The Second RP-PCA Factor and Crude Oil Price Predictability

Although it is notoriously difficult to utilize financial ratios to forecast the crude oil market prices, our study challenges this perception and reveals that the second risk premium principal component analysis (RP-PCA) factor may contain statistically significant information for both in-sample and out-of-sample forecasts of future crude oil prices. Our evidence illustrates that the second RP-PCA factor substantially outperforms many other popular predictors (approximately 30 conventional predictors) in forecasting crude oil prices and generating adequate higher values of economic profits. We conduct a range of informative tests, including bootstrap simulation, success ratio tests, alternative out-of-sample evaluation periods, and structure break tests. Furthermore, we illustrate that the forecasting ability of the second RP-PCA factor may stem from its ability to forecast oil market sentiment. Our study presents a novel and indicatable financial instrument for policymakers to predict crude oil prices robustly. The theoretical motivation of this study links to Cochrane's (2005) framework for general candidate factors in asset pricing.