« Two perspectives on EMH: Biz Skool vs. Econ Dept | Main | "Dispelling Canadian myths about foreign direct investment" »


Feed You can follow this conversation by subscribing to the comment feed for this post.

EMH relies on prices to reflect available information. It really should read, EMH relies value always representing market information. One needs to represent the distribution of value independent of the distribution of credit. So, value, as an independent entity, is represented by the distribution of sales, by rate and volume. When the stock market reflects information, then the variation is sales (dRate*dVolume) should be the same for all stocks, each stock on the index transmitting the same quanta of information for each sale.

Yes, we have tests for this.

This paradox:

"if everyone believes EMH is true, so that market prices already reflect all available information on fundamental values, then no individual has any incentive to collect and process that information, and everybody picks assets by throwing darts, or buys the index, so market prices will not reflect any available information on fundamentals, so EMH will be false"

is very similar to the argument against EMH put forth by Stiglitz and Grossman in their paper "On the impossibility of efficient markets". They argue that as long as information matters and is costly, market prices must always depart at least slightly perfectly efficient prices, otherwise their will be no profit to obtaining the costly information, and therefore markets will become imperfect, and at which point it will be profitable to gather costly information.

I think you have put it very well, btw.

Matt: I'm sort of half-following what you are saying there, I think.

Adam: Thanks. Yes. What I'm doing is essentially trying to represent the Stiglitz-Grossman idea in a picture that shows the interaction between the two forces, with (presumably) an interior solution.

Hey, I've seen that graph before - you've just changed the labels!

Back in 'skool days, when dos ruled and Lotus 123 was the killer app, the neat thing about it was that suddenly there was a whole world of coloured pie charts at one's finger tips. Two dimensional graphs with two lines just seemed so boring (mind you, the guy teaching applied economics was entertaining in a quirky sort of way).

In an earlier blog posting, I made a terse definition of finance: the transfer of wealth. Now I'd like to expand a bit, and in keeping with the environmental theme that has captured the imagination of many people, I'm labelling my pie chart Gaia.

Now Gaia and its atmosphere is so large that, by definition, the only externality of any significance is the sun which forever delivers energy to its surface - and is stored over eons in the fossil fuels we've grown so accustomed to (through decaying plants and animals).

Also, Gaia when it was formed had a finite set of elements - stored in various forms and compounds. In its core - lots of decaying radioactive material (hence the source of geothermal energy and volcanic activity). On or near its surface - water, air, and exploitable resources.

So, if you were to draw the Gaia pie chart, depending upon how many colours you had available and how small you wanted the pieces to be, that is the starting point at which I (as some anonymous blog commenter) think understanding finance should start.

And therefore, as humans, all that one is doing is competing for different slices of the finite pie - ie it's a zero sum game. All the accumulated knowledge to date simply is utilized to determine how big your piece of the pie will be.

So, there is no efficient market, no perfect information, just pie slice traders.

Now I must go and meditate, or have some desert...ohmmmmmmm :)

I think this is a brilliant insight. I've been keeping a small blog of similar insights that have fundamentally changed how I look at the world at www.notableinsights.com - no one reads it, but I hope you don't mind if I link to your post from it.

hmmm - not sure if that last comment was directed at me, but if I could be so presumptuous, slice and dice and link away.

And if it wasn't, as Emily Litella used to say (Gilda Radner character on SNL) "Never mind".

I've always wondered if Yogi Berra didn't provide the best summary of the EMH: "Nobody goes to that restaurant anymore - it's too crowded."

Nick, I've put forth a similar argument on Scott Summer's blog.

Anyway, when arguments have such structure, it is an indication that we are asking bad questions. Its similar to the Calvanist vs. Anabaptist debate in Protestant Christianity. They debated whether or not we have free will, when that isn't the real question. The real question is if we have responsibility or not.

In the same way EMH answers are bad question. The question shouldn't be "are markets efficient," the real question is "Can certain individuals given power consistently make better choices for a society than a market."

Three..three...three...THREE! Two...two...two...two..TWO! One...one...ONE! One..one...one...ONE! Corner on Barley!

Who has the bear and the bull?


But the shape of your "demand" curve changes with the number of market participants. Specifically, the more participants, the more the demand curve will bend out towards the upper right corner of the plot, as it takes a lesser percentage of EMH disbelievers to make EMH a better approximation to reality. In fact, a quick thought experiment tells me that if there is no possibility of collusion or coordination between them, the existence of just two EMH disbelievers with unlimited resources suffices to make EMH true for everyone. In the limit, as the number of participants goes to infinity, it seems the EMH is always true.

Jeff: Yes, agreed. I would call that a rightward shift in the demand curve. But since the demand curve can't cross the axes, thinking of a "shift" as a "bowing out", towards the North East corner, as you do, is more precise. What I like about my little picture is that it helps me think through your sort of experiment more easily. And it let's me see the interaction between the demand and supply curves. (Without that last point, we might forget that if markets process information into prices more easily, people will have less incentive to collect information before buying).

Doc: I've been thinking about your point. I don't think the existence of positive or negative feedback is what makes it a good or bad question. Adam Smith has negative feedback; Keynes' multiplier has positive feedback. That's just how the world is (or isn't). But re-defining a fuzzy question into a question about policy seems to be a sensible move.

Stephen: That nicely captures the "self-falsifying" (as opposed to "self-fullfilling") nature of the EMH.

Just visiting: "And therefore, as humans, all that one is doing is competing for different slices of the finite pie - ie it's a zero sum game. All the accumulated knowledge to date simply is utilized to determine how big your piece of the pie will be."

Nope. I strongly disagree with that. Let's ban ALL exchange, if exchange is zero sum. I think this exposes the fallacy of thinking about economics in purely environmentalist terms. But off-topic. so I won't explore further.

Jeff: a follow-up. " In the limit, as the number of participants goes to infinity, it seems the EMH is always true." But I don't believe that generally holds. Remember there is always some percentage of the market that is buying or selling for purely personal reasons, unrelated to the quality of the good. I'm selling my car because I'm emigrating. (Some economists call them "noise traders", but I think that's a misleading term.) Unless the percentage of fundamentalists in the market is large relative to the percentage of people buying and selling for personal liquidity reasons, it won't be 100% "efficient".


Unless the percentage of fundamentalists in the market is large relative to the percentage of people buying and selling for personal liquidity reasons, it won't be 100% "efficient".

No, the fundamentalists (think Warren Buffett or Peter Lynch or whoever's picking stocks for Calpers) only have to be large in resources, not numbers, compared to the individuals selling for personal liquidity reasons. Just saying "personal liquidity reasons" versus "Warren Buffett" is enough to make it kind of obvious which one is going to end up moving the market.

I'm not claiming 100 percent efficiency, except in the weak sense that nobody with sizable resources who understands fundamentals is going to leave easy money on the table. They will have to make enough profit to pay them for their research and analysis efforts, but profits over and above that are going to be competed away.

All of this is very standard and traditional economic thinking. Like Scott Sumner, I just have never seen any convincing demonstration that EMH is not a very good approximation to reality. Arbitrage is one of the most fundamental ideas there is in economics, and it's probably the one (along with comparative advantage) that most distinguishes the economic way of thinking.

"I just have never seen any convincing demonstration that EMH is not a very good approximation to reality."

So what sort of demonstration would convince you that the EMH is not a good approximation to reality?

Let's ban ALL exchange, if exchange is zero sum. I think this exposes the fallacy of thinking about economics in purely environmentalist terms. But off-topic. so I won't explore further.

I think your last point might be described by some as "sunk cost".

Carry on...

Beads and Manhattan
In January 1625 the ship Orange Tree left Amsterdam for New Netherlands with William Verhulst, who was to become the second governor of the colony and Peter Minuit, who was to succeed him. Verhulst had instructions from the merchant group known as the West India Company, who were financing the building of the colony. The instructions read in part:

In case any Indian should be living on the aforesaid land or make any claim upon it or any other places that are of use to us, they must not be driven away by force or threat, but by good words be persuaded to leave, or be given something therefor to their satisfaction, or else be allowed to live among us, a contract being made thereof and signed by them in their manner, since such contracts upon other occasions maybe very useful to the Company. [A.J.F. van Laer, trans. 1924 Documents Relating to New Netherlands 1624-1626 In the Huntington Hartford Library. San Marino CA, pp. 51-2.]

Further instructions were sent out to Verhulst on 22 April 1625 telling him much the same thing and specifically mentioning trade goods. So, the governor was explicitly instructed to pay something for the land they were to settle on if need be....



JP Koning: A convincing demonstration would be a trading strategy based on information that was easily available to everyone that consistently beats the market. In it's weak form, all the EMH says it that there is no such thing.

If you believe markets are mostly efficient except for information diffusion, then you may see a price move both as information and an inadequate response to it, leading you to become a momentum investor promoting instability rather than stability. If you don't delude yourself, you may even make money on the whipsaw, while if you do, you only become another sucker.

Peoples ability to rationalize their actions is pretty much limitless which emh largely is.

The problem with the example you pose to Nick (two fundamental traders in a world of EMH traders) is you assume unlimited resources for the former. But economics is the study of limited resources, and in the real world no one has unlimited resources.

It is possible that the EMH traders might all decide to simultaneously sell the securities the fundamental traders have accumulated, say if their darts all hit the same target by coincidence. Or they may all decide to sell stocks in general for real world liquidity concerns, out of fear, etc.

The effect of their selling may overwhelm the fundamental trader's limited equity, forcing them into bankruptcy. With the non-EMH believers out of the market, there is no way for the fundamentals to reassert themselves. A market driven by EMH can weed out a few fundamental traders, and thereby create a self-defeating prophecy.

In other words, "the market can stay irrational longer than you can stay solvent" - JMK

"No, the fundamentalists (think Warren Buffett or Peter Lynch or whoever's picking stocks for Calpers) only have to be large in resources, not numbers, compared to the individuals selling for personal liquidity reasons."

But the fundamentalists who are picking stocks do so as mutual or hedge fund managers, working on behalf of those very individual's selling for liquidity reasons. If these individuals choose to sell their fund units at once, it doesn't matter what the fundamentalist wants to do. He'll have to sell too in order to meet redemption requests. Those governed by liquidity concerns can overwhelm the desires of their fundamentally driven managers.

JP: All I can say is, if you really believe what you're saying, you should be able to discern when the market is irrational, take the other side, and become infinitely rich. Keynes is wrong. If the market really is irrational, but you aren't, you should be able to convince lenders of that. If you can, your solvency issue goes away, and again you will be infinitely rich.

I realize that it's not fashionable to say so in many quarters, but economics, particularly price theory, really works. It takes a model to beat a model, and the economic model of how the world works explains more than any competing paradigm. If you have a better one, you should stop hanging around commenting on blogs and go out and become infinitely rich.

If you have a better one...

Buy low, sell high.

[But to buy, you need a seller; and to sell, you need a buyer. - zero sum :) ]

Jeff: Goldman Sachs.


Well the competing model to the EMH is that financial markets are complex and constantly evolving systems. In these systems traders use rules that they learn in an evolutionary manner, keeping the rules that work, disposing of the ones that don't. The interaction of rule-using traders creates financial prices.

Fundamental analysis is an evolved institution, a knowledge base so to say, from which traders can draw rules to guide their choices. So is technical analysis. The EMH is also such an institution; it tells traders to use trading rules such as investing in passive index funds.

The theory that markets are evolving systems is able to contain within it the EMH theory by relegating its status to just one of many competing institutions, and therefore supersedes it.

^ In a nutshell: educated (maybe) guesses.

JP Koning: If you keep the rules that work, and they continue to work, you become infinitely rich. Nobody is doing that because the rules don't actually work because future price changes are not predictable. Furthermore, the EMH is not an institution, it's a hypothesis about how markets work. It fits the data very well. But it's more than just an empirical conjecture: there are sound economic reasons to expect it to be true. Precisely the opposite (on both counts) is true of technical analysis, which always has been and always will be a fraud.

The "theory that markets are evolving systems" is no theory at all. It makes no testable predictions, and explains nothing. A theory is more than vacuous generalities.

Patrick: The fact that the government has chosen to give some firms large amounts of the taxpayers money says nothing about the EMH. The EMH is about how prices reflect publicly-available information. It's not about private information, which is one way to think about corruption. One consequence of the bailouts is that now people place a non-zero probability on repeat performances. Witness the non-zero value of old GM stock, which is clearly worthless unless there are more government handouts. This is not inconsistent with the EMH, it simply reflects that at least some market participants think there's a chance they'll somehow get a further bailout that will give the old stock some value.

Just visiting: what are you trying to say? Buy low, sell high is not a theory or an explanation of anything. A theory explains something, and a scientific theory makes testable predictions.

What am I saying?

Basically, this is not science. It is an attempt to create formulas to predict human behaviour. Sure, if everyone uses exactly the same formula, you get the same outcome. Self fulfilling. And if more and more of the Wall Street traders go to the same biz schools and are taught the same formulaic predictors of future value of an entity, you get convergence.

Formulas are excellent at predicting past behaviour. They are not, however science, and therefore are not subjectable to the same sort of testable predictions. Might as well work on a perpetual motion machine.

In Nick's fine post: if EMH is believed true, "everybody picks assets by throwing darts, or buys the index..." Or decides that the only way to beat the index is to game the system by bribing employees of firms to obtain insider information, falsifying accounts if you are the manager, spreading false rumours, then making huge leveraged bets to exploit the imperfections you have created. Anybody here think the runup to the last financial crisis didn't show massive corruption?

Here's an interesting take: make the model dynamic.

If no-one believes the EMH, it will be more true, and people will start to believe it. As this belief grows, fewer people will do their own homework. When we get to the point where no-one is doing their homework, prices go wonky and the EMH will cease to be a useful description of how prices work. People will stop believing the EMH, and the cycle starts again.

Jeff : Stiglitz-Grossman! The existence of corruption, private information, privileged information, barriers to the flow of information, costs of collecting information, etc ... are precisely what makes EMH wrong. Exploiting information asymmetry is the trading strategy you demanded be produced. GS are the poster boys.

Patrick: Reread my challenge. I asked for "a trading strategy based on information that was easily available to everyone". I do not deny that private information can be profitably exploited. In fact, you can think of the extensive analysis and experience of successful fundamentalist investors (like Peter Lynch and Warren Buffett) as generating private information that they then exploit.

There are some who go further and maintain that market prices efficiently aggregate both private and public information. If everyone with private info fully exploits it, then it is probably true that the market price will reflect all of that info, but the exploiters will make some money in the process.

I asked for "a trading strategy based on information that was easily available to everyone".

But, the info only has "value" if it can be used in the existing models, widely subscribed.

If my model for wholesale gasoline prices in Ontario does not include corn yield in Mexico, then the fact that the crop failed has no value.

A couple of disparate comments:

Information is costly. Even publicly available information is costly to acquire and process. Consequently risk management is neither free nor cheap.

Just visiting from macleans: Capital markets provide, among other things, risk management services. The historical record is rather clear that capital markets provide socially enriching services; in aggregate, the performance is win-win and, more to the point, more productive than all the other real world institutionally-constrained options.

OFF TOPIC, just visiting, your earlier remark about capital markets being a zero-sum distribution game had me wondering how would one characterize natural resource wealth dissipation brought on by excessive entry. In passing, this is often a case of Small is Ugly and Destructive (with a wink to Schumacher), at least if viewed in terms of social opportunity cost.

How would you describe the social purging of resource rents? Does negative sum capture the social wealth destruction that occurs? Do terms like self-loathing have a place in economic policy analysis? :)

It is amusing to contemplate these questions in the context of the EMH because enormous windfall profits have been made during what are often highly publicized and controversial resource collapses, e.g., fisheries, and then the subsequent attempts to institutionally rebuild (typically rights-based catch share such as tradeable quota). Incumbent harvesters lacking the financial skills (read risk management skills) may quit early and then look back at a later date and conclude that they have "lost". Lost what and relative to what? The original wealth-destroying industry or the new rights-based arrangement?

end of OFF-TOPIC

"OFF-TOPIC" reply

First off, my admitted ignorance of much conventional wisdom within the field of economics offers both positives and negatives (your mileage may differ). On the positive side, a lack of group think. On the negative side, easily dismissed. But, I do fit on the underrepresented biz skool end of this discussion, with an engineering (applied science) foundation.

So, when I stand back and look at the matter under discussion (and let's be frank, some of the theoretical stuff Nick writes about I stear clear of with a ten foot pole) I ask myself rhetorically, is this a case of not being able to see the forest for the trees? So, how does one make sure that you are looking at the whole forest? Well, you make it as large as possible, and then start shrinking it down as required. Hence Gaia as a starting point.

So, now let me try to tackle a few of your q's (while acknowledging I may not be as familiar with the vernacular as others).

On all of your questions, I would ask: define your boundaries, which also has a time component. Are your observations about "risk management" and "socially enriching services" limited to the market players themselves, or just the societies where they live, or just for the present time? Because, until I know how you are measuring/scoring things, it's difficult to take issue with "The historical record is rather clear".

Now, I have some hands on experience with resource (oil and gas - in Alberta which seems to me where you are situated) exploitation, and could perhaps enter into a more detailed discussion with you on this topic (natural resource wealth dissipation), if you like. But, the only winking I might do to Schumacher would be to acknowledge that I've only just heard of him. Which may not be surprising to many.

end of "OFF-TOPIC" reply


British economist E.F. Schumacher published Small is Beautiful: Economics As If People Mattered in 1973. I sometimes wonder if it set back development in the 3rd world almost as far as Dependency Theory. Not necessarily because of the ideas in the book but how the book was interpreted by the development industry.

I'm on the west coast and enjoy losing money in the Alberta-based oil patch.


Sorry about your Alberta-based oil patch losses. I hope I haven't contributed to them.

I noted that Stephen Gordon wrote a blog a few years ago (Aug '07 - energy tag on sidebar) titled Extracting rents from the oil sands which linked to his earlier blog From Ghawar to Athabaska (May 07)

If you're a G&M reader, you'll note that this issue received national exposure in a couple of columns late May/June 06 by a familiar ROB columnist, and quickly rose to prominence shortly thereafter in the PC leadership race, underway in Alberta.

Btw, I knew the history of how the royalty scheme evolved for the oil sands, having been there at the time. Might give you an idea of my view of "excessive entry" and "resource rents" :)

"Adam Smith has negative feedback; Keynes' multiplier has positive feedback. That's just how the world is (or isn't). But re-defining a fuzzy question into a question about policy seems to be a sensible move."

Nick, I don't mean the feedback in the economic problem, I mean the feedback inherent in the question. If believing a theory makes that theory more true, then you have a well formed social construct. The belief in money is an example of this, it gets its value from how widespread it is. The more its believed the more true it is, so it is a stable, well-formed social construct.

You describe Keynes as having negative feedbacks, while his theory may have negative feedbacks, Belief in Keynes's theories doesn't have negative feedbacks, and if it did it either would be a malformed social construct (or possibly just incorrect). In fact because of the political economy, I would say that belief in Keynes's theories has positive feedback, which is why so many conservatives support conservative keynesiasm and leftists support spending stimulus.

Doc: Money (in particular Menger's theory of the origin of money as a medium of exchange) is a good example of positive feedback.

You lost me on Keynes (might be a typo/slip, or something). I said the Keynesian multiplier has *positive* feedback. An initial increase in desired expenditure increases income which causes a further increase... But yes, you are talking about feedback between a theory being believed to be true and being actually true (as in my original EMH case). I'm not sure if in that sense Keynesian theory has positive or negative feedback.

Ok, seems we misunderstood each other then. My point is that EMH needs to be reformed into a theory with positive feedbacks that has the same content for it to become a useful part of society.

Technical Analysis is probably outside the scope of the diagram. It searches for patterns in prices, e.g. systematic behaviour of market participants. It probably does so in a crude way as opposed to genetic algorithms (asf.). Forcasting such patterns is equivalent to making hypotheses about market participants' behaviour patterns that is independent of EMH. It probably requires that EMH is not (entirely) true in the first place. It depends more on actual price dynamics (and dynamics about expectations about price dynamics) than expectations about EMH.

Doc: OK we understand each other now. I think I might agree with you on that point. The Cretan saying all Cretans are liars wasn't making an especially useful statement. But I need to think about it some more.

Philipp: I think what you say makes sense. But then any particular theory of Technical Analysis will also tend to be self-falsifying too. If people believe it, and act on that belief, the pattern disappears. But if people don't believe in TA (and if EMH is false) there is probably money to be made from TA.

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad