齐次及非齐次马氏域变模型在股价泡沫检验中的应用  

Homogeneous and Non-homogeneous Markov Switching Models Applied in Stock Price Bubbles Test

作　　者：孟庆斌[1] 靳晓婷[2] 吴蕾[2] 

机构地区：[1]中国人民大学商学院 [2]南开大学经济学院

出　　处：《数量经济技术经济研究》2011年第4期124-136,F0003,共14页Journal of Quantitative & Technological Economics

基　　金：教育部人文社科青年基金(编号10YJC790196);中国人民大学青年基金(编号2010030125);南开大学人文社会科学校内青年项目(编号NKQ09017)阶段性成果

摘　　要：本文建立非齐次马氏域变模型检验股市价格泡沫,并结合对1996年1月至2010年6月间我国股市价格泡沫程度的实证度量,与齐次马氏域变模型的检验效果进行比较。两模型都能在一定程度上反映我国股票市场价格在各时间段的不同泡沫水平,但非齐次马氏域变模型所得到有泡沫概率和无泡沫概率区分度更高,比齐次马氏域变模型能更加精确的区分各时点上的泡沫水平。实证结果也表明,近期我国应对金融危机的超宽松经济政策对资本市场的刺激作用较为直接,而对实体经济的作用相对滞后,股市泡沫水平再度上升。To use non-homogeneous Markov Switching models, putting the economic variables into the state-switching probability function, we extend the tes- ting methods of price bubbles in stock markets. We also compare the effectiveness of homogeneous and non - homogeneous Markov Switching models by testing the bubbles of Shanghai - securities Index between January 1996 and June 2010 in China. The result shows that non- homogeneous Markov Switching models are better in distinguishing the bubble states in different periods and in distinguis- hing probability of bubbles states and non bubble states although both kinds of models can reflect the bubble levels. Meanwhile, as the financial crisis recently, the unconventional expansionary monetary policy stimulated capital market more directly than the real economy, and bubble level of Chinese stock price grew obviously.

关 键 词：齐次马氏域变模型 非齐次马氏域变模型 股市价格泡沫 

分 类 号：F222.3[经济管理—国民经济]