基于向量随机波动模型的股市收益关联研究

发布时间：2018-04-03 05:10

本文选题：随机波动模型　切入点：厚尾　出处：《长春工业大学》2017年硕士论文

【摘要】：为了研究股票市场的随机波动性,以及模拟股票收益的最大化。本论文梳理了国内外随机波动模型的研究成果,在以往的文献中发现传统的GARCH模型估计收益率序列通常表现出波动具有长记忆性特征和较高的方差持续性,这些特征可以由方差的结构性变点造成。本文在单变量随机波动模型的基础上,研究了阶段性随机波动模型,并且在总结股市收益研究方法的基础上,根据股市阶段性研究的需要提出检验股市收益转折点的检验方法,并且对上证股市和深成股市划分了阶段;根据研究需要建立了阶段性双变量SV模型;利用阶段性双变量SV模型模拟了我国上证和深成两个股市的收益和波动,在对比检验结果和我国股市实际阶段的基础上,得出了基于阶段性向量随机波动模型的一些实证结果。
[Abstract]:In order to study the stochastic volatility of stock market and simulate the maximization of stock returns.This paper reviews the research results of stochastic volatility models at home and abroad. It is found in previous literatures that the traditional GARCH models usually show long memory characteristics and high variance persistence.These characteristics can be caused by structural variability of variance.Based on the single-variable stochastic volatility model, this paper studies the stage stochastic volatility model, and on the basis of summarizing the stock market income research methods, according to the needs of the stock market stage research, puts forward the test method to test the stock market return turning point.And divided the Shanghai stock market and Shenzhen stock market into stages; according to the need to establish the stage of bivariate SV model; using the phased two-variable SV model to simulate the return and volatility of the Shanghai stock market and Shenzhen stock market.On the basis of comparing the test results with the actual stage of China's stock market, some empirical results based on the phase vector stochastic volatility model are obtained.
【学位授予单位】：长春工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F224;F832.51

【参考文献】