« Ownership, control & workers | Main | When China rules the world: a review »

August 06, 2009



This is the way I approach it (as a guy with a rusty degree in the sciences):

First you have observation. Then you develop theory and hypothesis.

That markets are "hard" is the old observation. Early work on "random walk" and "drunkard's walk" were observations of the market.

What developed later was they theory and hypothesis about the "why" or the "because."

The EMH said markets took the random walk because they were informationally efficient, and that their prices reflected available information.

As an old scientist I'm perturbed that the theory and hypothesis could really be shot down, and rolled back to a "weak" EMH that is just the observation again.

In my cranky scientific opinion, EMH has become a tautology and nothing more. Markets follow EMH because EMH has been redefined to be nothing more than observed market behavior.

There is no theory anymore!


Note: "No one can predict market moves with any reliability" is also an observation. It has long been an observation.

It only "proves" a theory that you make tautologically equivalent to "no one can predict the market."

The problem with "EMH" is the "EM" part. You have to jettison that (or at least hide it in your coat for a minute) while exercising the tautology.

Of course, the worst EMH boosters bring the "EM" part back out of their coats after exercising the tautology. "Since we've proven EMH," they say, "we know markets are efficient."

Luis Enrique


I think there is a theory to explain the observation of unpredictable future price moves, at least if I read Chris correctly.

Instead of positing a market of rational and informed actors that process all available information with the result that future price moves are unpredictable, he is positing a market of somewhat rational and somewhat well informed actors that process available information with the result that future price moves are unpredictable by other somewhat rational and somewhat well informed actors.

There is a theory here, which consists of understanding how prices adjust in asset markets. Many people do not (I didn't). The basic insight of EMH, that if it was possible to predict where a share price was going it would already be there, remains both important and non-obvious (at least, I can remember not grasping it). This weak version of the EMH leaves room for astute individuals to have systematic predictive ability, but they have to be compensated for by the non-astute, otherwise, well, we'd be in the strong-EMH world.

Just what proportion of market participants can posses predictive ability, before we'd be in a strong EMH world, I don't know. I think that's an interesting question.

I haven't kept up with what John Quiggan has been writing on CT, and I haven't read Justin Fox's book yet, but when it comes to the debunked version of the EMH that states prices are always "right", I'd like to understand what defines "right".

sean morris

"efficient" is not the same word or meaning as "perfect"

"efficient" means in this context better than something else compared alongside.

As an engineer I like my tools, some tools are better than others in solving a given problem, or repairing something broken, or calculating that something might be on the verge of failure.

"EMH" seems to me a pretty good tool, one that gives us more insight that we would normally have, afterall it was markets that spotted the oncoming crash not the politicos, was it not?

No I do not accept economics as a science, but that is not to devalue its usefulness to us. As our understanding grows so in due course it may well become one, but as someone who thinks humans, markets and the universe itself is irrational, it could be me thats the tool?


I could say that Luis is making additional observations (of more astute and more rational actors), and I'd agree. They are there.

I also agree with him that they are there in "some proportion" to other actors. Those other actors range from hyperactive day traders, to dogged index investors on a fixed strategy of market allocation, to semi-shrewd fund managers, to funds designed for marketing appeal, to millisecond arbitrage systems.

Information and information flows can also be observed.

I'm not sure that disproves my tautology.

The EMH is hanging on because the observations are important and valid and for so many the observations are bound to the EMH.

EMH as a "tool" sean? In what way?


The EMH is in essence a meaningless piece of deterministic circular reasoning. Of course no-one that operates within the market can ‘beat-it’. But it gets confused with the idea of a ‘rational’ market – and the notion that billions would be invested in dodgy loans seems to buck this concept. The market must always be ‘correct’ in the sense that there could be no other outcome, given the specific conditions in which it operated – but that’s pretty otiose.


"Of course no-one that operates within the market can ‘beat-it’. "

There is no reason to write "of course" here. The EMH makes a simple testable prediction - that there are no people who can consistently beat the market (of course, in any year, half of the players will beat it, and half will not, but that is just one year; the EMH is about consistency). One can imagine markets where the EMH might not be the case, say if a secret cabal of hook-nosed people controlled the market - then members of the cabal would be able to beat the market on a consistent basis. The fact that the EMH holds most of the time disproves the existence of such a cabal.

There are some interesting exceptions to EMH which show that it isn't a simple tautology. The dot-com bubble was made worse by some people who systematically overvalued companies that they could sell shares in. In that case, the people who profited from the dot-com bubble had information that they did not share with anyone else in the market - information like the fact that there is no way to make a billion dollars selling pet food online.


There's far too much over-theorizing about the EMH. As William says, it gives us a simple prediction - that you can't consistently out-perform the market unless you take on extra risk. And this prediction is hugely corroborated by the evidence.
If the EMH were wrong, fund managers would trade on their own account rather than take fee income; the Sunday Times rich list would be dominated by stock market traders (as dinstinct from fee gatherers); and Warren Buffett's experiment in how to beat the market would be frequently replicated.
None of these things are true.


Chis, William, you can't make a "testable prediction" that predates your theory.

Louis Bachelier talked about the "drunkard's walk" in markets around 1900, but I doubt he was the first.

Neither should you make one (in the instance of "weak form" EMH) that discards the original proposition about the incorporation of all available information in prices.


Shortest: There is no proof of the EMH mechanism, there is only the original observation.


Slow down Chris, you are smart enough to get this:

If the EMH were wrong, fund managers would trade on their own account rather than take fee income; the Sunday Times rich list would be dominated by stock market traders (as dinstinct from fee gatherers); and Warren Buffett's experiment in how to beat the market would be frequently replicated.

The observation of unpredictability in markets ("no free lunch") is separable from the contribution/assertion of the EMH, that it is "because" of certain aspects of information propagation.

Prices can be irrational and inefficiently prices and still unpredictable


Nice distinction in the title.


This isn't a valid test at all. First, a test of "whether you can beat the market" should clearly be carried out on the basis of gross-of-fees performance and on midprices (otherwise you're confusing overcharging with underperformance), whereas that data is provided net of fees and on a bid-to-offer basis. Second, you've cherry-picked the best-performing tracker fund in the dataset (the F&C one); if I were to take the Gartmore UK Tracker as representative of trackers, I could say that more than half of all unit trusts (159 out of 254) outperformed at least one tracker.


Yes the market is highly inefficient..but so is the state.

Everywhere, including financial markets and other financial services have been dominated by those using lean methods and using factory production approaches.

Professor John Seddon in a new podcast available on ITunes (ITunes>podcasts>search for the systems thinking review>download) identifies where lean went seriously wrong.

It is important to show that value is a long-term thing to be extracted. Perhaps where investors want short term results, they have to expose more money than normal with no extra rights or gaurantees and these are the last to be paid in bancruptcy proceedings?


You make two points precisely - either you get lucky or you take on more risk.In fact, our bust banks bear this out in spades - they got luck for a bit, took on extra risk - and now have paid the price.


Jensen's claim in 1978 has been shown to be wrong - EMH is not a viable theory. Why the behavioural theories then? And even Scholes makes he case against himself in "thinking much of the blame for the recent woe should be pinned not on economists’ theories and models but on those on Wall Street and in the City who pushed them too far in practice."

Exactly. When you have the same people insider trading who are making the rules and printing the money at the same time, using fractional reserve banking to create something out of nothing - and that model being followed - look at silver just now, then EMH is seen as naive.

Shleifer's theory of costs simply putting a spanner in the works with the informed beating noise traders. There are so many examples of the market not being rational.

Krugman said, as you'd be aware:

"most macroeconomics of the past 30 years was “spectacularly useless at best, and positively harmful at worst”

He might have added "LSE socialist economics" as well.


Eric Falkenstein says "Bank acquisitions were evaluated in part by their Community Reinvestment Act record, which necessitated lowering underwriting criteria on homes".

I've seen conflicting reports on the role of the CRA in the sub-prime lending boom; by some accounts only 20% of loans were from CRA regulated firms?


Mark Wadsworth

Oh dear.

People confuse two things:

1. Proper markets, i.e. for goods, labour, services etc, which are highly efficient, or at least infinitely more efficient than centrally planned.

2. Secondary markets in 'capital' assets, like shares, housing, which are subject to bubbles and irrational exuberance/despair.

The government is usually behind these bubbles because of rationing, legislative changes, subsidies, exchange controls, setting artificially high or low interest rates e.g. Labour's idea that you can bribe the people with ever-rising house prices.

The lefties say "Capital markets are all over the place. Therefore we must regulate them. Oh, and we'll regulate proper markets as well while we're at it"

The rightwingers say "Proper markets are highly efficient therefore capital markets must be efficient as well".

Both of them are wrong, as it happens.


"1. Proper markets, i.e. for goods, labour, services etc, which are highly efficient, or at least infinitely more efficient than centrally planned."

These proper markets are planned to the hilt. Only if you restrict the word centrally to bumbling autocrats does this even have a shot at being true.


The comments to this entry are closed.

blogs I like

Blog powered by Typepad