On the ergodicity of economic processes represented by time series

The article presents a method for testing the weak stationarity of a random process by means of its time series, which comes down to calculating the mean and the autocorrelation function over segments of the original series. Novelty of the presented procedures consists in estimating the stationarity of these values. The well-known theoretical criterion of ergodicity of a random process is adapted for practical application. The presented methods have been tested on a model example of a weakly ergodic random process and on a time series of the stock quotes of «Inarktika» company.