我国IPO市场的周期性行为研究:1994-2010

摘要

This paper focuses on describing characteristics of time-yarying volatility in hot and cold IPO cycles by establishing Markov regime switching-GARCH(MRS-GARCH) model, adopting quantile regression for count data (QR-CD) model to study the relationship between numbers of IPO offered and duration from offering to listing. The empirical results taking China' s A Share IPO samples from January 1994 to August 2010 are that; Average initial-day returns and duration from offering to listing are incentives of numbers of IPO offered, which results in phase difference of periodic behavior among these three main variables used for describing IPO behaviors; the decreasing of duration from offering to listing has a positive effect on numbers of IPO offered; Besides that Volatility persistence of numbers of IPO offered is strong in cold cycles, and the volatility persistences of other variables are weaker in both cold and hot cycles; for(adjusted)average initial day returns, numbers of IPO offered, volatility variances are greater in hot status than ones in cold cycles, while duration from offering to listing is opposite. The evolution of China' s A share IPO market cycles is dominated by issuing policies, macroeconomic factors and investors' sentiment.%对IPO冷热周期内的时变波动特征首次建立区制转换GARCH模型加以有效描述,以及用计数数据分位数回归模型研究发行数量和发行到上市持续期的关系.通对1994年1月至2010年8月间我国A股IPO市场新股的实证研究得到若干新结论:上市首日收益率和发行到上市的持续期是发行数量的诱因,导致这三个描述IPO行为主要变量的周期行为之间存在相位差;发行到上市的持续期的降低对发行数量的增加有积极地促进作用;除了发行数量在淡季状态波动持续性较强外,其他变量在淡旺季的波动持续性均较弱;旺季时(调整后的)上市首日收益率、发行数量的波动方差比淡季时大,而发行到上市的持续期则相反;我国A股IPO发行市场的淡旺季周期的形成主要受到IPO发行政策、宏观经济环境、投资者情绪的影响控制.