Can Commodity Price Uncertainty Indexes Be Improved by Capturing Media Information? The Case of Oil Price Uncertainty

Abstract

Uncertainty indexes are used in economics to assess the uncertainty regarding economic policy, and in finance to quantify the uncertainty associated with the price volatility of financial instruments and commodities. Existing uncertainty indexes are typically constructed using realized (historical) price volatility, implied volatility when options are available, and conditional volatility using various econometric models. In this article, the authors present an improved commodity price uncertainty index, an oil price uncertainty index that includes information from news articles and social media. They demonstrate the superior performance of their new index compared with existing indexes that do not include news and social media information. They also create an oil price volatility forecasting model based on media coverage and use it to trade an oil volatility portfolio, obtaining a return that exceeds the average monthly growth of oil price volatility. Altogether, media coverage is a powerful predictor of future price volatility in the oil market.

TOPICS: Volatility measures, performance measurement

Key Findings

• • Media information from newspapers and Twitter about oil price uncertainty is combined with the existing OVX oil ETF volatility index to create a new oil price uncertainty index (OPUX).

• • The OPUX model gives a significantly better prediction of 1-month forward oil price volatility, as measured by an increase in R-squared from 12.1% for OVX to 33.0% for OPUX.

• • A volatility trading model based on OPUX exceeds the average oil price volatility by 0.29% per month.