portfolio manager skill: is coin flipping a good analogy?

A few days ago I posted about the equity market views of well-known portfolio manager Bill Miller.  I knew that he had been  suffering through hard times after having beaten the S&P 500 every year for over a decade.   Still, when I looked up his recent record will writing the earlier post, I was taken aback by the extent of his underperformance over the  past several years.

Finance professors would have no trouble explaining this development.  In fact, as efficient markets adherents they might relish the prospect of talking about the stumbles of a renowned practitioner.

The standard academic argument runs as follows:

Take a large number of people (say, 1024) and put them all in a (big) room together, each armed with a coin to flip.  Start them all flipping their coins together on command.  At the end of round one, 512 of them will have flipped heads.  At the end of round two, 256 will have flipped two heads in a row.  At the end of round three, 128 will have flipped three heads in a row….At the end of round ten, 1 lucky individual will have flipped ten heads in a row.

Is he a coin-flipping genius?  No.  He’s just been lucky.  He’s the one in a thousand twenty-four that the laws of chance predict there will be.

Conclusion:  that’s all the “skill” of apparently successful investment professionals is–dumb luck, plus their clients’ lack of knowledge of basic facts about probability.

As an explanation, this is, of course, ludicrous.  It might be the basis of a good joke if the academic community didn’t actually believe it is a valid explanation.

Let’s try it out on AAPL.

AAPL was on the verge of bankruptcy when Steve Jobs was rehired.  But Steve is not a superior manager.  It was just dumb luck  that he launched the iPod.  Then he accidentally started opening Apple Stores, which enhanced the image of Apple and provided another distribution network for APPL products.  Then, through more dumb luck he created the iPhone, which doubled the size of the (now larger) company again.  After that, through even more luck the company has begun to sell the iPad.

Maybe Barry Bonds, or Mark McGuire or Sammy Sosa could hire a few finance professors to improve their public images.  The academics could explain that they didn’t grow those big bodies and hit all those prodigious home runs because they were juiced on steroids.  Instead, it was just but  dumb luck that they happened to repeatedly place their bats in just the right position to hit the pitched ball out of the park.  They had as little to do with the home runs as the coin flipper who repeatedly tosses heads had to do with his streak.

Nevertheless, if Mr. Miller is someone of unusually high skill in equity investing, how can we explain his recent extended fall from grace?

This is what I think happens:

Let’s say a  good professional investor spots an unusually attractive group of stocks or industries in advance of almost everyone else.  Since most US portfolio managers are former securities analysts, who will have studied one or two industry groups in great depth, it may be the undervaluation of the stocks he formerly covered that he sees.   For Warren Buffett, it was the value of intangible assets.  For Peter Lynch, it was the relative cheapness of mid- and small-cap issues vs. their larger counterparts.

In any event, it’s something. As time passes after our professional buys the stocks and his idea begins to play out, the investor’s rivals notice his outperformance, divine its probable source by studying the original investor’s portfolio, and begin to copy what he’s doing.

For mutual funds or ETFs, figuring out  portfolio structure is easier than it sounds.  The funds file lists of their holdings periodically with the SEC.  In addition, simply watching the daily changes in net asset value can reveal much more information than you’d think.   Some fund groups have had analytic tools to do this for at least twenty years.  The SEC filings will simply confirm the conclusions arrived at in-house.

For purely pension fund-oriented investment firms, the task is a bit harder.  But marketing his services to institutional clients will require that the original investor reveal a good deal about his strategy.  And, of course, there will be inevitable leakage from the brokers who do his trading.

Eventually, the misvaluation that the first portfolio manager sees will be arbitraged away. But this takes a much longer time than academic theory would suggest.   For Peter Lynch, the process took ten years.  For Warren Buffett, it might have taken twenty.  For other concepts, like the restructuring of the US industrial base during the junk bond boom of the Eighties–which began to work in the bull market of the early Nineties, the outperformance period may have been two or three.

Many managers, like many growth companies, are one trick ponies, even though the “trick” can last for a considerable period of time.  The best, however, are not limited to finding a single valuation anomaly.  They may see the first one more or less by accident.   But from this experience they learn how to look and are able to find a second or third instance as well.

The very long period of time of Mr. Miller’s outperformance suggests to me that he is one of the second type and that his string of index-beating years is the result of spotting several different kinds of misvaluation (although I’d like to see at least a flash of his old form before even thinking of sending him money to manage).

Leave a Reply

%d bloggers like this: