Firstly, it is imperative that I investigate the stochastic properties of each series considered in the model prior to estimating the effects of oil price shocks on macroeconomic activity. This is done by analysing the order of integration from using a unit root test. Specifically, in this project the Augmented Dicky-Fuller (ADF) test will be used.Asteriou & Hall (2011) write the three different types of ADF equations as;

Where Y_t = each variable (GDP, Inflation, Interest rates, Exchange rates, Unemployment and Oil prices.)

The difference between each equation is the appearance of  and  respectively. These are deterministic elements, noted as the constant and trend respectively.

For this test, the hypotheses are stated as;

The null hypothesis,  infers that a unit root exists, whereas the alternative hypothesis,  infers that there is no root. Once this test has been passed, an appropriate lag length will need to be determined for the VAR model. If any variable does not pass this test and contains a unit root, then it will be invalid and will not be analysed in the further stages.