TD_DICKEY_FULLER determines whether a set of time series data is stationary or nonstationary. It tests for the presence of a unit root in the data. The series being tested is a residual series, or any time or spatial series. When a series contains unit roots, then it is non-stationary. When a series contains no unit roots, the series may or may not be stationary, based on other factors involved in determining stationarity.
A time series is stationary if its statistical properties, such as mean, variance, and autocorrelation, do not change over time. A non-stationary time series has statistical properties that change over time, making it difficult to analyze and make predictions.
The test is based on the idea that if a time series has a unit root, it can be expressed as a random walk with drift. The test involves regressing the first difference of the series on its lagged values and testing for the significance of the coefficient on the lagged value.
If the coefficient is significant, then the null hypothesis of the test, which is the presence of a unit root, is rejected, and the series is stationary. If the coefficient is not significant, then the null hypothesis cannot be rejected, and the series is non-stationary. The test is used in econometrics, finance, and other fields where time series data is analyzed.
- Run TD_DICKEY_FULLER on the time series being modeled.
- Retrieve the results of the TD_DICKEY_FULLER test to determine if the time series contains any unit roots.
- [Unit roots are present] Use a technique such as differencing such as TD_DIFF or seasonal normalization, such as TD_SEASONALNORMALIZE, to create a new series, then rerun the TD_DICKEY_FULLER test to verify that the differenced or seasonally-normalized series unit root are removed.
- (Result shows series contains unit roots) Use TD_DIFF and TD_SEASONALNORMALIZE to remove unit roots.