新股发行周期波动的Markov三区制转换模型研究

The Detection of A-share IPO Market Cycles Based on the Three Regimes Markov Switching Model

本文将传统的IPO周期市场"热销"和"冷发"两种状态拓展为"热销"、"冷发"和"过渡"三种状态,并结合稳健性考虑和我国IPO市场实际对测度IPO周期的变量进行了改进,构造了一套新的包含IPO数量、抑价、市场条件、政府调控四类代理变量的指标体系,然后将变量分别应用三区制Markov区制转换模型进行回归并通过滤波迭代法得到滤波概率和平滑概率,继而得出每组变量对应的IPO周期,最后通过综合对比获得了我国A股市场1994年1月至2012年6月的IPO周期划分结果。结果显示我国IPO发行周期波动存在"热销"、"冷发"和"过渡"三种状态,并刻画了IPO周期与IPO数量、抑价、市场条件、政府调控之间的关系,研究丰富了IPO周期理论,并有助于IPO发行的有效决策。

This paper expands the IPO market conditions to three states,which include hot periods,cold periods and interim periods,and improves the variables in detecting IPO market cycles considering cycles’ robustness and China’s IPO market’s reality.These improved variables can be divided into numbers of IPOs issued,IPO underpricing,market conditions and government regulation,which include ten different proxy variables.Then we use the Markov regime switching model（three regimes） to make regressions with proxy variables respectively.By analyzing the regression results,filtered probability,and smoothed probability,we can get ten kinds of IPO cycles corresponding to ten different proxy variables.Further,this paper points the results of IPO market cycles in China’s A-share market from January 1994 to June 2012.The results confirm the relevance between IPO market cycles and numbers of IPOs issued,IPO underpricing,market conditions and government regulation,and also assist IPO issuers and investors to make effective decisions to some extent.