基于STAR模型的中国股市泡沫检验

Chinese Stock Market Bubbles Test Based on STAR Model

中国股市周期性出现的暴涨暴跌现象可能蕴涵着周期性破灭泡沫的成分,但是中国学术界关于股市泡沫的争论非常不利于政策的制定和实施。股市泡沫研究方法的差异是引发众多争论的主要原因之一。股市泡沫演化过程中的非线性、非对称性运动特征使得传统线性计量检验方法在检验股市泡沫时存在较大的偏差。通过平滑门限自回归模型(STAR)对中国股票市场的检验发现,以logistic函数作为转换函数的STAR模型能很好地描述中国股市泡沫演化的非对称、非线性运动特征,这些特征符合周期性破灭泡沫模型。研究结果证实中国股市存在周期性破灭泡沫,政府应从短期和长期两个视角对股市泡沫进行治理。

Big stock price movements appear in Chinese stock market cyclical likely contains periodically collapsing bubbles. But the academic debate on the stock market bubbles is not conducive to policy formulation and implementation. Many different research methods of stock market bubbles cause of many arguments. Nonlinear and non-symmetric evolution of bubbles in Chinese stock market often tend to make traditional testing methods invalid. We propose to use smooth threshold autoregressive model （STAR） on Chinese periodically collapsing bubbles. The result shows the stock price presents nonlinear and non-symmetric characteristic, which characteristics of bubbles are in line with model of periodically collapsing bubbles. So we can conclude that there’re periodically collapsing bubbles in China stock market. Basing on above empirical results, this paper proposes policy recommendations to reduce and prevent hazards of stock bubbles.