Abstract
ARCH model is a kind of dynamic non-linear time series model. It reflects a special feature of economic variables time-varying variances. As a new theory, ARCH model has caused extensive interests of economists and has been developed very fast since it came into being. The class has GARCH unit root test of conventional ADF degree of influence how critical is to analyze seriously ADF test, and attention.. Because GARCH model, more financial data exist in the fluctuation of leverage and lag coefficient of the ARCH and GARCH, at the same time, to describe a lot of time sequence, NGARCH and EGARCH model can describe the financial time series, and NGARCH and EGARCH as freely though they and the unit root test - as freely though they concern.
In this text，using the Monte Carlo eviews software through random simulation in limited samples through weighted least-square method are discussed in NGARCH and EGARCH under the condition of unit root test, exploratory analysis for normal distribution error when the critical test statistics and actual and potential, and the level of distortion are lagging order number on the test statistics. In the cases of limited samples, through random simulation, the disturbance of conditional variances of time-varying unit root test.
Keywords： ARCH; Unit root test; Test statistics
