Stationarity
A property of a time series where its statistical properties (mean, variance, autocorrelation) remain constant over time.
What is Stationarity?
A stationary time series is one whose fundamental data-generating process does not change depending on the time at which the series is observed. It has no long-term trend and no seasonal variations.
Why Stationarity Matters
Most standard time series forecasting models (like ARIMA) and econometric analyses mathematically require the data to be stationary. Analyzing non-stationary data often leads to spurious regressions (finding a false relationship simply because both variables are trending upward over time).
How to Interpret
Stationarity is typically tested using unit root tests, such as the Augmented Dickey-Fuller (ADF) test.
Example
The daily price of a stock is typically non-stationary (it trends upwards or downwards). However, the daily percentage change in the stock's price is often stationary.
Common Mistakes
- Running OLS regression on non-stationary variables without checking for cointegration, resulting in meaningless, highly significant but spurious results.