The data file var contains 100 observations on two generated series of data, (w) and (z). The
Question:
The data file var contains 100 observations on two generated series of data, \(w\) and \(z\). The variables are nonstationary but not cointegrated. Estimate a VAR model of changes in the variables. As a check, the results are (the intercept terms were not significant):
\[\begin{aligned}& \widehat{\Delta w_{t}}=0.743 \Delta w_{t-1}+0.214 \Delta z_{t-1} \\& (t)\quad(11.403) \\& \widehat{\Delta z_{t}}=-0.155 \Delta w_{t-1}+0.641 \Delta z_{t-1} \\& (t)\quad(-2.293)\end{aligned}\]
a. The residuals from the VAR model should not be autocorrelated. Is this the case?
b. Determine the impulse responses for the first two periods. (You may assume the special condition that there is no contemporaneous dependence.)
c. Determine the variance decompositions for the first two periods.
Step by Step Answer:
Principles Of Econometrics
ISBN: 9781118452271
5th Edition
Authors: R Carter Hill, William E Griffiths, Guay C Lim