基于非参数日内跳跃检验和高频数据的公司信息披露对股市价格波动影响研究

本文以2012年1月4日至2013年12月31日之间的共481个交易日作为样本期间，以样本期间上交所发布的“上证180”成分股中的上市公司的经营公告、财务报告及证券分析师根据上述信息披露的股评三种信息为主要研究对象，从验证方法选择、高频数据选取、信息考察窗口优化及基于面板数据多元Logistics回归模型构建等四个方面将跳跃与不同信息相联系，分析股价波动与不同信息披露的关系。研究结果表明，当信息范围为公司特定的经营公告、财务报告及分析师建议时，经营公告是最具影响力的信息披露渠道，而分析师建议并不是引起股价异常波动最重要的信息。同时，本文研究揭示仅有20%的跳跃与此类信息披露相关，当解释变量覆盖代表宏观信息“系统性事件”和行业、板块信息的“行业事件”时，也仅40%的价格跳跃发生和信息披露有关。本文的研究不仅表明哪一种信息可能更具有投资价值，而且揭示在此研究基础上继续探究引起股价异常波动的其他起因事件可能更具有重要意义。

A Study of Firm Specific Information Disclosure on the Price Variation with Nonparametric Intraday Jumps Detection in High Frequency Data

Using an intraday LM non-parameter jump detection technique in a multivariate framework, firm-specific information disclosure and its impact on stock price reaction are explored in this paper. The three common avenues are linked for information disclosure:analyst recommendations, earnings announcements, and management guidance with jump in price, the sample consists of stocks listed on the Shanghai 180 index for the two year period between January 2012 and December 2013. After controlling for confounding events and through multivariate logistic regression based on two different models, empirical research indicates that management guidance are more likely than either earnings announcements or analyst recommendations to cause a stock price jump, and analyst recommendations are not so appear informative in China. Meanwhile, although the paper can reveal that relative importance of recommendations, earnings announcements and guidance to investment decision, but those information collectively explain only 20% of all firm-specific jumps calculated using the nonparametric intraday jumps detection in high frequency data. Even considering the impact of macroeconomic news releases or industry information on volatility and prices, the rate only increase to 40%. Thus, more research needs to be done to identify the types of other unknown events that cause jumps in the market.