News Analytics for Energy Futures

Abstract

We investigate how crude oil futures markets respond to positive and negative sentiment in news, as measured by the Thomson Reuters News Analytics. We measure the impact of the news sentiment on daily futures returns and on the fundamental factors of the forward curve evolution, such as the forward curve’s level and the slope. We define positive and negative news events as those days when the aggregated daily sentiment is in the top 10% quantiles of the corresponding sentiment distribution. We then relate these events to abnormal futures returns and to significant changes in the forward curve’s level and slope.

We find that, for all maturities, positive resp. negative news events are accompanied by long periods of higher resp. lower than normal returns. These abnormal returns are observed already prior to the significant news events, suggesting that news are correlated with price momentum. Furthermore, after a negative news event occurs, returns continue decreasing at least for the subsequent 20 days. This is especially pronounced for nearby maturities, whose returns decrease by more than 3% on average. After a positive news event, returns increase on average by 2-3%. These effects are both statistically and economically significant. We also show a great asymmetry in the market’s behavior: although positive and negative events are both 10% of days, the negative events are accompanied by much greater losses than the gains surrounding positive events.

We analyze the complex effect of news on the slope of the forward curve, and show that negative news increase the slope both in backwardation and contango markets. This means that contango deepens and backwardation flattens following a negative news event. Positive news have the opposite, albeit smaller, effect on the forward curve’s slope. Generally, the reaction to news is significantly more pronounced in contango than in backwardation market.

The results of the analysis lay foundation for profitable trading strategies, based on taking a long/short position in oil futures of a fixed maturity immediately after a significant positive/negative news event. However, we expect such strategies to be quite risky. The analysis of news effects on the entire forward curve can be used to construct profitable (but much less risky) calendar spread trading strategies.

Keywords: Oil Futures, News, Event Study

JEL Classification: G14, G12, C53