SSRN-Market Reaction to Annual Earnings Announcements: The Case of Euronext Paris by Louhichi Wael: "Market Reaction to Annual Earnings Announcements: The Case of Euronext Paris "
As I am considering doing a paper on event studies, I have been looking at a few event study papers of late. In this research I stumbled upon this paper by Louhichi Wael. Wael looks at abnormal returns following overnight earnings announcements of French firms. The findings give several insights into market efficiency.
Probably the most convincing aspect of Wael’s paper is that the stock price moves on new information. While that is obvious, it is interesting to see exactly how this price change occurs. For instance the stock price change happens for both good and bad earnings announcements, but not for earnings that are "in line" with analyst forecasts. This is obviously consistent with semi-strong form efficiency.
There are several interesting things about this paper
1. The paper uses an event study methodology but uses minutes instead of days or months as the time period.
2. The author finds that for good earnings announcements (those above analysts’ expectations) there are on average no abnormal returns after the first 15 minutes of trading. However, even within this 15 minute window it would be difficult to make large returns as approximately 55% of the 1.74% positive excess return occurs on the first trade following the announcement, and a full 95% occurs within the first 15 minutes of trading.
The reaction for bad announcements is less pronounced. For bad earnings announcements, the firms experience a 1.04% drop for the day. However, this is a smaller drop than occurs on average after the first 30 minutes of trading where the stock price tends to bottom out at -1.28%. This is evidence of a slight overreaction that occurs within the first 30 minutes of trading following bad earnings announcements.
3. Bid-Ask spreads increase immediately after the announcement. The spreads and volume return to normal more quickly (within 15 minutes) for good news.
4. Volume is unusually high before and after the announcements.
5. Actual "price volatility remains abnormally high thirty minutes following the announcement of good news and fifty five minutes after bad news."
All in all some pretty convincing evidence that while the market is not perfectly efficient, it is pretty good (and fast) at incorporating new information.
Cite: Wael, Louhichi, "Market Reaction to Annual Earnings Announcements: The Case of Euronext Paris" (January 2004). EFMA 2004 Basel Meetings Paper. http://ssrn.com/abstract=498502