Kollias et al (2011) use a nonlinear BEKK-GARCH model to see how war and terrorism affect the covariance of oil prices and the four main stock market indices (S&P500, DAX, FTSE100, CAC40). They found that war has a lasting effect on the covariance, while terrorist attacks have a one-off shock on the oil price, DAX and CAC, while the S&P500 and FTSE100 were not significantly affected since they are more efficient markets and deep enough to withstand the consequences of a terrorist attack. Their research demonstrates that war/terrorism events are important to consider when examining the price of oil, which in turn influences some stock market indices and would impact the stock price of multinational oil companies. The impacts these events have on oil prices are shown in the graph below. For example, 9/11 clearly increased the price of a barrel of crude oil (EIA, 2014). 9/11 is one of the events being evaluated and I'm interested to see if it significantly affected the stock price of this super major. d) How changes in oil prices affect stock market returns In this subsection, I will examine how changes in oil prices affect stock market returns with reference to academic papers Ready (2012) found that changes in the supply of oil (which ultimately shifts its supply curve causing the price to change) have a large impact on stock markets around the world. Ready explicitly highlights the fact that countries that are heavily dependent on imported oil have seen their stock market returns react more than other countries. Specifically, he mentions the US stock market as being hit more than others, which is in line with statistics showing that the US is actually the big… center of the chart… risk sensitivity of the stock to the market and Alpha (α), which is the average unexplained return (i.e. the return not explained by the market). (Weston et al, 2004). Weston et al state that because the market model takes risk into account, it is the most commonly used method for event studies in general. MacKinley (1997) agrees with this due to the simplicity of the market model (it is a one-factor model), its linearity and the limitations of multifactor models, also stating that it represents an improvement over other models. Other event studies of a similar nature to mine, such as Zevenbergen's (2008) paper on oil spills impacting the stock price of publicly traded oil companies, use the market model to find the AR, CAR, AAR, and CAAR of each company in a given period. Therefore, I believe that the market model will be the best model to use for my proposed event study.