新股收益率波动分析——基于创业板成立初期股票数据
Analysis of IPO Return Volatility—Based on Initial Stock Data from the ChiNext Market

2009年10月30日，中国创业板市场正式成立。但创业板的新股发行首日收益(IPO抑价率)，并不能从本质上反映创业板IPO的定价效率。虽然创业板新股首日收益率的波动性受人为影响因素干扰较多，但随着时间的推进，价格会逐渐反应股票本身的价值。为此，在充分理解和借鉴传统研究方法的基础上，本文决定延长创业板新股数据的采集区间，通过研究创业板首月(21个交易日)收益率的波动性，力图实现最大可能地降低政策规定在创业板新股首日收益率对于IPO定价效率的替代性方面的影响，从而更加全面准确地反映创业板新股初始收益率情况。首先描绘了创业板2010年6月至2018年12月历史走势图。其次，利用统计方法计算出新股票IPO的首月平均初始收益率与标准差。再次，利用相关数据来寻找创业板新股首月收益率的平均值与波动性两者之间的相关关系。然后，运用ARMA模型并结合EGARCH模型来探究时间序列变化对创业板IPO初始的收益率和对其分布的影响情况。最后，本文对发展创业板市场提出一些政策建议。
On October 30, 2009, the ChiNext market was officially established in China. However, the first-day returns of IPOs on ChiNext (IPO under pricing) do not inherently reflect the pricing efficiency of these IPOs. While the volatility of first-day returns for new stocks on ChiNext is significantly influenced by human factors, over time, prices tend to reflect the intrinsic value of the stocks. Therefore, this paper extends the data collection period for new stocks on ChiNext beyond the initial day. By analyzing the volatility of the first-month returns (21 trading days), the aim is to minimize the substitutive effect of regulatory policies on the first-day returns regarding IPO pricing efficiency, thereby providing a more comprehensive and accurate reflection of the initial returns of new stocks on ChiNext. Firstly, this paper depicts the historical trend of ChiNext from June 2010 to December 2018. Secondly, statistical methods are used to calculate the average initial returns and standard deviation of new stock IPOs for the first month. Thirdly, the paper explores the correlation between the average first-month returns and their volatility. Furthermore, an ARMA model combined with an EGARCH model is employed to investigate the impact of time series changes on the initial returns and their distribution for ChiNext IPOs. Finally, the paper proposes some policy recommendations for the development of the ChiNext market.