Procedury testowania w badaniu stacjonarności szeregów czasowych procesów ekonomicznych

Abstract

The aim of this paper is to present some useful techniques of stationarity analysis of time — series concerning economic processes. According to the first technique time — series is divided into two parts and then t-test and F-test are used for each of them. If a zero-hypothesis of statistically nonessential differences between compared parameters accepted, stationarity of time — series is proved. It means that mean value and variance of this stochastic process are time — independent. ... According to the second proposed technique a hypothesis of consistency of distributions in both parts of time — serie is verified by means of χ2 — test and Kolomogorov-Smirnov-test. Testing procedures are based on the empirical data on market processes in the GDR