Unit Root
A characteristic of a time series indicating that it is non-stationary and has statistical properties that change over time.
What is Unit Root?
A time series with a unit root shows a random walk pattern, meaning its mean and variance are not constant over time, and shocks to the system have permanent effects rather than decaying away. Statistical tests, such as the Augmented Dickey-Fuller (ADF) test, are used to detect the presence of a unit root.
Why Unit Root Matters
Running standard regressions on variables with unit roots can result in spurious regressions, where completely unrelated variables appear to be highly correlated simply because they both trend over time.
Example
An econometrician finds that GDP data has a unit root. To use it in a regression, they must first take the difference (current quarter minus previous quarter) to make the series stationary.
Common Mistakes
- Failing to test for unit roots before estimating regressions with macroeconomic time series.
- Over-differencing data that is already stationary, which introduces unnecessary noise.