Part 7

Question 1

Cointegration meaning

Answer 1

Cointegration means that two or more time series are non stationary but a linear combination of them is stationary (has constant mean, variance and co variance). So equilibrium relationship remains constant over time

Question 2

Engle-Granger Approach drawbacks

Answer 2

1. Unit root and cointegration tests have low power in finite samples. May not detect cointergration
2. Variables treated asymmetrically: one specified as the dependent and the
   other as independent variables. Can imply causality that exists which affects cointergration equations

Question 3

Ljung box tests - calculate T stat

Answer 3

T (T+2) x …

Question 4

Ljung box tests - if t stat more than critical value (x^2 ones)

Answer 4

If the test statistic is greater than the critical value in the Ljung-Box test, then we reject the null hypothesis that the first three autocorrelation coefficients are jointly equal to zero, indicating that there is significant autocorrelation.

Question 5

Ljung box tests - if t stat less than critical value (x^2 ones)

Answer 5

we fail to reject null hypothesis that the first … autocorrelation co-efficients jointly are not significantly different from zero. So jointly insignificant

Question 6

Testing autocorrelations whether significant - how did you tell - outside what parameter (5% level)

Answer 6

If lags are greater than (-0.196…0.196), they are signifcant

Question 7

Testing stationarity quadratic

Answer 7

solve in book