Oil Price Uncertainty Index: Capturing Media Information

Abstract

We develop a new oil price uncertainty index (OPUX) based on newspaper and social networking coverage. Our index improves on the existing oil ETF volatility index (OVX) because financial markets are inefficient insofar as prices do not reflect all information (Shiller, 1981). Several pieces of evidence, including a media audit of 14,233 newspaper articles and 502 tweets, indicate that our OPUX predicts changes in the uncertainty of oil prices. Specifically, our results demonstrate that the OPUX improves ability to predict oil price volatility (the standard deviation of daily prices for a given month) between 24% and 30%. The intuition behind our results is that the existing OVX is fairly smooth, whereas actual oil price volatility shows considerably more fluctuation in step with media coverage (see Figure 1, Panels A, B and C). Collectively, newspaper and social media coverage are powerful predictors of future volatility in the oil market.

Keywords: Commodity, oil price uncertainty, forecasting

JEL Classification: Q02, E17