马尔科夫转换-GARCH模型参数估计方法的研究和探讨
首发时间:2015-04-16
摘要:金融市场波动可能存在结构变化,经典的GARCH模型由于系数保持不变,不能反映该结构变化,使得波动的预测不够准确。本文将马尔科夫过程引入GARCH模型中,通过设置状态变量,马尔科夫转换-GARCH模型(MS-GARCH)更好的捕捉到波动特性。由于模型的路径依赖,极大释然估计将不适用,本文基于MCMC算法对MS-GARCH模型参数估计进行了研究和探讨。
关键词: 波动性;GARCH模型;MS-GARCH模型 Gibbs抽样
For information in English, please click here
Research on the parameter estimation of Markov-switching GARCH model
Abstract:The volatility of financial market may have structural changes, the coefficients of the classical GARCH are constant, and GARCH can't reflect the structural changes of volatility, so GARCH is not accurate enough for volatility forecast. This paper introduces a Markov-switching GARCH model (MS-GARCH) in which the parameters are allowed to switch over time, the switching is governed by an unobserved Markov chain. MS-GARCH may be better to capture the volatility characteristics. Because of path dependence, maximum likelihood estimation is not feasible. Parameter estimation of MS-GARCH model carried out a detailed study and discussion based on MCMC algorithms in this paper.
Keywords: volatility GARCH model MS-GARCH model Gibbs Sampling
基金:
论文图表:
引用
No.4637662104549214****
同行评议
共计0人参与
勘误表
马尔科夫转换-GARCH模型参数估计方法的研究和探讨
评论
全部评论