Monday, March 28, 2016

Follow academic research at your own risk


A newly published research piece studied out of sample and post publication return predictability for a number of variables that have been shown to predict cross-sectional stock returns. They ran tests over 97 relevant variables. The researchers found that most variables show declines in expected returns out of sample. On average, portfolio returns decline by 26% which serves as an upper bound of data mining effects. More importantly, the authors find that returns decline 58% post publication of the research on any variable. Assuming that we conservatively take the 26% decline from data mining away, the researchers believe there is a  32% decline in returns from traders who have taken advantage of this research or market mis-pricing. See "Does Academic Research Destroy Stock Return Predictability?"

Traders read the research that is conducted by academics and they exploit any pricing anomalies that are found in the research. After the research is widely disseminated, the excess returns are taken away by the actions of traders. If you read the research that looks good, don't expect to make any of the real money from it. There are plenty of others who are doing the same thing and the opportunity is diminished. Markets are efficient in that if there are identified mis-pricings, the returns will be significantly bid away. You cannot get a free lunch from piggy-bacing on the research of academics. If you read academic research, read it early in the process. By the time it gets published it is old news.
In some sense this is not surprising. By the time you read news in the Wall Street Journal, it may already be old news.

Sorry folks, you will have to do your own research and keep it to yourself. 

No comments: