute percent change of the S&P 500. An accurate model of this metric would greatly increase profitability of option trading strategies such as straddles and iron condors. In this project, novel features were created based on historical data and fed to machine learning algorithms such as Decision Trees, Rule Based Classifiers, K-mean Clusters, and Kernels. Based on our findings, Decision Trees and Kernels showed an accuracy of 80% when predicting absolute percent change, while Rule Based Classifiers had an accuracy of 88%.