The Effects of Crude Oil on Stock Markets with use of Markov Switching Models

Abstract

In this thesis, a two regime Markov switching (MS) model is implemented to examine the relationship between crude oil, both brent oil and WTI, and stock markets. In particular, the model is applied to stock markets in both oil importing and exporting countries which include Canada, China, Japan, Germany, Netherlands, Norway, the United Kingdom and the United States. This paper first evaluates the significance of oil parameters in the detected regimes, where the two regimes respond to low mean and high variance (bear state), as well as high mean and low variance (bull state) respectively. We find evidence of stronger significance of oil returns in high volatility regimes. Overall, crude oil plays a significant role in determining stock returns. There is a stronger and more consistent relationship between oil and stock market in oil importing nations, regardless of regimes. The paper also presents an estimation of the correlation between oil and national indices for both regimes. The results provide further evidence of consistently higher correlation in high volatility regimes. The correlation ratio between the regimes are higher for oil importing nations, indicating that such nations are more strongly affected by volatility regimes.