基于马尔可夫域变模型的上海股市收益率非线性特征分析

An Empirical Study on the Non-linear Characteristic of Return of the Shanghai Stock Market Based on Markov Regimes Switching Model

针对股市收益率在不同时期内具有不同的均值、波动性和持续性等非线性特征,引入马尔可夫域变模型(MRSM)对上海股市收益率的均值与波动性的对应关系以及高、低收益率状态转换特征进行分析,结果表明马尔可夫域变模型与GARCH类模型相比较,显著地提高了对股票市场行为的描述能力。它不仅可以从动态角度明确刻画金融市场的"收益与风险"相对称的特征,而且可测定不同状态持续的可能性和由一种状态转向另一种状态的概率。

The rate of return of stock market has the non- linear characteristic with different means, volatility and duration in a different time; this paper analyses the symmetrical relation of average rate of return and volatility as well as the duration of high - and - low - yield state of the Shanghai stock market based on Markov Regimes Switching Model （MRSM）. The result indicates that Markov Regimes Switching Model has more remarkable description ability to stock market behavior than the GARCH model. MRSM can not only explicitly portray the characteristic that the return and the risk is symmetrical in financial market from dynamic perspective, but also can determine the probability of duration and the average duration of the different regime and the probability of switching from one regime to another.