Thursday, November 19, 2009

The Efficient Market Hypothesis


Here, Robert P. Murphy takes defenders of the Efficient Market Hypothesis to task. My students of finance and investing will find Murphy's arguments and his links to other articles particularly interesting.

Murphy's basic observation is that the EMH is a tautology. If that's so, then the EMH cannot fail to be true any more than 2+2=4 can fail to be true, or that Bob=Bob can fail to be true. Tautologies are true by definition --- true because of what we mean by the words or symbols we use to make the statement.

It is certainly true that market prices for financial securities incorporate "all relevant and available information." To the extent that the EMH said nothing more than this, it would be a fairly unremarkable statement. For what would it mean if market prices did not include all available and relevant information? It would mean that traders, investors, and money managers were buying or selling securities randomly; that they were making buying or selling decisions on what they, as individual buyers or sellers, took to be non-information or false information. That would be preposterous, of course.  Conclusion: Murphy is right. The EMH is abundantly true; it cannot fail to be true, given what we mean by "information."  The real question is, so what? What are the implications of the EMH being true?

What the EMH says beyond this very reasonable proposition is that financial markets are efficient in the sense that they incorporate all relevant and available information really quickly, and even if only a few participants in the market know and act on the information.   

Most defenders of the EMH conclude that if the EMH is true, then an individual investor cannot "beat the market" consistently over the long run (here, the term “long run” does not focus mainly on calendar time, but on repetition; beating the market over and over, reliably and consistently). What does "beat the market" mean? It means to earn a risk-adjusted rate of return that is higher than the rate of return experienced by an appropriate risk-matching index of the broad market. In the terminology of stock traders and money managers and Wall Street, it means earning “alpha,” through the application of what are taken to be superior skill and understanding of the implications of information.

In financial theory, we speak of the "market portfolio," which comprises all assets in the world! Obviously, no such real portfolio exists. But we do have highly diversified index funds and ETFs that we take to give us a decent estimate of the rate of return for the market portfolio. Holding one of the available S&P 500 index funds, or the Wilshire 5000 index fund, or the Vanguard Total Stock fund comes close enough to the market portfolio for most practical purposes.

So, does a true EMH imply that individual investors cannot ever earn a rate of return higher than the rate of return available from holding a Wilshire 5000 index fund?  No.  The EMH doesn’t rule out any particular investor earning excess returns.  That outcome can and does happen routinely.  The EMH implies that investors cannot systematically and consistently earn excess returns over a lengthy time horizon of several years.   

Some notable investors have outperformed the market over a long enough period to rule out chance.  Warren Buffet flies to mind (at least over the first half or so of his notable career); Peter Lynch might qualify (although Lynch’s success with the Magellan fund might not have been measured appropriately on a risk-adjusted basis); William O'Neil and the CAN SLIM program might qualify; the Value Line System appears to "beat the market" (although some researchers argue that neither CAN SLIM nor the Value Line System earn “alpha” after considering all costs of using those systems).  Logically, if someone devises a system that can consistently and reliably earn excess returns on a risk-adjusted basis, that someone better keep the system a secret.  For if everyone comes to know the system, excess returns will be arbitraged away.  Economists who conduct research about the EMH claim that no one has ever devised a market-beating system.  Well, maybe someone has, but keeps the system a closely-held secret, in which case, the researchers will never learn about the system.

Some investors and money managers have predicted market crashes with what appears to be enough time-frame accuracy to avoid the large losses suffered by most investors when the stock market plunges. Some people (but only a few) do avoid market crashes by exiting stocks and bonds before a market crash occurs.  What is not clear is whether the so-called “smart money” headed for the exits soon enough by luck, or by calculated and skillful interpretation of information, or perhaps even by possessing insider information.

The EMH says that market prices already have "baked in" all "relevant and available" information. We can quibble about what "relevant and available" means (which is just what the three forms of the EMH --- weak, semi-strong, and strong --- do), but that really isn't the point either.  The EMH really says that no other superior source of information is available to tell investors whether a financial asset is over or underpriced, aside from the market determined price itself.  In other words, the market price simply is the best estimate of the value of the asset, given all relevant and available information at the instant the market price is formed.

People often ask how could it be that market prices for securities were unbiased, best measures of value on day 1, when on day 2 the market crashes, losing 10% of its value.  Weren’t prices on day 1 obviously “wrong.”  No.  Prices on day 1 were not “wrong,” except from the vantage point of day 2.  Looking backward is very different from looking forward.  The EMH says that security prices on day 2 are no more nor less “right” than prices were on day 1.

Market prices reveal information about what “others” think value is.  We ourselves may have a different judgment of value.  We may turn out to be right, when “later on,” others come to the same value judgments we already hold before the “others” do.  Or, we may turn out to be wrong, when we “later on” change our judgment of value.  What seems entirely clear, though, is that security prices formed through the interaction of voluntary exchanges among suppliers and demanders necessarily reflect what “others” think value is. 

If an individual investor is somehow going to “beat the market” consistently and reliably,  that investor is going to have to somehow reach judgments about future value of financial assets that a multitude of “others” somehow do not reach.

How could someone reach those superior judgments of value, leading to outperforming an appropriate (risk adjusted) market index of financial securities? First, they could be lucky. But we wouldn't expect someone to be lucky over a large number of trials.  So, Warren Buffet does not appear to be just lucky.  Second, someone could have information that guides her trades that no one else has. No mystery here.  Possession of insider information and the ability to use it without detection of regulators would certainly generate excess returns.  But because the volume of trades made using inside information would ordinarily be insignificant compared to total volume in a particular security, changes in market prices for the assets traded would  likely be imperceptible.  A third way for someone to earn excess returns is to interpret the same information that everyone has, but interpret it differently and more “appropriately” for buying and selling financial assets.  Perhaps this possibility explains at least part of Warren Buffet’s phenomenal success.  Or, maybe Mr. Buffet has just been lucky in an extended run of trials.  The theory of probability does not eliminate that possibility, it just makes it highly unlikely.

What can we conclude? The EMH is certainly tautologically true, just as Murphy observes.  But does a true EMH rule out the possibility of some investors outperforming the market portfolio consistently, measured over many trials on a correctly measured risk-adjusted basis?  No.  The EMH does not rule that possibility out.  But the EMH does rule out the possibility that we all can somehow earn excess returns, if only we knew a little more or could exercise a little more skill.

As for inside information, we probably wouldn't have laws against it if it weren't something that happens. As for interpreting the same information differently and more correctly, we wouldn't have some investment advisors that get paid millions per year, year after year, if what they were selling had absolutely no value. You really can't fool all those high-net-worth individuals year after year for ever, do you think?  Abraham Lincoln’s thoughts about fooling people comes to mind.

So, Professor Fama is certainly right; financial markets (at least in the developed world) are definitely efficient. But the assertions that some writers about the EMH routinely make are not correct.  What is doubtless true is that only a few investors are able to outperform the market consistently. They are those with great luck, those with inside information, and those with unusual insights into what information means for the future. But the huge majority of investors (literally, most of us) will not be able to outperform the market consistently over the long haul. If the EMH and its implications were stated that way, I doubt if anyone would argue much about it.

No comments: