A reliable way to spoil a market-beating investment strategy is to publish an academic paper about it, according to research by David McLean, a visiting professor at MIT Sloan School of Management.
Prof McLean and Jeffrey Pontiff, a professor of finance at Boston College’s Carroll School, examined 82 strategies that were shown to predict stock returns in 68 papers published in finance, accounting and economics journals*. They found that once these papers were published, the average strategy’s return fell by more than 30 per cent. This may be due to these papers attracting the notice of sophisticated quantitative investors who trade on a given strategy, which corrects the market mispricing that made it profitable in the first place. Prof McLean discusses his research:
What sparked your interest?
We had wondered if the investment strategies outlined in academic papers worked outside their original samples. It was an interesting question, but it was a daunting prospect to replicate so many strategies. During the summer of 2011 we did [the empirical work for] about a dozen of the strategies. We saw that the predictability was a lot weaker from the original sample to post-publication.
And what did you discover?
We found that the average strategy’s return declines by 35 per cent after an academic paper has been published. In other words, if you are an investor trading on a strategy outlined in an academic paper that promises an additional return of 5 per cent per year, you should expect to get only 3.75 per cent in the years after that paper is published.
What explains this effect?
When an academic publishes a paper about a strategy, investors learn from it and trade on that strategy. This trading impacts prices, bringing them more in line with fundamental values. This process also makes the strategy less profitable. It seems the strategies outlined in these papers were right for their original sample, but once practitioners learnt about them the predictability weakens.
Which investment strategies saw the greatest post-publication decline?
The returns of strategies that involve larger, more liquid stocks declined by more than 35 per cent. This is probably because investors are more likely to act on a published strategy if the cost of trading in the strategy is low. But when it’s more expensive to trade – smaller, more volatile stocks – investors are less apt to try a new strategy. We found that these costlier strategies keep on getting high, abnormal returns even after the paper [they are outlined in] is published.
Are your finance colleagues dispirited by your findings?
There’s nothing that makes the profession look bad. It suggests that academic research affects the real world.
Do they ever think, what’s the point?
Academics are interested in testing market efficiency and predicting stock returns is one way to do that. The point of these papers is not to claim ‘Hey here’s this great way to make a lot of money’. I think this research shows that markets want to be efficient, but there are costs and risks associated with getting there, so the market doesn’t end up in a perfectly efficient place.
You say there’s a silver lining to your research. Explain.
Our interpretation of these results is that academic research makes financial markets work better. Once the papers were published, the strategies did not work as well, suggesting that the market mispricing has been at least partially corrected. As a result, stock prices are more accurate estimates of what companies are worth. You prefer to live in a world where prices mean something.
*Does Academic Research Destroy Stock Return Predictability? Social Science Research Network.