题目：Good News, Bad News, and Market Efficiency
庄额嘉，武汉大学经济与管理学院特聘研究员，博士毕业于台湾大学财务与金融学系，研究领域为金融计量、风险理论。研究成果发表于 Journal of Econometrics, Insurance: Mathematics and Economics.
The efficient market hypothesis (EMH) assumes that new information is quickly and correctly reflected in its current security price (Lim and Brooks, 2011, Journal of Economic Surveys). This paper seeks to answer two important questions related to the EMH: how quickly and how correctly? If the impact of an event is absorbed by price changes of corresponding security, then return samples before the event and that after the event should not exhibit any dominance relationship in the sense of Hollifield and Kraus (2009, Management Science), which defines bad news as a change in return distribution that decreases both investors' expected utility and demand of security. This definition of bad news is further extended to conditional distribution to account for the fact that investors often have either public or private information to assist their decision making (Vega, 2006, Journal of Financial Economics), with a statistical test developed for practical validation. The test is then applied to 260 events in the U.S. stock market, containing earnings per share over/under-expectation, and Moody's change of credit ratings. It is interesting to find that the market does not respond to the events correctly for around 30% of all events, and does not respond to some events quickly enough, neither, with the inefficiencies usually lasting for a couple of days.