Investment Returns from the Long Tail and the Marginalization of Wall Street Research

A particular statistic has been burning a hole in my head since I first read it in Alan Abelson's Barron's column this weekend - according to Rich Bernstein, Chief Investment Strategist at Merrill Lynch, the best of 40 trading strategies tracked by the firm's quants in 2006 was to buy the 50 S&P 500 stocks least covered by the sell-side analyst community. Huh? The return of this strategy was 24.6%, fully 11% better than the 13.6% return generated by the S&P 500 for 2006. This isn't outperformance - this is an a**-kicking. But the real question is why, and what might the implications of this be?

Alan made a funny reference to the legendary investor Gerald Loeb, and his view on the usefulness of security analysts: "In bull markets, you don't need 'em; in bear markets, you don't want 'em." Now Gerald wasn't the most socially sensitive guy, but you kind of get his point. I am personally more curious as to why "orphan" stocks outperformed better-covered shares, because, on its face, one might find this counter-intuitive. Could it be that more information levels the playing field, rendering the ability to develop and act on truly differentiated information extremely difficult? Or that there is so much data, good and bad, around an actively covered share that it is hard to separate information from noise? Might it possibly be that a valuable piece of information on a lighly-covered stock has much greater marginal value, both because of lighter trading volumes (thereby pushing up the stock on a big buy program) and the resulting follow-through from momentum-based traders?

I'd argue that it is these and many other factors that can explain the difference in performance cited by Rich, but that it is hard to argue that the marginal value of information on a less-covered stock is apt to have a greater impact on price than that same piece of information on a more actively-covered stock. Why? Because the information is less expected; is less likely to be discovered before-hand; and the trading volumes are likely to be smaller. If one buys these arguments, and if one views the S&P 500 as a microcosm of the overall equity market, then it stands to reason that there is one very straight-forward way to generate outsized returns: get more and better information on poorly-covered stocks in the face of weak sell-side analyst support. But how?

By using powerful tools to look for "long-tail" information that can augment information that is generally available through conventional media outlets. My company and a handful of others have recognized this and are trying to bring this capability to the investment community. I am firmly convinced that we are entering the "Third Wave" of information for investors: with Reuters pioneering dissemination of news with undersea cables in the late 1800s; Bloomberg revolutionizing granular cross-market pricing data and analytics leveraging client server technology in the 1980s; and tools for extracting valuable and timely information from the Internet in the 2000s. And this Third Wave has the potential of being even bigger than the first two, principally due to the proliferation of content on a global basis finding its way onto the Web, and the increasingly high-quality sources contributing to the global dialogue on companies, products, and macro themes on a 24-7-365 basis.

This is extremely exciting stuff that is screaming for new tools and technologies. And we are very early in the evolution of the Third Wave. It is just when people like Rich Bernstein crunch the numbers and highlight the fascinating yet counterintuitive that it makes the inexorable trend towards more and better internet information increasingly clear.

Comments

[howard Lindzon]

great post and all true. Great quote from Loeb and really true - until Monitor 110 :)

Obviously you are on to something big.

[Bill aka NO DooDahs!]

Is +24.6% the best they could do? Wow. they need better quants!

PS - they need to include dividends on both sides of the comparison.

[EC]

I agree with Roger that this is likely to NOT be a small-cap phenomenon. If this were the 50 least followed companies period that would be one thing, but the 50 least followed S&P 500 stocks are not likely to be small one hit wonders, as implied by George.

Also, it helps to back up arguments with data. 2006 was a marked change from 2005 and "the trend" on the outperformance of small caps in my opinion.

http://tickersense.typepad.com/.shared/image.html?/photos/uncategorized/average_us_stock_performance.jpg

That's a chart of how the "market", chunked into 10 buckets, did in 2006. Note the lack of material outperformance of small caps this time around, and the fact that super large caps were the 2nd best performing segment of the market. So for 2006, small caps didn't outperform. Roger's point is more nuanced than that.

What *did* do well was distressed and high FCF yielding companies-- contrarian investments. I hate to make arguments without data, but my guess is a positive correlation to contrarian investments. My rationale would be the simple fact that investment research is, in my opinion, trend following. Their revenue model attracts them to whatever happens to be hot at the time. For S&P stocks this is especially true-- micro-cap stocks might be left off simply because of lack of liquidity, even if they're in a hot market, but with the S&P, the liquidity factor is fine and those stocks all have capacity. All those other residual factors aren't there for S&P stocks. On average, I tend to believe the trend following nature of investment research shuns contrarian stocks. Contrarian stocks did especially well, hence the correlation.

This would be an easy test to run because the data is readily available. Just take the past 10 years, grab the performance of small caps, the S&P, the 50 least covered stocks of the S&P, distressed stocks and high yielding stocks. Either the correlation is there or it isn't. Would be interested in the results. And in your opinion!

[Roger]

Yaser, I totally agree. Data is limited but the phenomenon is interesting nonetheless.

[Roger]

George, as with Yaser, I wouldn't call the Bernstein's analysis revealing of a small-cap phenomenon. The S&P 500 is not really comprised of small-cap stocks, which is the universe against which Rich and Merrill's analysis was performed. Thanks for the comment.

[Yaser Anwar]

Sir, I think there is a lack of info to judge why the 50 least covered stocks outperformed.

They could be from any industry and size. Could be a hybrid of small-caps mixed with business turnaround situations. Like Telcos and material companies stocks did the best in 06, so it could be some least covered stock in that universe.

The possibilities and probabilities are endless. I haven't read the Barrons article but if RB is just saying 50 least covered stocks, thats quite a broad classification IMO.

[George S.]

I think it's not only a matter of finding "better" information or uncovering secrets through the noise, but I think the business models of the underlying companies could simply have riskier return profiles.

I'm sure many of these less covered companies simply less covered by virtue of their smaller size. A smaller company with fewer product lines and diversified business ventures will have riskier cash flows because of this fact alone.

On a larger company such as Hershey, we don't care how each candy old and new is doing, but we care about the law of large numbers and how a large enough # of candies will do if they are executed fairly well. However on a small candy company, the idea better be right on their first candy bar or the entire candy company is going under.

Therefore, yes the news is fewer and farther between and there are less people making the market efficient leading up to the news event, but in the end even the CEO may not know what the next month holds for his tiny company with no sell-side research.

In review, I think this phenomenon is probably driven by risky ventures of smaller companies (as size is correlated with coverage) and not a lack of noise. The market is already discounting these guys a ridiculous amount (where you can pick up edge if you are risk neutral) because they hit home-runs or go down in flames.

[Roger]

You're right about small-caps, Yaser, but I wouldn't classify the 50 least covered S&P 500 stocks as small-cap. Not even close. So I don't think this explains the phenomenon Rich Bernstein identified.

[Yaser Anwar]

I'm assuming the stocks were small-caps? Small-caps usually outperform when the broader market does well.

Also, every now and then a small-cap company has an innovative product/solution which turns it from a small-cap to large-cap, much like Starbucks in 1992 when its market cap was $220 million. Today, its market cap is $23 billion.

This can be blamed partly due to lack of Wall St. coverage, since small-caps have thin floats and big mutual fund and other portfolio managers can't take big positions. Thus creating opportunity.

You should check out this book
http://www.findingthenextstarbucks.com