Purchasing Power Parity - Essay Example

A new Eviews workfile is generated from the main menu of the Eviews by selecting File/New/Workfile, which opens up the create workfile dialogue box. Dated-regular frequency is chosen as the workfile structure type, frequency is chosen as monthly with the start date as 1990-1 and end date as 2011-03 as shown in fig1. The data from the excel is then imported by selecting the Proc/Import/Read Text-Lotus-Excel options of the main menu and choosing the dataset Data_Canada_PPP.xls. The number of series is entered as 3 in the Excel spreadsheet import dialogue box as shown in fig2.

The data is imported successfully and is verified with the original data in the excel sheet by opening the generated data as shown in fig 3. Fig 1 Generate New Workfile Fig 2 Enter the number of series of Data Fig 3 Verifying the imported data 2) Generating Real Exchange Rate qt Real Exchange Rate qt is obtained by the formula: qt = st – pt + pt* -------- 1 where st = log(Exchange_rate) -------- 2 pt = log(CPI_Can) -------- 3 pt* = log(CPI_US) -------- 4 The formulae 2 to 4 are first generated using the Genr option in the workfile.

The value of qt is then generated using the formula 1. The generated qt is shown in fig 3. e Fig 3 Value of qt 3) Plotting the graph (qt) Fig 4 shows the graphical view of series of values of qt. . This can be tested using: Interpretation from Graphical representation: A non-stationary series produces lines with definite upward and downward trend with the passage of time, whereas a stationary series does not produce any such lines. Observing the Correlogram or Autocorrelation function (ACF): For a stationary process, the ACF will decline to zero in a quicker fashion whereas for a non-stationary process, the ACF declines in a linear fashion.

From the graphical representation of real exchange rates between Canada and US shown in fig 4, it can be noted that the real exchange rate is likely to have some sort of random walk-up and walk-down pattern over the period of time. The presence of random walks indicates that the series qt seems to show non-stationarity in behaviour. However, the random walk does not show any increasing or decreasing trend. 4) Unit Test Root Non-stationarity of a process is characterized by the presence of unit root.

In order to test whether the process is stationary or not, it would suffice if we can check for the presence of unit root. This check can be performed by employing Augmented Dickey-Fuller’s test. The overall objective of this test is to test the null hypothesis that ? = 1 in: yt = ? yt-1 + ? + ut against the one-sided alternative ? < 1. So we define the hypothesis as H0 : yt = yt-1 + ? + ut (qt is non-stationary, ? =1) H1 : yt = ? yt-1 + ? + ut (qt is stationary. ? < 1) Subtracting the above equation with yt-1, we get the simplified equation as: ? yt = ? yt-1 + ?

+ ut Where ? = ?-1. Now the hypothesis for the presense unit roots can be written as: H0 : ? =0 (qt contains a unit root and is