The Effect of Performance Metrics and Sentiment Scores on Selecting Oil Price Prediction Models

Predicting crude oil prices is an important yet challenging forecasting problem due to various influencing quantitative and qualitative factors. To address the growing number of potential prediction models and model parameters that researchers and practitioners need to consider during model selection, we suggest a systematic comparison of alternative prediction models and variables. In this article, we provide a novel perspective on oil price prediction models by comparing a variety of different forecasting models and considering both their statistical and financial performance. To assess the usefulness in a practical setting, we evaluate the predictions in a simulation of a simple trading strategy. We show that the ranking of different approaches depends on the selected evaluation metric and that small differences between models in one evaluation metric can translate into large differences in another metric. Finally, we show that including qualitative information in the prediction models through sentiment analysis can yield both statistical and financial performance improvements