Stock Portfolio Prediction by Multi-Target Decision Support

Abstract

Investing in the stock market is a complex process due to its high volatility caused by factors as exchange rates, political events, inflation and the market history. To support investor's decisions, the prediction of future stock price and economic metrics is valuable. With the hypothesis that there is a relation among investment performance indicators, we applied multi-target regression (MTR) methods to estimate 6 different indicators aiming at creating an automated prediction tool for decision support. The experiments were based on 4 datasets, corresponding to 4 different time periods, composed of 63 combinations of weights of stock-picking concepts each, simulated in the US stock market. We compared traditional machine learning approaches with four state-of-the-art MTR solutions: Stacked Single Target, Ensemble of Regressor Chains, Deep Structure for Tracking Asynchronous Regressor Stacking and Multi-output Random Forest (MORF). With the exception of MORF, traditional approaches and the MTR methods were evaluated with Random Forest and Support Vector Machine regressors. By means of extensive experimental evaluation, our results showed that the most recent MTR solutions can achieve suitable predictive performance, improving all the scenarios (12.6% in the best period, considering all target variables). In this sense, MTR is a proper strategy for building stock market decision support system based on prediction models.