基于隐藏马尔可夫模型的证券市场波动分析

Analysis of Stock Market Fluctuation Based on Hidden Markov Model

文章从证券市场股票价格的波动结构转变出发,采用隐藏马尔可夫状态转移模型来分析2014-2017年沪深300指数每日收盘价的波动特征,同时建立AR模型和GARCH模型进行对比分析。模型结果表明,沪深300指数收盘价的波动确实存在真实的转移特征,即低风险波动、中等风险波动以及高风险波动三种状态,且各个状态的平均持续时间各不相同。比较三种模型的拟合结果,发现隐藏马尔科夫模型优于另外两种模型,说明其可以更好的刻画证券市场波动的规律,为投资者及相关研究者提供了一种新的思路。

Starting from the change in the volatility structure of stock prices in the securities market,the article uses the Hidden Markov State Transfer Model to analyze the volatility of the daily closing price of the CSI 300 Index during the 2014-2017 period.The model results show that the fluctuation of the closing price of the Shanghai and Shenzhen 300 Index does have real transfer characteristics,namely low-risk fluctuations,medium-risk fluctuations and high-risk fluctuations.By calculating the transition probability matrix between each fluctuation state and the average duration of each state,we find that stock price fluctuations will continue to maintain the volatility of the previous day with extremely high probability,and the average time of various states will be different.The high volatility state,such as the stock market,maintained the longest average such as the circumstances in 2015,but the probability of this phenomenon is extremely small.Hidden Markov model can better describe the law of volatility in the securities market,providing investors and related researchers with a new idea.