基于区制转换的股市泡沫动态监测模型

提出了两阶段区制转换模型,考虑了在两个时区之间的随机转换和退出泡沫时期的概率,使用序贯贝叶斯学习方法实时估计模型状态和参数。模拟试验表明,相比现有的其他方法,两阶段区制转换模型对于检测泡沫来说具有更好的效力。实证检验证实了模型的优势,并且得到一个重要的结论是2008年之后的A股市场的泡沫化程度较低。本文模型可以为机构投资者和监管部门动态监测股票市场可能存在的价格泡沫现象提供一些启示。

Dynamic Monitoring Model of Stock Market Bubble Based on Regime Switching

A two-stage zonal conversion model is proposed,and the model takes into account the probability of a random transition between two regime and the probability of exiting the bubble.This paper uses sequential Bayesian learning method to estimate model state and parameters in real time.Simulation tests shows that,that the two-stage rgime-based transformation model is more effective in detecting bubbles than other methods.Empirical analysis based on actual data of SHCI confirms the advantages of the proposed model.The empirical test confirms the advantages of the model and gets an important conclusion that the bubble of A share market after 2008 is relatively low.This model can provide some inspiration for institutional investors and regulators to monitor the possible price bubbles in stock market.