Tuesday, July 2, 2019

Academics help make markets efficient - Unfortunately, mining research is no free lunch

Who says nobody reads academic research? A close examination of academic research in finance shows that new factors that predict cross-sectional stock returns identified by academics fall substantial once the research is presented as a working paper or published. As soon as academics find these new predictors, investor are reading the research, weaponizing the idea, and generating returns. 

This active employment of capital, however, actually shrinks the measured predictor returns once measured out-of-sample and post publication. This shrinkage suggests that many of these anomalies are associated with both mispricing and the flow of capital diminishing any excess return that investors may receive. 

Mclean and Pontiff in their carefully researched paper "Does Academic Research Destroy Stock Return Predictability" study just under 100 variables that seem to show cross-sectional stock predictability. The returns fall just over 25 percent out-of-sample and over 50 percent lower post-publication. It is also notable that the return declines are greatest for those that had the largest in-sample returns and low liquidity. These in-sample returns may represent statistical bias, mispricing, or more likely the process of markets becoming efficient.

These identified factors may not be compensation for a specific risk but inefficiencies that are eliminated once investor identify and commit capital. This is just another example of market efficiency and the strongly competitive market of finance. If excess returns continue, some these returns are associated with the limits to arbitrage. Investors find opportunities, learn how to exploit them, and then eliminate them when possible. 

While this research is a few years old, it highlights the importance of careful analysis when investing new research ideas from academics or Wall Street firms. There should be some expected excess return discount for new ideas that get in the public domain and reach a wider audience. Perhaps there are some cheap lunches, but there are no free lunches.   

No comments: