Question

econometrics. what are the potential implications of including stationary variables as both dependant and independant variables in a regression model. 4 marks

describing how it might affect forecasting.

describing the potential confusion causes by spurious regressions.

Answer 1

If a time series is stationary,it's mean, variance and autocovariance (at lags ) remain the same no matter at what point we measure them;that is, they are time invariant.lf any of these conditions are violated, time series is non-stationary.ln other words non-stationary time series will have a time varying mean or time-varying variance or both.

Including non-stationary variables in the regression model will lead to the following consequences.

* if time series isn't stationary,we can study it's behaviour only for the time period under consideration.

* Each set of time series data will be for a purticular episode.

* It's not possible to generalize it to other time periods.Therefor the purpose of forecasting will be of little value .

* Another disadvantage is the emergence of spurious regression or non-sensical regression . It was discovered by Yule.

* Yule shouwed that spurious correlation could persist in non-stationary time series even if the sample is very large . If coefficient of determination is greater than d statistic , we should suspect the existence of spurious regression.

* Coefficient of determination as well as the t statistic from spurious regression will be misleading.