Structural breaks in mean reverting processes: Empirical study of WTI-Brent futures spreads

Abstract

The purpose of this study is to examine the implication of structural breaks in mean reverting processes on the expected return of spread trading. Previous research focuses on the effective- ness of threshold filters in mean-reverting models when deciding trading strategies to exploit arbitrage opportunities within the spread of two highly correlated commodity futures. It is often assumed that high levels of co-integration are persistent in these futures prices, therefore ignoring the risks associated with structural breaks. Conducting an event study, this thesis uses an intensity-based model to measure the risk and return associated with structural breaks and changes in the properties of the spread process. The relationship between the two oil futures WTI and Brent have recently experienced considerable structural changes after a long period of stable relation, making the point of change an interesting event to study. The results aim to show how the changes in the mean levels affect risk and return.