摘要
研究用马尔科夫链蒙特卡罗(MCMC)算法估计随机波动模型的参数问题.基于“前向滤波,后向抽样”方法提出一种新算法,并将新算法同原有算法进行了比较.然后利用新算法对上海股市进行波动性分析,发现中国涨跌停板制度对波动的持续性估计有着重要的影响,忽视这些因素将会导致波动的持续性被高估.
In this paper, A new markov chain monte carlo algorithm for estimating stochastic volatility model is given. And the new algorithm is compared with old algorithms. Then authors apply stochastic volatility model to analyze the volatility of Shanghai stock market by the new algorithm. Empirical results indicate that stochastic volatility model performs well. On the other hand empirical results also indicate that price limit have much effect on the persistence of volatility.
出处
《系统工程理论与实践》
EI
CSCD
北大核心
2006年第4期27-31,共5页
Systems Engineering-Theory & Practice