Investor sentiment and the price of oil

The literature on oil prices considers real economic factors as the main drivers of changes in oil prices. Using parametric and nonparametric methods, this study provides evidence that behavioral factors have the power to predict oil price movements. Based on monthly, weekly and daily data for 1986 to 2016, we find that investor sentiment, captured by nine different proxies, has a significant effect on oil prices. In addition, we demonstrate that volatility in these sentiment indices spills over and can explain part of the volatility in oil prices. Our findings are even more significant during and after the early 2000s, when oil-based financial products became a popular asset class for many funds and portfolio managers. We also use daily search query data from Google Trends to establish that oil shocks Granger-cause the attention of retail investors, and that a heightened number of searches can predict an increase in volatility in the trading days that follow.