Nonlinear behavior of the Chinese SSEC index with a unit root: Evidence from threshold unit root tests

TL;DR

This paper examines the nonlinear dynamics of the Chinese SSEC index from 1990 to 2007 using a threshold autoregressive model, revealing two regimes with unit roots, which has implications for understanding stock market behavior.

Contribution

It applies a two-regime threshold unit root test to Chinese stock data, demonstrating nonlinear behavior and regime-dependent unit roots in the SSEC index.

Findings

The SSEC index exhibits nonlinear behavior with two regimes.

Both regimes show the presence of unit roots.

Threshold effects significantly influence stock market dynamics.

Abstract

We investigate the behavior of the Shanghai Stock Exchange Composite (SSEC) index for the period from 1990:12 to 2007:06 using an unconstrained two-regime threshold autoregressive (TAR) model with an unit root developed by Caner and Hansen. The method allows us to simultaneously consider non-stationarity and nonlinearity in financial time series. Our finding indicates that the Shanghai stock market exhibits nonlinear behavior with two regimes and has unit roots in both regimes. The important implications of the threshold effect in stock markets are also discussed.