Wednesday, January 5, 2011

Betting Against The Shorts

A few weeks ago, potential sources of stock ideas were discussed on this site. The post made reference to contrarian indicators that signal that a stock is hated; this hatred can result in a low price, which value investors love. One such contrarian indicator is a stock's short interest.

Twice a month, the exchanges release how many shares of an issue are held short (with a two-week lag). This can be a helpful resource in not only identifying which stocks Mr. Market hates, but also which same stocks Mr. Market plans to buy later. This is because when a share is shorted, it must be bought back later; as such, a high number of short shares suggests a higher than normal demand for shares in the future.

Recently, the NYSE reported that short interest was lower than it has been in about a year. This suggests that sentiment is rather positive (at least by recent standards), which should sound a bit of a warning to investors. But information like this on aggregate market data is difficult to act on. Much more interesting to bottom-up investors is the short-interest level on individual stocks.

Highshortinterest.com offers a list of the most highly shorted stocks (relative to their floats) on the major US exchanges. It currently contains about 100 of the most hated stocks in the US, and plenty of stocks with value potential litter that list. For value investors looking for potential ideas worthy of further study, this is a decent place to start.

For example, currently at #4 on the list is Conn's Inc, which has come up a few times on this site. Further down the list are other names that have been bandied about on this site such as LodgeNet, DSW, and GameStop. Other names on the list which are currently in the news as potential value plays include Sears, Ebix, St. Joe, and Barnes and Noble.

When sentiment on a stock is extremely negative, value investors should pay attention. As such, one of the best places to look for stock ideas is the short interest list. Not only are these stocks unpopular and therefore cheap, these short shares will have to be bought back in the future.

4 comments:

Unknown said...

It is rare for me to disagree with you, but on this subject, I definitely disagree.

Obviously, shorts can and do get it wrong, but they tend to target firms with lots and lots of downside risk. To me, value investing is all about avoiding downside risk.

Another drawback to investing in heavily shorted companies is that they often have unique problems or muddy accounting, which increases the need to do lots of homework.

Empirically speaking, there also seems to be evidence that shorts are pretty good at what they do. For instance:

http://www.aiecon.org/advanced/suggestedreadings/PDF/sug322.pdf
http://kvenkataraman.cox.smu.edu/papers/dkv1.pdf
http://fisher.osu.edu/~diether_1/papers/sho.pdf

If I buy into a heavily shorted company, I require either a greater margin of safety, or a greater understanding of the company, than I would for a non-shorted company.

Saj Karsan said...

Hi Parker,

There are some good arguments made in your comment and in the papers you referred to. Papers 1 and 3 deal with short-interest versus returns, but from my understanding their return estimates are based on very short time periods (semi-monthly and monthly, respectively) so I'm not sure they are useful for long-term investors such as ourselves. Paper #2, which deals with short-selling versus accounting restatements, is certainly interesting.

Unknown said...

Thanks for the reply and consideration.

Based on my studies, it looks like just about every contrarian indicator indicates value (at least on a statistical basis). This includes poor historical price movement, low price to earnings/book/sales/cashflow, low earnings growth, low institutional ownership, smaller market cap, low actual stock price, etc, etc.

For some reason, high short interest seems to be the only exception to this contrarian=value rule.

More evidence (that's not so short term).
http://www.en.atfin.ru/Aretr.pdf

From page 13 of this study:

"Future abnormal returns decline monotonically with the level of short interest. For firms with no short positions, the average one-year ahead abnormal return is 2.3 percent, while for firms with over five percent shorted, the average abnormal return falls to –18.1 percent. For each of the categories with short positions the average abnormal return is significantly lower than the average abnormal return for the firm-years with no short positions."

I'm the only person I know 'in real life' who would read research like this for fun, so its enjoyable to discuss it with people like you. Thanks for blogging :)

Saj Karsan said...

Hi Parker,

Thanks, that is an interesting and comprehensive study. The authors note that shorts participate more often in "low fundamentals to price" situations (e.g. low B/P and low E/P). Could that be the main factor behind the low returns on heavily shorted stocks, the fact that they are generally companies with expensive ratios? What happens in the rare case where there are a high amount of shorts on companies with cheap ratios?