ABSTRACT
The concept of stationarity is of crucial importance in modern time series econometrics. If variables of interest are found to be stationary then it is meaningful to include them in a regression model for the purpose of short- and long-run analysis using the cointegration methodology. On the other hand, if one or more of the variables are found to be non-stationary, their inclusion into a regression model typically results in spurious regression and faulty analysis. It is therefore important to first test for the stationarity of time series variables to establish their status. This testing procedure is carried out via the unit root testing methodology, which constitutes a key stage in modern time series econometrics. This chapter provides an introduction to the unit root testing methodology, using a number of examples to illustrate the applications of key procedures in practice. Key topics
Dickey-Fuller unit root tests
Problems with the unit root tests
