David R. Henderson  

Bob Murphy on EMH

When Is Uncertainty an Argumen... What's Really Wrong With Cryon...

Ever since I was an assistant professor at the University of Rochester's B-school (now called the Simon School) in the late 1970s, I have believed in the Efficient Markets Hypothesis. The basic idea is that market prices reflect all available information. See this article in the Encyclopedia for a fairly up-to-date statement of the issues and the evidence. But I think that at some point I started to hold the EMH as an article of faith rather than as something that could be wrong.

In a recent article, economist Bob Murphy takes on three defenders of EMH. It's well worth reading. A couple of highlights:

On Bob Lucas's defense:

Lucas's arguments here are typical in this debate. He offers a seductive mixture of assertion and non sequitur to make his case. First, the EMH is itself under dispute, so it hardly helps to cite the EMH and its implications. (This is akin to a Christian quoting the Bible to an atheist to prove the authority of Scripture.)

Now, in what sense has it "been known for more than 40 years" that it's impossible to predict sudden falls in asset values? Didn't Mark Thornton and others warn us that the housing bubble was too good to be true several years before the crash? What more could an Austrian cynic do to disprove the EMH, than to predict that "the market" was all wrong when it came to housing prices, risk premiums, and so forth? Investors who heeded the warnings of Thornton and others got out of the stock market, didn't buy houses to flip in 2005, and, otherwise, managed to outperform other people who were caught up in the euphoric boom. If that's not "beating the market," what is?

Quoting Bill Easterly's defense:

[E]conomists did something even better than predict the crisis. We correctly predicted that we would not be able to predict it. The most important part of the much-maligned efficient-markets hypothesis (EMH) is that nobody can systematically beat the stock market. Which implies nobody can predict a market crash, because if you could, then you would obviously beat the market.

and then commenting:

Now c'mon -- that's just plain funny.

Comments and Sharing


COMMENTS (14 to date)
Patrick R. Sullivan writes:

I've only read the blog post, not the linked to article, but it seems that Murphy doesn't understand the EMH. All it says is that the 'best' theory of investing has proven to be a long term, buy and hold strategy of a diversified portfolio of equities. That such a strategy occasionally falters is irrelevant.

To disprove the EMH one has to come up with an alternative theory that provides higher long term returns to an investor. If I go to Las Vegas and hit my lucky number at roulette I'll make a lot of money, but I wouldn't bet on it as a strategy to outperform EMH.

Phil writes:

Was it really obvious that the housing crash would happen, or is it only obvious in retrospect?

If person A things there's a 5% chance of a crash, and Mark Thornton thinks there's a 70% chance, and a crash happens ... how do you prove who was right? Maybe there was a 6% chance, just as the market predicted, and Thornton got rich because he was lucky?

I think the tech stock bubble is much more relevant than the housing crash. Houses were overvalued by ... what, 30%? Tech stocks proved to be overvalued by more than 100%, and, even at the time, there was no good explanation of how the prices would be justified by future earnings. *That* is a story I think of how there wasn't a good explanation at the time -- not just in retrospect.

david writes:

@ Patrick R. Sullivan

Yes, that's the weak EMH. The weak EMH is easy to defend, but unfortunately it also has weak implications - for instance, it doesn't give any reason to think that prices at any given period reflect all available information. Maybe you're in one the periods where an aberration is occurring, lucky you! There's that famous saying about being dead in the long run, I believe. Or staying solvent longer than the market.

There's a mistake that some EMH defenders make where they defend the weak form against attack but use the strong form to draw conclusions. Don't make that mistake! If your defense of the EMH is that "all" it says is that the best long-term portfolio is such-and-such, then you can't then turn around and say, as Lucas did during the housing bubble, that the Fed cannot legitimately act to deflate a bubble because current prices reflect all available information (the strong EMH). Or say post-burst that the bubble couldn't have been forseen.

kebko writes:

Why is this different than prices in any other market?
The analogy I use is that a passive investor, or a market theorist, is like a shopper in the grocery store. All the items are there, priced, and waiting for you to buy them. But, there are butchers, cattle breeders, and all sorts of other players. The shopper might decide that the store is an example of EMH, because would likely do no better for himself if he went out & bred & butchered his own cow for steaks. But, it doesn't look like an efficient market to the successful breeder, who says, "I've been profiting from this business for years. It's not that hard to do."

Why do economists treat financial markets like they are different? Sure, if all the shoppers decided they could be butchers & breeders, they would generally fail. Sure, breeders & butchers frequently fail. None of this proves EMH. It just means that there are some consumers who will do best to take the prices that are given, and those prices are really the result of the actions of many butchers & breeders, some who have skills that will earn them excess profits over time.

Why are financial markets any different?

fundamentalist writes:

Patrick: "To disprove the EMH one has to come up with an alternative theory that provides higher long term returns to an investor."

I have seen research from the University of Chicago that says value investing can outperform the market. And I think you can use the Austrian business cycle to do even better.

"The basic idea is that market prices reflect all available information."

Does that mean that the market reflects all of the information that every individual person has, or just public information? I think the distinction makes a big difference. Of course the market reflects all publicly available information. And it reflects the consensus interpretation of that information. But the consensus is often wrong. But a few individuals can hold information that others don't have that can give them an edge in investing. Obviously, once that special knowledge becomes public, the advantage goes away.

Personally, I have beat the market for the past three years with little more effort than applying what I know about the ABCT. But I can do that only because the ABCT is not the consensus view of the economy. If it were, I would lose my advantage, which is why I hope that mainstream econ continues promoting random shocks as the only cause of depressions and Keynesian moonshine as the only cure.

wm13 writes:

"Didn't Mark Thornton and others warn us that the housing bubble was too good to be true several years before the crash? What more could an Austrian cynic do . . . .?"

To disprove EMH in a useful way, Thornton (or anyone else) has to issue a string of useful predictions, outnumbering any wrong predictions, such that someone following his advice outperforms the market over a prolonged period. Anyone can have either (i) an occasional insight or (ii) a fixed monomania which is occasionally correct. I'm not a student of economists, but there are certainly financial columnists in both categories, and their sometime predictive successes don't refute EMH.

While I am very sympathetic to the Austrian school they are very poor on the EMH. I think this flows from not understanding the theory well before jumping to a conclusion. Similarly, not liking how it is normally described as opposed to understanding how markets become efficient. I would have thought that a story about investors competing to gain an informational advantage would sit well with those economists that emphasise entrepreneurship.

random writes:


Fama pointed out 30 years ago that the EMH was only testable with a model of market equilibrium. Rejecting the EHM is rejecting an untestable joint hypothesis. I prefer to believe that the people in the market aren't dumb, it's just that the modelers who don't know the correct model of the world are over-arrogant.

Rimfax writes:

Investors who heeded the warnings of Thornton and others got out of the stock market, didn't buy houses to flip in 2005, and, otherwise, managed to outperform other people who were caught up in the euphoric boom. If that's not "beating the market," what is?

Wasn't it essentially proven on another thread (here or at MarginalRevolution) that riding the market right up to, and largely into, the crash is still a winning strategy over a premature bear position? In other words, while being a loser in a crash sucks, being an even-slightly premature bear still sucks worse (unless stability has a value in and of itself for you).

fundamentalist writes:

Rimfax, In theory, riding the bull market until after the crash is a winning strategy if you don't consider any other factors. The problem with that strategy is that everyone is heading for the door at the same time when the crash becomes obvious. That's the problem hedge funds have. If you exit before the crash while everyone else is optimistic, you're likely to sell at a good price. If you wait until the crash is obvious, you're going to have to ride the bear down a long dark tunnel as everyone else tries to get out with you.

R. Richard Schweitzer writes:

This may sound a bit off course or naive:

Is there not also the possibility of market prices reflecting the comprehension and interpretation of the available information (which may never be "complete") by the dominant participants in the particular market?

Any information, as such (blank facts), becomes interpreted, connected with other bits of information, in many differing ways by participants who vary in capabilities and objectives.

So, what EMT mighy more validly state is that markets reflect is the totality of the information "processed" by the participants in seeking varying objectives

Ryan Vann writes:

Mr. Schweitzer,

That is probably how your typical Austrian, or Complexity Economist would interpret the EMT. If we assume that the EMT is simply the optimal aggregate comprehension of scattered information, it is still entirely possible to beat the market, as well as be beaten by it.

So no, I don't think you sound naive at all; you just use a different framework than others (I tend to agree with that framework).

mark writes:

If the EMH is stated as "no one can systematically outperform the stock market", it does not follow per Mr Easterly that "no one can predict a market crash". One is systematic, the other is isolated. The use of the word "systematically" plainly accommodates suggests that EMH recognizes occasional outperformance by individuals. It's like saying you can't make a living playing the lottery everyday vs no one can ever win the lottery anyday. In addition, the words "systematically" and "predict" are both vague and can easily accommodate overlap.

Jim Glass writes:
Easterly: "Which implies nobody can predict a market crash, because if you could, then you would obviously beat the market."

Now c'mon -- that's just plain funny.

Gee, I must have lost my sense of humor somewhere.

When an asset class appreciates greatly compared to historical norms, how does one know that the result is going to be a crash -- as opposed to a return of the long-term norm through a period of reduced appreciation, or no appreciation, or slow and gradual decline in value? All of which will return values to the prior norm, no crash necessary. (Assuming a full return to the long-term norm to begin with.)

Let me know so that next time I can beat the market too!

A practical test of whether a bubble like in housing makes it easy for smart investors to systematically beat the market is to see how many smart investors used it to systematically beat the market. Because a fortune could be made by those who did.

How many really smart investors made fortunes this way? Buffett? No. Soros? No. The big banks who had plenty of skin in the game? Uh, no. The big mortage companies that had even more skin in the game. Um, no. The central banks of the US, Britain, Europe, Russia, Hungary, Iceland, all of whom plenty of skin in the game? No, not one.

Well, surely, Robert Shiller called the crash of the housing bubble before it happened, right? Um ... nope.

Did Bob Murphy make a lot of money by calling the crash in advance?

Hmmm ... maybe it's not so easy to call the crash of a bubble after all. First you have to know it's a bubble, and then you have to know that it will in fact crash, instead of adjust more slowly and calmly ... how do we know that again? In a way that explains why so many smart investors, including so many who had so much skin in the game, didn't know it last time?

The EMH says one can't systematically beat the market using only publicly available information (without correspondingly jacking up risk.)

To systematically beat the market using the publicly available information that "a bubble exists and will bust", first the bubble has to exist ... then it has to in fact bust rather than run another more moderate course ... then one has to know how to time both stepping into and out of the bubble so as to beat the market -- on a systematic basis.

(I once saw Jim Grant be publicly applauded for calling the bubble, and he replied -- much more honestly than most other supposed bubble callers -- "No, I was wrong. I called it too soon. Calling too soon is as bad as calling too late. Any fool can see when prices are 'too high'. Knowing what to do about it and when to do it is the only thing that deserves credit". The other bubble callers on the panel didn't appreciate that. ;-) )

If the housing bubble so obviously refutes this test of the EMH then let's see the evidence of how smart investors systematically made a beat-the-market profit off it. Considering the general wipe-out of the last 18 months, I'd say such evidence is probably lacking.

Today is Monday, and today everybody knows all the plays the coaches should have called in all the college and pro football games played over the weekend -- even though (especially when!) the professional coaches paid fortunes small and large to call the plays were too dumbass to call them.

Saying now that it is laughably obvious that the housing bubble made it so easy to profitably predict a housing crash -- after the general wipeout of all the pros with skin in the game who didn't see it coming, and the massive missed opportunities of the other pro investors who didn't call it -- seems pretty much the same sort of after-the-game quarterbacking.

Comments for this entry have been closed
Return to top