Tuesday, January 04, 2005

Han and Wang--Constraints, Efficiency, and Profits from Momentum Investing

What a great way to kick off the new year! This one definitely qualifies as I^3 (interesting, informative, and important)

Super short version:

Han and Wang have an interesting paper that examines the “impact of institutional investment constraints.” Their finding (that these constraints play an important role in the pricing of securities), gives us a much better insight into both institutional behavior as well as momentum investing.

Slightly longer version:

We have seen repeatedly now that things that limit arbitrage (for instance short sale constraints, high transaction costs etc.), limit market efficiency. This paper by Han and Wang is no expection. However, it looks at self-imposed contraints.

Lakonishok, Shleifer and Vishny (1997) ; Chan, Chen, and Lakonishok (2002); and Cohen, Gompers and Voluteenaho (2003) have all shown that institutional investors tend to not take positions significantly different from their benchmark portfolios. This may be because of a reluctance to accept non-systematic risk or incentives induced by contracting and/or agency costs.

In this paper Han and Wang investigate whether this behavior has consequences on market pricing and market efficiency. (In other words, they are testing whether these constraints—which are largely self-imposed—lead to predictable differences in the financial markets.) And lo and behold, the authors find that the constraints do matter!

“Using quarterly data on institutional equity holding between 1980 and 2001, we find that institution investors appear to be constrained from buying stocks that they already overweight and selling stocks they underweight. Such demand distortion may lead to price inefficiency for the stocks affected by these institutional investment constraints and generate cross-sectional return predictability as the mispricings get corrected.”

In simpler terms: if institutions already have a large position in a stock (i.e. overweighted), they are less likely to buy more, even if the stock price is going up. And vice verse—they are less likely to sell shares that they already underweight, even if the stock experiences more bad news.

Why is this important? Several reasons:

1. It shows again that where there are imperfections (even self imposed) that markets are less than perfectly efficient.
2. It shows that self-imposed constraints can be quite binding and that they do lead to predictable differences in the asset market.
3. It gives us a much better understanding of how momentum investing can be allowed to continue. In the author’s words:

“...we find that the momentum strategy which invests in the winner stocks that the institutions overweight and the loser stocks that institutions underweight is significantly more profitable than the simple momentum strategy that invests in all winners and losers. By contrast, buying the winners that institutions underweight and selling the losers that institutions overweight do not generate significant profits.”

Wow! You may want to reread that last quote! If true (and they have pretty much convinced me), they have given us the best proof yet as to why some researchers have been able to find abnormal profits from momentum investing.

Given these far reaching implications the time that Han and Wang spend looking at various alternative explanations is well spent. They find these other explanations all wanting in some important area. Thus, it appears that the constraint hypothesis is the best explanation.


The paper is also available through SSRN: Han, Bing NMI1 and Wang, Qinghai, "Institutional Investment Constraints and Market Efficiency" (October 2004). Dice Center Working Paper No. 2004-24. http://ssrn.com/abstract=628683

and yes Han is a former PSUer! (is there such a word? there is now! ;) )

No comments: