Models and the Efficient Market Hypothesis

James Hanley

James Hanley is a two-bit college professor who'd rather be canoeing.

Related Post Roulette

151 Responses

  1. James Hanley says:

    Oh, and if James K or J R critiques me, listen to them. If they say I’ve gotten something wrong, I’m certainly going to be inclined to believe them.Report

  2. D Clarity says:

    The EMH as defined by “if enough information is widespread enough, an investor cannot consistently outperform the market” can only possibly be true to the extent that people behave as if it is not true.

    Can a theory that becomes false if everyone believes it be described as true?

    If I knew 20 stocks that were going to be worth $0.00 sometime within the next 20 years, how could I make money off of that?

    Let’s say you have a hypothesis about the behavior of the market. Prices in the market fall by 22.6% in six and a half hours with no significant new available information. That is not enough to get you to reject your hypothesis. Is “Efficient Market Hypothesis” really a good name for your hypothesis?

    Is it possible that Tiger Woods was no better than the average golfer over his career, but just benefitted from some good luck?Report

    • James Hanley in reply to D Clarity says:

      If I knew 20 stocks that were going to be worth $0.00 sometime within the next 20 years, how could I make money off of that?

      Is that public information or not?

      Let’s say you have a hypothesis about the behavior of the market. Prices in the market fall by 22.6% in six and a half hours with no significant new available information. That is not enough to get you to reject your hypothesis.

      How would proposing a new–obviously untested–hypothesis be enough to get me to reject a prior hypothesis? That’s not really how hypothesis generation and testing work, is it?

      Is it possible that Tiger Woods was no better than the average golfer over his career, but just benefitted from some good luck?

      Is Tiger Woods’ success dependent just on having and using better information? I’ve seen this question asked with Michael Jordan in pace of Woods, but for practical persons let me shift to another basketball player, Larry Bird. Bird used to sometimes tell defender which hand he was going to shoot with–giving them the same information as himself–and then he’d score on them anyway. Obviously sports requires more than just information, so they’re a bad analogy.Report

      • D Clarity in reply to James Hanley says:

        Is that public information or not?

        Say that it is – exactly that. It is publicly available information that the 20 stocks are going to $0.00 within the next 20 years, yet they are not at $0.00 now. How do I make money in this situation?

        How would proposing a new–obviously untested–hypothesis

        I was not proposing a new hypothesis. I was stating what happened on October 19, 1987.

        Obviously sports requires more than just information

        And making above-market returns investing doesn’t? Where is your evidence of this?Report

  3. Road Scholar says:

    I accept the EMH. But I have just one question: Given that fund managers can’t reliably beat the market, just why the hell do they make so much money? Seems sort of scammy to me, like Dr. Oz bullshit alternative medicine scammy.Report

  4. Alan Scott says:

    What about Warren Buffet? As a case study, he seems to be the exception that disproves the rule, doesn’t he? But it’s always good to step back from a single-case and look at the world statistically, at least when we have a large enough n. And when it comes to investors, we’ve got a damn large n. And what would we expect to be the distribution, purely from random chance, of such a large n? We’d expect to see most people clustered near the middle, with some outliers at each end. That is, even if all investors were poo-flinging monkeys (and who’s to say they aren’t?), over a lifetime some would wildly underperform the average, and others would “impressively” outperform the average.

    The economic stuff is mostly over my head, but the math guy in me is a bit suspicious of the attribution of Buffett’s success to chance.

    After all, it’s unlikely but possible that a fair coin flipped 50 times will result in 49 heads. But if the coin is truly fair, we’d expect the next fifty flips to be around 25-25, rather than 49-1 again. So is Buffett’s past performance a predictor of his future performance? Or is he no more likely to pick winners today than a million other investors? If it’s the former, then we should really be looking for a better explanation than the fact that sometimes there are outliers.Report

    • zic in reply to Alan Scott says:

      To understand Buffet, you have to look at his investing habits. He didn’t invest in hot stocks (most people rush to buy stocks when they’re at their peak). He purchased boring stocks when nobody else noticed them, and he held onto them, he is not a flipper playing the margins. Buffet also profited on the series of recessions we had over the years; when people panicked and dumped stocks for companies with sound fundamentals, he picked them up for cheap.

      Buffet was patient and had enough wealth to not only ride out the troughs that ruined other people, but to profit from them.Report

    • Mo in reply to Alan Scott says:

      A) You don’t know where in his 49 out of 50 streak Buffett is
      B) He has underperformed the market in 4 of the last 5 yearsReport

      • dhex in reply to Mo says:

        his most interesting trait is his public persona. he’s got “one of the good ones” billionaire cred amongst people generally not in the habit of referring to any wealthy individual as anything other than a rapacious force of nature.Report

      • Kim in reply to Mo says:

        dhex,
        “As good as they get” isn’t much of a compliment. Still it’s better than what people say about the Kochs.Report

    • Chris in reply to Alan Scott says:

      A better example might be Soros. In the lead up to the bursting of the dot-com bubble, Soros (and a bunch of other people) recognized the bubble, indicating inefficiency, and their rational reaction was to sell short. However, simply recognizing the bubble was not enough to know when the bubble would burst, and Soros lost a lot of money as a result. So, from an EMH perspective, even in a bubble pricing was relatively efficient (that is, it had some efficiency, it just wasn’t absolutely effiicent).Report

      • zic in reply to Chris says:

        I think it also matters that both Soros and Buffet have access to a lot of information the average Joe investor simply lacks. I can’t afford to go out and purchase an Forrester report on an industry I’m considering investing in next week; they both can easily do this. Hopefully, the managers of funds I might invest in can do this, too. But I know I will, at best, only have a very limited view of the information available that efficiently shapes markets.

        Revealed information is often private research that’s for sale, not public information that’s readily available to average small-potatoes investors.Report

      • Chris in reply to Chris says:

        I think it also matters that both Soros and Buffet have access to a lot of information the average Joe investor simply lacks.

        That’s important: if prices reflected all information, and were not merely relatively efficient (which, duh, that’s kind of what “price” means in a market), that is they didn’t merely contain some of the information, it wouldn’t matter that some people have more access to non-insider information than others, because all of that would already be reflected in the price.Report

      • Alan Scott in reply to Chris says:

        So, maybe that’s where a lot of my skepticism is addressed.

        As it’s been explained, the EMH suggests that there are no such thing as bubbles, or that bubbles cannot be reliably identified as such until after the fact–which seems suspicious to me as a layman. But is all it really saying is that you can’t predict WHEN a bubble is going to pop? Because that seems like a much more reasonable conclusion.Report

      • Michael Cain in reply to Chris says:

        However, simply recognizing the bubble was not enough to know when the bubble would burst, and Soros lost a lot of money as a result.

        John Maynard Keynes: “Markets can remain irrational longer than you can remain solvent.” Soros and Buffett can remain solvent a heck of a lot longer than you or I can :^) They’re playing with money that has zero effect on whether the kids can go to college, or if the next mortgage payment gets made, or the next meal.Report

      • James Hanley in reply to Chris says:

        @alan-scott
        As it’s been explained, the EMH suggests that there are no such thing as bubbles,
        Well, no, it doesn’t really say that.

        or that bubbles cannot be reliably identified as such until after the fact–which seems suspicious to me as a layman.
        But this means that even if some correctly predicts–or “predicts”–a bubble thus time, it could just be because every time the market booms some folks start hollering “bubble.” Some of them, then, are going to holler at the right time. But will they be able to do it consistently and not yell it when it’s not a bubble? That’s what the idea–whether right or wrong–is getting at.

        But is all it really saying is that you can’t predict WHEN a bubble is going to pop? Because that seems like a much more reasonable conclusion.
        That’s an interesting thought. I think it’s actually saying more (again, rightly or wrongly) but perhaps that’s something like a minimal version of it? I don’t know, but it seems not unreasonable.Report

      • Kim in reply to Chris says:

        Alan,
        it sounds like what you’re describing is the Greater Fool Theory, rather than the EMH.Report

    • Patrick in reply to Alan Scott says:

      Buffett’s performance may indeed be above-average due to skill, but a large portion of it may well indeed be perfectly attributable to chance.

      Some part of it may be attributable to skill.

      In any event, the part that is attributable to skill may still not outperform the market. Here’s an example:

      Let’s say I’m perfectly capable of predicting 5% growth because I have the skill to always choose the fairly conservative but reliable economic winner.

      Let’s say you have no skill whatsoever, but you place your bets entirely over the market. You hit some big winners when some penny stock becomes worth $100 in five years, but you lose an equal amount in the N penny stock bets you made that were all losers.

      At the end of the ten years, we’ve both made 5% per year. The difference between us is that I’ve made the 5% every year, and some years you made 50% and others you lost your ass.

      If what you *want* is a 5% gain every year, (assuming my skill level evaluation is correct) I’m the better person to choose as your fund manager, because you can fairly reliably say that your $100 will be worth $105 next year and $110.25 the year after that and so on.

      If what you *want* is 5% gain over infinite years, the investment strategy doesn’t matter, because the scattershot strategy will, in the long run, produce the same result.Report

  5. Road Scholar says:

    And a side comment: I really despise the way economists use the word “efficiency.” To my engineering-trained brain, efficiency is a measure of output given an input, basically bang for your buck. So if you tell me an engine is efficient I can infer that it consumes less fuel for a given output than a less efficient engine. If economists used the word the same way, an efficient market would be one with low transaction costs (the financial equivalent of friction).

    I think the dichotomy between the way economists use that word and practically everyone else uses it contributes to a lot of misunderstanding around concepts like EMH.Report

    • zic in reply to Road Scholar says:

      That’s part of it.

      But I think that there’s more: market’s correcting is a process, it takes time, and there is a lot of turmoil in that time; as we’ve witnessed over the last six years. I don’t think the problem is not trusting that the markets will (eventually) correct so much as how you handle the muck churned up in that process. This is, it seems to me, the real discussion going on here; liberals aren’t for the most part against capitalism and free markets; but they are not happy with they damage those churning events do to vulnerable people caught up in the churn.

      The EMH assumes that over whatever period of time is necessary, things will correct; critiques rest on the delta of change and its costs to losers in the market through that process. This is a powerful-enough force that the appearance that risk had been eliminated drove the housing bubble; so I suspect that the fear of churn of market correction is a big concern of EMH fans.Report

      • Kim in reply to zic says:

        My critique simply says “most public information is not read by enough investors to actually disseminate”. See ENRON, which put all their scams in the fineprint for investors to read.Report

    • Patrick in reply to Road Scholar says:

      I agree with RS that economics, as a discipline, uses a taxonomy that is (to the outside world) LOADED with normative connotations.

      I suspect that this might be a problem with how economics views its own self, from a neuroscience standpoint… but that’s only a suspicion.Report

      • Kim in reply to Patrick says:

        i want all the people in favor of child labor to start teaching their kids how to be little robber-barons, running their own little sweatshops in other countries.Report

    • James Hanley in reply to Road Scholar says:

      @road-scholar

      Re: efficiciency. I think you’d be interested in my long quote from Noah Smith, below. He comments on the use of the term efficiency in this instance. I snipped most of the quote, but you can follow the link for more.Report

    • Brandon Berg in reply to Road Scholar says:

      In general, that is what it means in economics: Producing the maximum possible output for a given set of inputs and constraints. So we talk about Pareto efficiency (when no one can be made better off without making someone else worse off) and Kaldor-Hicks efficiency (when no one can be made better off without making someone else worse off by a greater margin) and we say that a policy is inefficient if it creates deadweight loss.

      The way it’s used in the EMH is different, but it makes sense to me in a way I can’t quite articulate.Report

      • Road Scholar in reply to Brandon Berg says:

        Well, yeah, @brandon-berg, but do you not see the normative assumptions built into those concepts? Particularly when you start from a base assumption of LF market capitalism and most particularly when you deny the concept of diminishing utility.

        You like to accuse your ideological opponents of not understanding economics when the real issue is that we don’t share certain moral priors that have been smuggled into the discipline.

        So basically I don’t care about Pareto efficiency. KH is more promising if diminishing utility is admitted as a prior. I mean it’s fine to talk about these things in the abstract, but when they become rhetorical points in a political argument then I have to start asking when “efficiency” became a moral value as opposed to a pragmatic measure.Report

      • dhex in reply to Brandon Berg says:

        aren’t you getting a little bit into “it’s called evolutionary THEORY” jesus on a dinosaur gotcha territory with this?Report

    • James K in reply to Road Scholar says:

      @road-scholar

      I can understand your frustration, since we’re talking about 3 kinds of efficiency here, but they all have the same nickname.

      The efficiency you’re talking about is referred to in the econ literature as Technical Efficiency. We don’t really talk about it a lot (its not our area of expertise after all), we just call it that to distinguish it form the other types.

      Most of the time, when economists say “efficiency” they mean Allocative Efficiency. This is about how well market match supply to demand across the various goods and services that exist in an economy. But that’s not the kind we’re talking about here.

      The “efficient” in Efficient Market Hypothesis is Informational Efficiency. This is about how well financial markets aggregate people’s information about the future into a forecast that is a best estimate, given the available data.Report

      • Road Scholar in reply to James K says:

        @james-k, it’s not that I’m really frustrated or confused particularly. I’ve studied econ, I practically, though not formally, minored in it in engineering school.

        Here’s the thing… the word “efficiency” or “efficient” doesn’t really mean anything until you specify what you’re trying to accomplish. So I don’t have much objection to the uses of the term you outlined above since the adjective modifier makes it clear what task you are more or less efficiently pursuing. But regardless of the particular context the word carries a normative connotation such that “more efficient” is better than “less efficient.” And in general we seek more of the good and less of the bad, so it becomes, at least tacitly, a moral value, or at least sounds like one to outside observers who aren’t entirely informed of the disciplinary jargon or the fact that it’s being used in a purely descriptive sense.Report

      • James K in reply to James K says:

        @road-scholar

        This is a common problem when technical jargon escapes into the general populace, case in point: “theory”.

        I do intend to talk about allocative efficiency more, and what it means to pursue it as a policy goal. But I have quite a bit of writing to do before I’m ready to post on that.Report

      • James Hanley in reply to James K says:

        I’m not sure I’ve ever met an economist who didn’t realize that utility was a value.Report

  6. zic says:

    My understanding of EMH stems mostly from my father-in-law, who studied economics in Chicago after WWII on the GI Bill and helped found one of the big finance industries. I spent 30 years debating this stuff with him. And one of the lessons I got was that ‘market’ is, specifically, the stock market. It’s not the housing market; not privately-held companies, not global trade; it’s publicly-traded stocks.

    So I have a great deal of apprehension about suggesting EMH is or is-not proved by the Great Recession; and I would never make that claim; too much of it rested on the message that if you don’t purchase a home now, you’ll be priced out forever, and homes are not publicly-traded stock.

    I would argue that synthetic derivatives, as an opaque market, shielded information from people; it wasn’t just opaque, it could only be read by inference. If it’s proof of anything, it’s proof that to avoid damaging churn in economies, transparency and disclosure are essential; that unregulated markets are inefficient; not that efficient markets are wrong.Report

    • j r in reply to zic says:

      I agree with @zic here. Part of the problem that I see with this EMH conversation is that lots of people seem to want to treat it as a proxy for a whole free markets vs. regulation conversation. And that just doesn’t make much sense.

      When it comes to international trade, immigration, labor markets, I am about as laissez faire as they come. And when it comes to things like utilities or health care or even commercial banking, I favor a very light regulatory touch. Financial markets though, are different and when it comes to financial markets I put a lot – not all, but a lot – of my thinking about other markets on the shelf.Report

  7. Michael Drew says:

    Is the EMH an economic model, properly speaking?Report

    • Chris in reply to Michael Drew says:

      In its weakest form it is not a model, it is just a fairly abstract theory about the relationship between asset prices, namely that they’ll be random. This is why even when there is excess volatility, say, adherents of the weak version can say, “But there is also some randomness, therefore markets are efficient.” There are models that can quantify this theory, with the addition of some assumptions, and the theory was actually developed in light of an actual mathematical model, the random walk theory, in finance.Report

  8. j r says:

    @james-hanley

    I appreciate the vote of confidence, but I have to say that financial economics is a bit outside of my wheelhouse. I am an international econ guy.

    That being said, I tend to think that the EMH critics are over-playing their hand. There is truth to the notion that the idea of self-correcting financial markets had an undue amount of support in the years leading up to the global financial crisis. I stop short, however, of saying that deregulation is to blame for the crash or the resulting rescission. What happened in 2008 was a perfect storm of causality that traces its roots in a dozen different directions.

    My problem with the EMH is that it does not appear to be falsifiable. Making a determination about whether a market is pricing efficiently requires some objective measure, which, as far as I know, does not exist. The only ruler that we have is time. With that in mind, I tend to think of the EMH as being most useful as a method of taxonomy. That is, we can talk about weak, semi-strong and strong market efficiency and use those concepts to identify how one financial market differs from the rest. For instance, holding US Treasury securities is about the closest thing to holding cash, so you would expect the market for US Treasuries to be pretty close to strong market efficiency and much stronger than the market for contingent convertible bonds, for instance.

    As I said above, part of the problem is that people want to use this as a proxy fight in the larger ideological war and that never makes for good analysis. The real debate is not so much about regulation vs. deregulation or left vs. right. The real debate is about to what extent the set of biases and behavioral phenomena identified by behavioral economics renders the EMH false or whether those behaviors are small anomalies that markets work out over time. My guess is that the truth lies somewhere in the middle.

    In practice, I do think that it is almost impossible to consistently beat the market, which is why any investing that I do, I do through passively managed funds. Successful strategies that beat the market are almost never pure price plays. Value investing, for instance, is about picking under-priced stocks, but its also about making money on dividends and by minimizing transaction costs. Even hedge funds that exist to exploit price movements rely on very small price discrepancies that don’t last very long. Think about playing black jack. The odds are in the houses favor, no matter what strategy you use. If you are some kind of walking computer, you can count cards and even the odds a bit, but that is still not a winning strategy. The winning strategy is to except that you are going to lose more than you win, but if you know the count, you can bet more and win more on the hands that are in your favor.Report

    • Michael Drew in reply to j r says:

      It strikes me there may be a bit of motte-bailey action going on on this topic. The M&B model, though, I realized, suffers from an assumption of a unified strategic decision maker, and that doesn’t really apply to the battlefield of ideas as it actually plays out in modern media. There’s not just one actor that speaks for everyone making roughly associated arguments. So at one and the same time, people can truthfully say they are responding to people who were arguing for an undersupported bailey set of positions on a particular topic (even if they don’t realize it’s a bailey and not a motte), while at the same time some on the other side can truthfully say say that they’ve all along only for the motte, and that only the motte is the motte, it doesn’t include any bailey, and they don’t care about any bailey anybody else is talking about.

      I wonder if the EMH has suffered from having a number of ill-informed people arguing for a bailey around it for many years, so that now the debate around it really is mostly on the bailey, to the extent that many arguing for it (the bailey version) don’t even know what the motte is, and likewise those arguing against that bailey. The question becomes, if combatants on both sides tell you they care the most about the bailey they’ve been arguing about, is there really anything wrong with having that argument, even if a label is being wrongly applied (according to the terms of the motte) to the idea being argued over? And if there is something really wrong with it, whose responsibility is it most to correct the terms and substance of that debate? What if the developers of the motte idea, while continuing to personally argue only for the motte, have personally benefitted from the great popular interest that has developed around the debate over the bailey that has grown up around it?Report

      • Tod Kelly in reply to Michael Drew says:

        @michael-drew Your guess here seems simultaneously insightful, likely, and probably inevitable. Put in your terms, I could have said the exact same thing five years ago during the healthcare debate when everyone was talking about how health insurance worked. I wonder if that isn’t what happens whenever we have any fields that require special expertise be part of s public policy discussion.

        I should probably do a broader post on this, but one of the things I am starting to think about a lot is that this notion that the internet making knowledge free and easily accessible leads to a smarter and wiser populace might be very, very wrong. I’m even beginning to wonder if it might not lead to the opposite.Report

      • Michael Drew in reply to Michael Drew says:

        I wonder if I might not tend to agree, and whether I should fear I may be part of the problem. Indeed that we all who want to discuss civic matters on which we are non-expert must inevitably be, to some extent.Report

    • LWA in reply to j r says:

      Most of the arguments over economics in public policy have the motte-bailey problem.
      Economists, whether Keynsian or Austrian, say the motte of X; policy wonks push the bailey of A,B…n based on the logic.

      Think about why we are even having this thread; It was written by a non-economist, and commented on by non-economists. Why? Because public policy, massive decisions about what party or candidate to support, are at issue.

      This is why I keep coming back to a moral theory- about what normative concepts we want to adopt. It really doesn’t matter to me if the EMH is true or not. Its like laypeople arguing over the hockey stick graph- everyone becomes a jailhouse lawyer, an instant expert in matters they are actually clueless in.
      But moral theory is actually accessible to everyone- you don’t need a divinity degree to talk about it- we are all gifted with a sense of moral correctness, of justice and injustice.

      What matters is whether the outcome of policy is just or unjust, or results in greater or less human misery or fulfillment.Report

      • j r in reply to LWA says:

        But moral theory is actually accessible to everyone- you don’t need a divinity degree to talk about it- we are all gifted with a sense of moral correctness, of justice and injustice.

        I am skeptical on this. Most of the time that I hear someone arguing policy from the standpoint of morality or justice, it’s just plain bunk. We all have moral intuitions, but many of those intuitions are subject to the same sort of psychological and behavioral problems that plague economic behavior and many of them are just plain wrong.Report

      • North in reply to LWA says:

        All well and good LWA but it elides the basic problem.
        In order to debate if policy A producing outcome X is morally good and desirable we have to first establish that policy A -will- produce outcome X. It’s in that latter assertion that most of the debate is raging.Report

      • Jaybird in reply to LWA says:

        Let alone the debate of whether forcing someone to follow policy X creates more than a negligible amount of moral wrong (or even enough moral wrong to negate the good caused by policy X).Report

      • LWA in reply to LWA says:

        “we have to first establish that policy A -will- produce outcome X”
        Exactly so; and just like with climate change, minimum wage, or foireign policy there is plenty of need for technical experts to opine.

        But most of the debate isn’t raging among the experts; what underlies most of the debate among laypeople is moral precepts.
        For instance- is the current debate over health care reform one of which approach best reaches the end goal of universal coverage?

        No, it isn’t. The debate is really about whether universal coverage is a desirable goal.
        Likewise, is debate over economics about the best way to reduce the wealth inequality? Or is it whether wealth inequality is a moral good, bad, or neutral thing?Report

      • North in reply to LWA says:

        Maybe we’re reading different articles/threads but all of this EMH debate and economic theory debate seems to me to be people arguing about how economics work; not whether certain economic outcomes are desirable.Report

      • LWA in reply to LWA says:

        We probably are reading different things.
        But just as a hotly debated article on global warming is almost never commented on by climatologists or physicists, the articles on EMH are almost never by economists; they are- (like this one)- written and debated by laypeople, who are bringing their own moral axes to grind.
        The EMH is just a proxy for the real argument.Report

      • James Hanley in reply to LWA says:

        I can’t speak for others but I’m not proxying for any policy issues because with the weak version if EMH that’s all I’m willing to make a defense if, I don’t see any certain policy implications.

        What I’m really targeting is hubris.Report

    • James Hanley in reply to j r says:

      My problem with the EMH is that it does not appear to be falsifiable.

      I think that’s a stronger argument than, and one that excludes, the “it’s been falsified” argument.Report

      • Chris in reply to James Hanley says:

        You are the only one who used the word “falsified” in the comments to the last post.Report

      • James Hanley in reply to James Hanley says:

        2008 was one of the final nails in the coffin of the EMH,

        Of course you didn’t mean falsified, or debunked, or proven wrong, or anything else that might smack of falsification. You meant…?Report

      • Chris in reply to James Hanley says:

        I meant dead, as I said on that thread (seriously, at this point it’s like talking to a child ). Since it’s not falsifiable in its weakest form (something I also said on that thread; see child, talking to), I wouldn’t say it’s falsified. I would say that its viability as a guide to practical models of pricing is nil, and its theoretical utility is, as even its supporters seem to admit, that of a useful heuristic, an ideal observer model against which to contrast actual markets’ various inefficiencies, and perhaps help explain them (is out ltm, herding, risk-aversion, a lack of speculative investment, etc.?).

        Seven years ago its supporters were all semi-strong and this theory is everything.Report

      • James Hanley in reply to James Hanley says:

        I’m just puzzled as all hell how an idea can be dead when so many members of a discipline still think it’s live. And why anyone should take the word of a guy so far outside that discipline?

        Do me a big ol’ favor. Go over to Sumner’s blog, tell him the idea’s dead, and tell him like you told James K that he obviously learned about it in grad school and hasn’t followed it since.

        Let me know how that goes.Report

    • Michael Drew in reply to j r says:

      But isn’t part of what the OP is saying that it wouldn’t really be a problem if the EMH weren’t falsifiable, at least not for the purpose that the EMH is really meant to be used. I.e., that it’s basically just an assumption about markets with perfect information that’s meant to be use to study the effects that imperfect information has on actual markets? Does it make particular predictions about how various kinds of imperfect information (distributions) will in fact affect markets? You can study how information affects markets by starting with the EMH without needing to have it confirmed or falsified that the EMH is correct in what it says about idealized markets. It probably would be interesting to figure out if the what the EMH says about ideal markets ought to be right, but shouldn’t that be theoretical work? (Was Shiller’s critique of it mostly theoretical or was it empirical?) Regardless, it doesn’t seem like a central question for those using the EMH as a background assumption to study information effects.

      What does seem to be worth hashing out are claims by people who aren’t the keepers of the True EMH Flame that the EMH actually predicts that markets will tend to be very close to perfectly efficient, because information in them tends to be good enough to satisfy the conditions to actually make the assumption of the EMH hold for real markets. Can anyone deny that that claims like that, whether inspired by the EMH or in fact claiming that it holds that to be the case (I don;t think that’s all that important a distinction in practice) have come down from many quarters over the years? Is there any reason not to have big fights over them?Report

      • Chris in reply to Michael Drew says:

        Ideal gases.

        “Assume a spherical cow.”

        That sort of thing. Except when it makes its way into pricing models that are used by actual people to allocate money. Then, you know, it’s not just a theory about ideal asset markets.Report

    • Jim Heffman in reply to j r says:

      “My problem with the EMH is that it does not appear to be falsifiable. ”

      Neither is the theory of gravity.

      Oh, sure, it’s conceivable that there could be done some experiment that disproves the theory that masses are mutually attractive. But any time someone seems to have shown this, it turns out that there was some perturbation–electromagnetic force, solar pressure, inaccurate measurement, failure to take frictional losses into account, improper experiment design.

      And it’s conceivable that there’s some reason why prices don’t accurately reflect the rational decisions of consumers based on the information available to them. But any time someone seems to have shown this, it turns out that there was some perturbation–incorrect or incomplete information, unexpressed preferences, government intrusion skewing costs.

      EMH is a law just like the law of gravity. “Oh, but I can’t write an equation for prices the way that I can for gravitational attraction!” Right, but that equation for gravity is the result of a curve fit to experimental data. And it involves a lot of simplifying assumptions; imagine trying to write an equation for gravitational attraction when every bit of mass has its own gravitational constant, and a different power-law relationship to every other bit of mass in existence.Report

      • Kim in reply to Jim Heffman says:

        Jim,
        only for certain scales. There are research papers out there disagreeing with our current formulation of gravity, as it really doesn’t experimentally fit the data for spiral galaxies without having to invoke squirrely things like dark matter.Report

      • Jim Heffman in reply to Jim Heffman says:

        Even those weird ideas and alternate theories still take it as given that mass attracts mass. They aren’t going after the basic notion, just how it’s expressed; what’s constant, what’s not, whether other forces are involved.

        Just like EMH. “Does this behavior appear irrational? Nonsense, all behavior is rational. There must be information that is not apparent to an observer. If you find it, you will understand why that decision was made.”Report

        • So you’re saying that what we have is an apparent phenomenon like gravity – that all prices appear to accurately reflect the rational decisions of consumers based on the information available to them in the same way that mass always seems to attract mass? Btw, is that all the EMH says? That prices respond to consumers who have some information or whatever?Report

      • Jim Heffman in reply to Jim Heffman says:

        Yes, that’s what I’m saying.

        I’m also saying that it’s insufficient in and of itself as a predictor for what prices would actually be, just like “mass attracts mass” is insufficient as a predictor for what will happen when I let go of a watermelon. I figure it will go down, because other things have gone down in similar circumstances. But if it hovers three feet above the ground, I don’t say “oh wow, looks like the law of gravity was bullshit all along”, I look for the guy playing silly buggers with strings.Report

      • Mike Schilling in reply to Jim Heffman says:

        GRAVITATION, n. The tendency of all bodies to approach one another with a strength proportion to the quantity of matter they contain— the quantity of matter they contain being ascertained by the strength of their tendency to approach one another. This is a lovely and edifying illustration of how science, having made A the proof of B, makes B the proof of A.

        The Devil’s Dictionary, by Ambrose Bierce

        (It’s a fair criticism. Classical physics has no explanation for why inertial mass and gravitational mass are identical, though general relativity avoids this problem by describing gravity geometrically.)Report

  9. Mike Schilling says:

    Not everyone agrees, naturally, but I do. I always think of some of the assumptions of physics, such as a universe without friction. We all know that objects in motion don’t really stay in motion forever, that they are in fact affected by real-world forces, but nobody says (I hope), “Newton’s first law is sooooooo stupid.”

    Newton’s First Law:

    An object at rest stays at rest and an object in motion stays in motion with the same speed and in the same direction unless acted upon by an unbalanced force.

    That is, the reason we don’t say that Newton’s First Law is stupid is that we have taken the time to read it all the way through.Report

  10. Don Zeko says:

    This is only somewhat related, but if we’re talking about financial markets I have a question for the group. Does it provide any useful information to anyone for the news to report daily fluctuations in the Dow, Nasdaq, and so forth? If you’re watching financial markets to get a sense of the overall health of the economy, then it’s an indirect indicator at best and any given day will reflect far more noise than signal. if you’re interested in the health of your portfolio, again, you want to know the average movement over the course of months or years, not a day. And if you’re trading stocks in such a short time horizon that a day’s movement matters, you want to know about your particular stocks, not the overall movement of an index.

    As far as I can tell, this is completely useless information, yet it is faithfully reported on a daily basis by all of the major papers, networks, and NPR. Heck, they even frequently provide us with more detail than that, telling us that the market was low in the morning then rallied or what have you. Why is there a consensus that this is a good use of scarce broadcast time?Report

    • Kim in reply to Don Zeko says:

      The DOW is a stupid average, and shouldn’t be used as a measure of the market or the economy as a whole (worse, they do change what they put in it). S&P 500 is okay, and it’ll tell you something about how the war in Mexico is going. Nasdaq is a decent measure of “how is tech doing”.

      I believe these prices,as reported, do some fun things to consumer confidence — which is why your press barons report on them.Report

      • Don Zeko in reply to Kim says:

        My question is whether or not daily reporting on these indexes make sense, not whether the indexes have any value at all. I have no problem with the notion of reporting this stuff on a weekly or monthly basis, similar to how we report things like unemployment statistics.Report

    • Mike Schilling in reply to Don Zeko says:

      The same reason that newspapers carry the baseball scores. It’s what the fans enjoy reading.Report

      • <Tongue-in-cheek>
        If they didn’t have indices to report daily, the automatons that generate the headlines for the business section of the general form “Markets [higher/lower] on [random event of the day]” wouldn’t be able to do their job. I swear that’s how those lead stories get written and headlined. Sometimes it’s some random theory about the market rather than a news event. If the indices drop after three up days, somebody is going to run a story headlined “Markets lower on profit-taking”.
        </Tongue-in-cheek>Report

      • Mike Schilling in reply to Mike Schilling says:

        What are they going to say, “Market up today due to random fluctuations in the aggregation of millions of idiosyncratic decisions”? That would violate the basic laws of physics^H^H^H^H^H^H^H economics!Report

      • Kim in reply to Mike Schilling says:

        Michael,
        You should see what they come up with for King Day. Or Rosh Hashanah, or other days when most of the Live Traders are off.
        “Markets go crazy when we only use automated systems. Will correct tomorrow.”Report

  11. Kim says:

    Alternative:
    http://en.wikipedia.org/wiki/Greater_fool_theory

    Note that nearly all stocks are worth more than their current assets — you also price in future earnings.Report

  12. Chris says:

    A couple things. First, the posturing comment was not to your defending the EMH. That’s all well and good. It was that both you and James, as is often the case in discussions of economics, responded to my saying that evoking the EMH to explain 2008 was ironic, by simply saying that I didn’t understand EMH. That was posturing.

    Second, to this:

    The research program of the EMH is not to prove that markets are efficient.

    So when someone says, “the model’s stupid, people don’t have all the information and markets aren’t efficient,” that is not a refutation, nor even a critique, of the EMH.

    This was all well and good in 1969, when he wrote that, and before the EMH became not only a research program but accepted dogma in actual practical (not theoretical or merely research) models of financial asset pricing, and therefore actually drove behavior. There is a reason some people (and not just Krugman) blame the EMH for 2008, and that’s because those models. For an example of a description of the embeddedness of the EMH (from):

    One can note that the market efficiency hypothesis, derived from the no arbitrage assumption and implying that prices reflect all available information, is retained by all major theoretical models used in finance, including the Capital Asset Pricing Model and the Arbitrage Pricing Theory, and represents the basis for the pricing of all financial securities, including that of all forms of derivative products. Translated into corporate finance, the no arbitrage assumption gives rise to the famous Modigliani-Miller results, which actually were at the origin of the whole no-arbitrage concept. The world of efficient markets is thus a world of constrained Pareto optimization, where both the perfect expectation and no-arbitrage assumptions prevail and where enterprises are indifferent between financing themselves by equity or by debt.

    So people aren’t saying the model was stupid, people are saying the model led people to behave suboptimally, and suboptimally in a way that had dire consequences. Now, you can dispute that, though you haven’t really done so here, but saying, “It’s just a research program” with a quote from the 60s is just wrong.Report

    • James Hanley in reply to Chris says:

      If you want to say the whole thing was falsified, then you have to deal with the whole thing, not just the parts that suit you.

      I’m not the guy that tried to lump all of it together.Report

      • Chris in reply to James Hanley says:

        Excess volatility and short and potentially long-term correlations show that asset prices don’t follow a random walk, and bubbles show that not all public information is in the prices (this is why people can see that there’s a bubble). If all you want to do is say, “OK, now it’s just a research tool, a heuristic, not something we actually believe and we’re not really going to set prices based on it because after 2008 we’d be insane to do so,” that’s cool. I doubt even the most ardent anti-EMH economists would care. Hell, they’d probably agree with you that it’s pretty useful in that regard, particularly if you can come up with a measure of relative efficiency.Report

      • James Hanley in reply to James Hanley says:

        Excess volatility and short and potentially long-term correlations show that asset prices don’t follow a random walk, and bubbles show that not all public information is in the prices (this is why people can see that there’s a bubble).

        I’ve addressed all that, but instead of addressing my response to it you just repeat it as though it’s self-evidently true. An intellectual tour de force.Report

      • Chris in reply to James Hanley says:

        As far as I can tell, you don’t touch on long-term memory or short term correlations really at all, and your own comment on volatility amounts to this:

        Krugman writes as if the volatility of stock prices alone disproves market efficiency,

        It’s not the very existence of volatility, it’s excess volatility, volatility that causes deviations from randomness, that is the problem for EMH models.

        I am not an expert enough to educate you about these things myself, but ’cause I’m a nice guy, I’ll do you a solid and link you to an actual lecture on the subject by Schiller himself (powerpoint and transcript here):

        Now you’ll know what you’re talking about. Enjoy. (P.S., you can watch all of the course lectures there, I believe. I watched his 2011 course series a couple years ago, which includes another discussion of EMH, embedded below.)

        Report

      • James Hanley in reply to James Hanley says:

        It’s not the very existence of volatility, it’s excess volatility

        Maybe. I’m willing to accept that as a possibility. I’ll take a look at it. But I want to note the difference between us here: Shiller says it, and you take it as indisputable truth. The EMH folks say no, but I’ll take a look at it and consider it.

        You do understand, don’t you, that Fama and Shiller co-won the Nobel the same year for conflicting theories? So just pointing out that Shiller says it is not sufficient to settle his argument with Fama I could as easily just point to Fama and act like his saying something is sufficient to settle his argument with Shiller. But neither of us was trained that way.

        Let’s flip this around, and talk in the abstract for a moment. Two high profile psychologists, Smith and Jones, with conflicting theories, each claiming to have a considerable amount of evidentiary support, each with lots of other psychologists in agreement with them. An economist reads up on them to some extent, enough to be an educated amateur, but not enough to claim actual expertise–admittedly not enough to educate others–and pronounces with absolute certainty that Smith’s theory–still supported by a substantial number of respected psychologists–has had the final nail driven in its coffin, for this set of reasons given by Jones. Another psychologist who thinks Smith’s argument still has some value says to the economist, “you don’t understand this very well” (that’s the proxy figure for James K, or a Sumner, etc., not me–in this scenario I’m just another educated amateur; that is, I could be dead wrong ). The economist replies, “You must not have studied this since grad school, because you obviously don’t know what Jones says.”

        What are the odds the economist in that story is right? What are the odds he really understand the whole debate better than the psychologist? What would you wager in his favor, given even odds? When the economist admits he can’t really explain the theory or rebuttal of it well, but it’s indisputable that he’s right, are you going to think he’s acting like a professional scholar?Report

      • Chris in reply to James Hanley says:

        Schiller and Fama are not the only participants.

        It’s also a great illustration of the weirdness of economics as a science: economists get taught that it is the most empirically supported theory in economics, or even all of the social sciences, while the data and conceptual arguments against it are easily found, not because it is in fact the best supported, but because a particular theoretical paradigm is so dominant that they can get away with it. It’d be like going to school in France in the late 17th century and being told that Cartesian vortex-based astophysics were established empirical reality even as Newton was telling the world how he got that bump on his noggin.

        OK, that’s hyperbole, but given the economists’ penchant for delusions of the discipline’s scientific grandeur, I don’t feel bad about it.

        Chicago is the late 20th century equivalent of the Hegelian in the German universities during the fist half of the 19th century. That’s not hyperbole.

        And remember, this started because I sneered at James for suggesting that the EMH explained ’08, when I think it’s pretty clear now that ’08 only made the EMH a more contentious issue among the actual experts (if you look at stuff from the 90s, you’ll already find puerile calling it one of the most contentious issues in economics, even if they forgot to note this in New Zealand econ programs; there’s a reason they both won the fake Nobel). Evoking it the way he did pretty ironic.

        Also, the story Dave linked is pretty good, if you haven’t read it.Report

      • James Hanley in reply to James Hanley says:

        Schiller and Fama are not the only participants.

        And of course no one has suggested otherwise.

        I sneered at James for suggesting that the EMH explained ’08,

        Which is not what he said.

        when I think it’s pretty clear now that ’08 only made the EMH a more contentious issue among the actual experts

        Sure, no disagreement on that.

        there’s a reason they both won the fake Nobel).

        Nobody said differently. You’re the only one who’s condemning just one side.

        even if they forgot to note this in New Zealand econ programs;

        The analogy was designed to get you to step outside yourself and see how astounding your hubris is.

        And that’s really the issue here; not that you don’t find the EMH arguments convincing–you’ve got respectable company there–but that as a scholar in one discipline you feel competent to pronounce so authoritatively upon another discipline to which you’ve given minimal formal study. That’s not being scholarly.Report

      • Kim in reply to James Hanley says:

        Chris,
        as far as I know, that seems to mirror the schools of CBT and Freudian psychoanalytics in psychology/psychiatry. Ne’er the twain shall meet, and each one is firmly convinced the other is just plain wrong.Report

      • Chris in reply to James Hanley says:

        Well, you’ve gone from saying I don’t have any idea what I’m talking about to agreeing with the initial point that led you to say it, and merely saying I’m not being very scholarly. So at least that’s progress.Report

    • Mark Thompson in reply to Chris says:

      So this is going to be a “talk to me like I’m stupid” question since I don’t understand this debate enough to have much of a position; if pressed, I’d say that the EMH, at least the softer versions, is obviously true to the point of being close to a tautology, but not really useful outside an academic context – a measuring stick or perhaps a taxonomy as @j-r indicates.

      In any event, my question is why the EMH is or was necessary for the “major theoretical models used in finance, including the Capital Asset Pricing Model and the Arbitrage Pricing Theory,” such that it “represents the basis for the pricing of all financial securities, including that of all forms of derivative products.”

      It seems to my completely uninformed mind that, in my very minimal understanding of these concepts, any reliance on the EMH would mean looking to the current market price for the asset in question. Any calculations stemming from that market price would be based on assumptions outside of the EMH, right?

      So my question then is: why do we need the EMH to use current market price of an asset? Isn’t the current market price sort of the only reasonably objective and readily calculable measure, however imperfect it may be?

      The paper you link to concludes by indicating that there is a need for an alternative finance theory, but doesn’t indicate that any such alternatives exist, much less indicate why such alternatives would be more consistently reliable. So if the problems stemmed from relying too heavily on current market pricing to value assets and we don’t have any proven, objective, transparent, and readily calculable alternatives to current market pricing, isn’t it more likely that we’re relying so heavily on those prices because they’re the only thing we have to an objective, transparent, and readily calculable valuation than it is that we’re relying on them only because of a less-than-50 year old academic theory?

      I know, for instance, that the term “fair market value” has been around in the law for at least 100 years, has always been defined as a price that a willing buyer would pay to a willing seller, and has always seemingly been understood to have a strong preference for then-current publicly traded market prices where those prices were available.Report

      • Chris in reply to Mark Thompson says:

        I’d say that the EMH, at least the softer versions, is obviously true to the point of being close to a tautology, but not really useful outside an academic context

        To the extent that the EMH says that prices reflect available information, it is tautologically true. In well-regulated markets, that’s what prices do (as opposed to being set by the government, say, or being arbitrary, or involving cheating as in a Ponzi scheme, say). However, that’s not all that the EMH says. It also says that it contains so much information that it is impossible to predict future prices from historical (including current) prices. That is, that future prices will vary randomly (there are different ways to quantify this, and statistically it turns out to be a really difficult quetstion to pose, much less to answer, which is why there’s still any controversy at all theoretically). In this case, it is only tautologically true if your definition of price is that it includes all information.

        As for an alternative finance theory, there are fairly well specified, but perhaps not well-tested models (I’ll let someone who actually studies finance talk about them, ’cause I don’t know them well, just that they exist via discussions of the EMH). Generally, there are the behavioral models (e.g. this, this, and this) that James mentions, as well as “technical analysis,” but these can lead to multiple different quantitative approaches. Quantitatively, the random walk model (essentially the EMH quantified, though it predates the EMH) has as competitors evolutionary models (e.g., models that predict that markets become more efficient over time), models with long-memory (because one of the ways it appears stock prices violate efficiency is over long periods, according to these models because prices have a long-term effect on future prices), fractal models, and from what I can tell, a battle between the EMH models and non-EMH models on what accounting for risk aversion means. I can’t say much about any of these, because I only know of them from discussions of EMH. I read the after-battle reports (for lay folk), I don’t watch the battles as they play out.

        Man, I hope all that html works.Report

      • Thanks, @chris – first, the HTML worked, so well done. I guess I’m still not understanding why the EMH view that future prices are unpredictable would impact the above-referenced models beyond telling them to use current market pricing as a starting point. In fact, wouldn’t EMH hold that doing anything more than that to value an asset is folly?

        Again – this is a talk to me like I’m stupid question. I minored in Econ and generally get Econ theories, but the financial markets are generally beyond me. My general sense in that department is that those markets really are unpredictable and that the very well paid analysts (many of whom I went to college with) whose job it is to beat the market really aren’t very good at that.Report

      • Chris in reply to Mark Thompson says:

        Man, they’re beyond me, too.

        The weak version explicitly says that you can’t predict future prices from current ones (that’s relative, of course: tomorrow’s prices will be close to today’s most of the time, it will just fluctuate in unpredictable ways).Report

    • Dave in reply to Chris says:

      @chris

      I think this goes to what you’re saying.

      http://www.ritholtz.com/blog/2013/10/how-shiller-helped-fama-win-the-nobel/

      Also, while I haven’t kept up on the reading over the years, there seem to be some significant discrepancies between the way some economists look at this and people in finance, especially money managers that reject the random walk theory root and branch.

      I enjoyed the post and will try to comment on it later. I’m out and about today.Report

  13. Jaybird says:

    The problem that I see with the EMH is that it does a lot better job of accurately reflecting reality when nothing else that is interesting is going on.

    When the only people playing the stock market are the Sunday morning show talking head types, it does pretty good. When taxi drivers are giving passengers stock tips? We are now in crazytown and the EMH no longer applies. The dumb money creates churn and irrationality (and the market can remain irrational longer than something something) and only once the dumb money has settled (IRAs, for example) then can the market get back to having the only players be the Sunday morning show talking head types and the EMH goes back to applying.

    The problems in the last 30ish years or so (1986, the internet bubble, the housing bubble) were all caused by dumb money.

    So when the dumb money gets in, you should probably get out. Well, not right away, of course… there’s going to be a wave going up and you’re going to want a piece of that.Report

    • Chris in reply to Jaybird says:

      The problem that I see with the EMH is that it does a lot better job of accurately reflecting reality when nothing else that is interesting is going on.

      This is when stocks do follow a random walk: when absolutely nothing is going on.Report

      • Jaybird in reply to Chris says:

        Well, it’s one of those things where once the lumpenproletariat learns about the EMH, the EMH ceases to accurately reflect the market. Once the lumpenproletariat lose their money, the EMH slowly settles back into an accurate reflection.Report

  14. James Hanley says:

    Here’s a nice blog post that avoids being either a pro or anti EMH idiot, and gratifies me by touching on the behavioral econ stuff that I talked about.Report

  15. Michael Cain says:

    Warning, anecdote ahead. Ten or so years ago, I had some money that needed to be invested. Playing with historical data for the mid-cap stock index, I found a timing pattern which allowed movement in and out of the market about twice per month, using mutual funds that had been constructed to allow trading at that sort of frequency, which produced market-matching returns but with much lower volatility. I traded for three years using that model, with the same result (market or better returns, much lower volatility) before the pattern disappeared back into the noise.

    I looked at it from a bunch of different statistical directions. The most telling, IMO, was the result that said the probability that the returns for the days I was in the market and the returns for the days I was out of the market were drawn from the same distribution was very small. I ran my analysis past a couple of friends who are practicing PhD economists. Both agreed the pattern was real, and clearly violated all of the forms of the EMH. One of them traded a piece of his retirement using the model for the same three years I did. The other suggested that, based on when on the calendar the model had me in the market, it was predicting periods when the big mutual funds were likely to be moving money in and out.Report

  16. clawback says:

    But it seems to me that the model incorporates the ideas of bubbles and busts–they’re part of the random walk of the market

    Can you provide evidence that markets in general exhibit random walk characteristics? Because most analyses contend the distributions have fat tails — and precisely because of the bubbles and busts.Report

    • James Hanley in reply to clawback says:

      I think people are–understandably–overinterpreting the “random” concept. I just came to this realization that some people are reading the term more precisely than it’s necessarily meant, and it’s really the fault of writers (including me now, I guess) who aren’t explicit that the term is being used loosely. That is, random walk is used casually, but as far as I can tell, it’s been known that it’s not truly random in a mathematical sense, so much as just not readily predictable.

      As a relevant addendum, see my extensive quote from Noah Smith, below.Report

  17. James Hanley says:

    Here’s Noah Smith, a guy who often tends to disagree with people I agree with, on the EMH. I’m just grabbing a few quotes, but I recommend his whole post. He’s another person who’s taking a balanced approach on it.

    The “Efficient Markets Hypothesis” is a popular target of anger and derision among lay critics of the econ profession. How can financial markets be “efficient” when they just crashed and took our economy down with them? And when sensible people like Bob Shiller, Nouriel Roubini, Bill McBride, et al. were screaming their heads off about a housing bubble years before the pop?

    Of course I have some sympathy for these complaints. But the more I learn about and teach finance, the more I learn what an important and useful idea the “EMH” in fact is. I don’t want to say that the EMH is unfairly maligned, but I do think that its vast usefulness is usually ignored in the press.

    First of all, people should realize that the EMH is misnamed – it’s not really a hypothesis, it’s not about “efficiency” in the economic sense of the word, and it’s not unique (so it shouldn’t have a “the” in front of it). Some of this miswording was just semantic clumsiness on the part of the people who came up with the theory. Some was sloppy science.

    The “efficient” part of “EMH” doesn’t mean that financial markets lead to a Pareto-efficient outcome. …

    The “efficient” actually just refers to information-processing efficiency. What that basically means is that if there’s some piece of information out there – some fact about a company’s balance sheets, or some pattern in past prices, etc. – the market price should reflect that piece of information. That’s what “efficient” means here.

    But exactly how should prices reflect information? Here’s the bigger problem with the term “EMH” (the “sloppy science” part) – it’s not really a hypothesis. How prices reflect information will always depend on people’s preferences. In finance, preferences include preferences about risk. So without a measure of risk, it’s impossible to scientifically test whether or not prices incorporate information. To be a real hypothesis, the EMH needs to be paired with a specification of risk (or, more generally, a hypothesis about people’s preferences with respect to uncertainty and time, and a hypothesis about the sources of risk). And since there are many possible such specifications, there isn’t just one “EMH”…there are infinite.

    To complicate things, “the EMH” says nothing about how long it takes for the market to process information. So even if an EMH happens to be true at one frequency (say, daily), it might not be true at the 1-second frequency.

    And to top it all off, theory says that the strong form of the EMH can’t even be true.

    So “the EMH” is very limited as a scientific hypothesis or physics-like law of nature. And I think that ever since many of these points were pointed out … financial economists have stopped talking about “the EMH” as such, except in a vague hand-wavey way during informal discussions. Sloppiness has been much reduced.

    But I do seem to recall that the title of this post was “In defense of the EMH”. So I had better get around to defending it! What I want to defend is the idea behind the EMH. Even if the data rejected every single EMH, the idea would still be incredibly useful for the average person.

    Let’s call this idea the Random Markets Idea, or RMI.

    … Basically, if it was pretty easy to see where prices were headed, a lot of people would see it, and try to make free money by trading on it. Since people in the finance industry are doing a lot of work – watching the news like a hawk, doing constant analysis of changing numbers – chances are that the price change will happen so fast that you won’t have time to get in on the action. So from the perspective of any of us who doesn’t have a supercomputer in his head, prices movements must be unpredictable and surprising. They must seem random.

    That’s it. That’s the RMI. Note that this is very different than saying “On average, people don’t beat the market average.” This is more than that. This is saying that even if you manage to beat the market average for a year or two years or even ten years, you shouldn’t expect to be able to repeat your performance next year.

    That may seem counterintuitive, or even silly. “Hey,” you think, “I beat the market last year, so I must be one of the smart guys! And that means I should be able to repeat my performance…right?” Well, maybe. But the RMI says that that’s actually very, very unlikely. It’s far more likely that you just got lucky.

    Now here we get to why the RMI is so useful to you and me and most people (and to the managers of our pension funds and mutual funds). It provides a check on our behavioral biases. Probably the most robust findings in the field of behavioral finance is that individual investors do badly. They are overconfident. They trade too much and take losses on trading costs. They suffer from biases like disposition effect, probability mis-weighting, recency bias, etc. And as a result they lose money, relative to the wise folks who just stick their money in a low-cost diversified portfolio and watch it grow. As for institutional investors – mutual fund managers and pension fund managers – we don’t know as much about what they do, but we do know that very few of them manage to consistently beat the market, year after year (and most don’t beat the market at all).

    It’s interesting to note that people usually think of behavioral finance as being an alternative to efficient-markets theory. And sometimes it is! But in the case of personal investing – i.e., the single most important way that you will probably participate financial markets – the two ideas support each other. The RMI says “You can’t beat the market”; behavioral finance says “But you’re probably going to lose your money trying.”

    Of course, even the RMI isn’t quite true. There are some people – a very few – who correctly guess price movements, and make money year after year after year (I work with a couple). But you’re very unlikely to be one of those people. And your behavioral biases – your self-attribution bias, overconfidence, and optimism – are constantly trying to trick you into thinking you’re one of the lucky few, even when you’re not.

    The RMI is an antidote to this! Just remind yourself that market movements should be really, really tough to predict. Then, when you start to think “It’s so so so obvious, why can’t people see AAPL is headed for $900, I’m gonna trade and get rich!”, you’ll realize that no, it can’t be that obvious. And you’ll restrain your itchy trigger finger. And when you start to think “My money manager is awesome, he beat the market the last 5 years running, I’ll pay his hefty fee and he’ll make me rich!”, you’ll stop and realize that no, it was probably luck. And you’ll think about putting your money in index funds, ETFs, and other low-cost products instead.

    Report

    • Kim in reply to James Hanley says:

      Seems like an investor ought to use the EMH as a null hypothesis, and require some decent confidence that it isn’t happening, before he chooses to invest. (cited example about lead painted toys — look at time-averaged (because random) stock price both before and after release of data. If there’s no statistical difference, then one can reasonably conclude that the market hasn’t priced in a good “lead toys” national news story)Report

    • Smith is awesome. He does (obviously) tend to take up for the basic ideas of his profession against outsiders from the left pretty often, so this isn’t unusual for him. He writes this kind of post pretty often. When he’s warring those on the right in his profession it’s (seemingly) done mostly within the bounds of professional disagreement inside the field.

      There is no econblogger who I think does a better job at it than Noah Smith.Report

    • So here’s my stupid question:

      If the EMH (or RMI) is right, and it means basically “You shouldn’t think you can beat markets that incorporate good information in the future, basically ever,” what are we supposed to learn from that more than that we’re probably idiots or at least irrational to decide to pick individual stocks or other assets rather than buy market indexes? Smith says that, “personal investing [… is] the single most important way that you will probably participate [in] financial markets. But is that actually what all the storm und drang over this is about? Have people claimed there are broader lessons to be learned from this that other people take as more significant for them than the strict investment advice it seems to give (since to the extent they invest, many probably already do incorporate Fama the Younger’s insight just as a matter of employing a competent investment advisor)? Should people make such claims about broader lessons (possibly – duh duh DUH – policy lessons) to take from the EMH? Should we take such claims seriously?

      I’m not sure Smith’s discussion there really tracks or addresses the stakes in the discussion he is addressing that people are actually interested in on this topic.Report

      • James Hanley in reply to Michael Drew says:

        I hope James K writes his post on EMH, but I’m not sure if there is much more to learn. Strong endorsement, eh?

        But it’s an important enough lesson on its own, perhaps. I’m too lazy to look it up again right now, but one blogger I found talked about teaching finance students, who all think they’re going to make their gajillions beating the market. I think Smith’s on board with that. If EMH does nothing more than make us aware of how cognitive biases mislead us into thinking we can regularly beat markets it’s still a valuable theory, and certainly not “dead.” But I’m just a political scientist, so maybe I’m wrong and economists like Noah Smith really should let a psychologist tell them what theories they should and should not teach their finance students.Report

      • But I think people nevertheless have acted through the years like there is much more to learn. That’s where I think all the heat around it comes from. But maybe I’m concluding from the fact that there’s so much heat around it that people have acted that way (and said as much). Do you have an opinion on that?Report

    • To complicate things, “the EMH” says nothing about how long it takes for the market to process information. So even if an EMH happens to be true at one frequency (say, daily), it might not be true at the 1-second frequency.

      There are people being paid handsomely to write software that recognizes new patterns that may persist for only a few seconds, trade on them, and then look for the next pattern. To be competitive in that business requires you to beat the “slower” people to the trade. It’s profitable enough that some of the big high-frequency traders in London and New York paid for their own trans-Atlantic fiber cable because it would be 5.2 milliseconds faster than the cables everyone else uses.Report

      • morat20 in reply to Michael Cain says:

        In first-person shooters, those people are called “LPBs” (low-ping bast*rds) because their shorter information loop gives them a critical edge.

        That doesn’t even get into the stuff the real HFT’s are doing, the ones who actually work inside the regular trade loop to steal money. I’m sorry, I mean “provide much needed liquidity”. It just looks like a table rake to the uninformed, because apparently providing that liquidity is ridiculously lucrative.

        Seriously, I don’t know what you’d call the US stock market, but “transparent” does not come to mind. “Gambling” does.Report

      • “Gambling” does.

        It’s really kind of surprising how much financial literature — at least popularized accounts — openly refer to the lead-up to the 2007-8 financial crisis as too many banks and shadow banks “all making bets in the same direction.”Report

      • Kim in reply to Michael Cain says:

        Michael,
        the problem was, they got so busy betting they forgot to figure out who was the designated bagholder. And give them a golden parachute.Report

    • Mike Schilling in reply to James Hanley says:

      First of all, people should realize that the EMH is misnamed – it’s not really a hypothesis, it’s not about “efficiency” in the economic sense of the word, and it’s not unique

      In other words, the Habsburgs should be running it.Report

  18. Mad Rocket Scientist says:

    Nicely done, James. I wish I could add something to the conversation, but this is pretty far outside my wheelhouse at the moment (except insofar as it treats modeling in general).Report

  19. Major Zed says:

    Picking up on something Chris said: “reflect[ing] available information… [is]… what prices do (as opposed to being set by the government, say…)” Let’s not forget that one of the most important prices in the financial system, the short-term interest rate, is set by government fiat, not by the action of buyers and sellers. Since adding full employment to its mandate (previously just low inflation) the Fed has become a serial bubble machine.Report

  20. James Hanley says:

    For the interested, an Austrian critique of EMH. It’s both readable and non-whackadoodle (which isn’t the same as claiming it’s necessarily correct).Report

    • Chris in reply to James Hanley says:

      Looking back, the first time I ever heard of the EMH was around ’03, and it was in a blog post by an Austrian pimping his favorite stock prediction technique. Looking at his blog now, it looks like he’s no longer pimping it (he used to have a huge chart on one side of the blog), but I ain’t going to link to it, ’cause he’s an odious human being (not because he’s an Austrian, though that doesn’t help).Report

  21. Road Scholar says:

    I think perhaps the EMH was more applicable when it was formulated than it is today. Allow me to grope toward a point here. As stated, the EMH seems to assume the kind of investor who is primarily interested in accurately determining the “true” value of a company and detecting discrepancies between that valuation and the market valuation.

    But is that what most active traders are really trying to do?

    When capital gains taxes and the top marginal rates were cut in the 80s, the focus of investor activity shifted somewhat from value seeking to profiting from trading itself. The penultimate example of this being the totally automated HFT that basically tries to do a Maxwell’s Demon thing on the stock market.

    So now the goal isn’t about accurately determining the value of a company so much as it is about accurately determining and predicting what other traders are thinking. After all, if you’re only maintaining a market position for a few milliseconds what the hell do you care about fundamentals?

    I’m not sure what all this means wrt to the EMH, but it at least seems like the market we have now is a different beast than the one Fama was contemplating.Report

  22. James K says:

    Thanks for posting this James, to avoid informally starting an Efficient Markets Symposium I won’t do a full EMH post, but here’s the quick version:

    What is it?
    The EMH is a set of three hypotheses regarding how well financial markets process information. In particular, they ask what you would need to know to be able to construct a trading strategy that would out-perform the market on average (that is to say, out-perform the market without just getting lucky).

    What is it Not?
    The EMH (even the strong form) does not assert that markets are infallible. Markets can only know what their participants know.

    The EMH does not assert markets are rational. Rational markets would be sufficient for efficient markets, but they are not necessary. For a bubble to violate the EMH, the bubble’s behaviour would have to be predictable enough that you could plot out a trading strategy that was more lucrative than buy-and-hold.

    The EMH has little to say about the relative merits of more or less government intervention in financial markets, or any other markets. I’ll talk more about policy implications later, but there aren’t very many.

    What are the three forms?
    The EMH comes in three forms: Weak, Semi-Strong and Strong. Each places increasingly strict limits on what information is needed to beat the market:
    1) The weak form says that you can’t beat the market by using it’s own past to predict its future. SO you need current events to beat the market, not old news.
    2) The semi-strong form says that anything that is publicly known is already included in the price, so to the beat the market you need to know something the other traders don’t.
    3) The strong form rules out beating the market at all, except by accident.

    Is it falsifiable?
    Yes. Each form is tested a little differently.

    1) The weak EMH is tested by looking for regularities in the time series characteristics of share prices. Finding usable regularities is evidence against the EMH.

    2) The semi-strong EMH is tested by looking at how prices respond to unpredicted events. The EMH predicts that unpredicted events will show a sharp rise (or fall) in repose to the event, but that’s it. A piece of news should not cause a sustained change in the trend of a share price. This is hard to test with just one event, which is why you look at a lot of events at once. Widely-expected events should show little-to-no effect on prices.

    3) For the strong EMH, the only events that should show a price response are those that were utterly unexpected by anyone. Insider information being made public should not affect prices.

    What are the empirical results?
    The weak EMH has rock-solid empirical support, the strong EMH doesn’t hold up, and while the semi-strong EMH is supported on balance, though the evidence doesn’t all go one way. Admittedly something could have change the 9 years since I left uni, but the evidence was pretty settled at that time.

    What are the policy implications?
    Few. The only thing it argues against is trying to intervene in a market by responding to a crisis just before it happens, because that requires timing the EMH won’t permit you to have. The strong EMH would also suggest that banning insider trading is a waste of time, but as I noted there isn’t a lot of empirical support for the strong EMH.

    So why does everyone care about the EMH?
    Hell if I know.

    If you are an academic economist, or a fund manager or possibly an official in the SEC or Fed you have some reason to care about the EMH. Why everyone else makes so much fuss about it doesn’t make a lot of sense on the merits.

    My guess is that it has something to do with the name. As @road-scholar noted upthread economists have some specialised definitions of efficient and I suspect this has led to people drafting the EMH into the endless “Markets, white magic or black magic?” debate that seems to consume all discussions of economic policy.Report

    • Michael Drew in reply to James K says:

      Thank you , @james-k !Report

    • Michael Drew in reply to James K says:

      ..,Another way to come at it might be to ask what it would mean if a market is not efficient in this sense. From the definitions you give (and from other definitions I’ve seen), it looks like this is often defined in terms of “beating the market.” So: if the market is “beat”able (with the various kinds of information pertaining to each form of the hypothesis) then it’s not efficient in this sense. But “beat the market” seems a surprisingly nontechnical criterion on which to hang what sounds like a pretty technical hypothesis. What is it, quite exactly, to “beat” a market in the sense being used in these statements of the hypothesis?Report

      • James K in reply to Michael Drew says:

        @michael-drew

        “Beat the market” may sound colloquial, but a lot of finance jargon is like that. Basically, beating the market means that you can construct a trading strategy that will allow you to get higher average returns than a broad-based market index without taking greater risks than holding a market-based portfolio would require taking.

        If markets were weak-inefficient then you’d be able to forecast market prices, and econometricians would be rich. If markets were semi-strong inefficient then reading the financial section of the paper would actually help you make money in the stock market.Report

    • James Hanley in reply to James K says:

      Thank you, James K. I always find it valuable when you chime in, which we all know you do too little.Report

    • Road Scholar in reply to James K says:

      Thanks, @james-k , for this most excellent primer. And I don’t see a lot here that strikes me as particularly redoubtable. I would note that HFT seems to partially defy the weak version on the shortest time scales but maybe not since some things I read on it seems to imply that the algorithms are utilizing what amounts to private information.Report

    • zic in reply to James K says:

      So why does everyone care about EMH?

      Because anyone lacking understanding that 1) EMH is about publicly traded stocks, where there are certain levels of disclosure required, and 2) that intervention after a bubble forms won’t prevent the bubble allows a host of nonsensical arguments to be made. For instance, the problems in the health-insurance market will be solved by the market. Or regulation of any sort distorts free markets. But you get my drift.

      Since this became a near-symposium, I’ve been considering the difference between venture/angel investing and stock-market investing. We tend to think of that difference in terms of risk; a mutual fund is a relatively safe investment, an angel investment a highly risky investment. I don’t know if these numbers have changed, but in the 1980’s and ’90’s, about one in seven venture investments were profitable. The other six were losers. If you want to do angel or venture investing (which requires you to have access to large amounts of liquid capital that you can risk losing six out of seven times,) you do due diligence, and the most successful investors do really thorough due diligence. Having spoken to entrepreneurs who’ve been through the process, they felt it pretty invasive; they not only revealed their life in a way that might put an FBI background check to shame, but they gave up significant amount of control over the fate and direction of their company in exchange for the working capital venture investing provides.

      This matters because there are different types of investment markets with different types of risk; and EMH (as a non-stock theory) fails to get at this, while it’s used to argue against market regulation. For instance. Consider the risk between public disclosure of publicly-traded companies, where we can read the annual and quarterly SEC filings, where there are boards elected by share holders, and some level of public information readily available, both about the particular company and the market sectors it occupies, vs. venture investing, where much, if not all, of this information is not available. EMH, applied venture investing doesn’t work. The market isn’t self correcting because the information at the root of the market isn’t public; which is why good VCs and Angels do that due diligence, and still lose out six out of seven times.Report

      • James Hanley in reply to zic says:

        Because anyone lacking understanding that 1) EMH is about publicly traded stocks, where there are certain levels of disclosure required, and 2) that intervention after a bubble forms won’t prevent the bubble allows a host of nonsensical arguments to be made.

        Lucas and Fama are such idiots, aren’t they?Report

      • zic in reply to zic says:

        I didn’t say that, did I? I said the fools were the people who used EMH to explain non-publicly-traded markets.

        Have you ever taken a company public? Sold a company? Split a company? Settled an estate invested in private companies? Taken a company through bankruptcy?

        Valuation’s a bitch, and doesn’t comport with EMH as far as I can tell, even though there are many things common to both pricing methods. EMH may work well and explain publicly traded companies and stock markets, but that’s about it. Yet it’s used all sorts of laxness in corporate oversight because “the market’s efficient.”

        Most companies are not publicly traded; Hobby Lobby’s a good example much discussed here. Were it publicly traded, it would not have even brought it’s case over contraception to the SC. Should it ever go public, it will lose it’s religious exemption from complying. When the Greens eventually die, their heirs will have to value the company for estate-tax purposes; but it will still not be public information; and should they opt to do an IPO, the public-market price, using available information, will be determined some time after the IPO — the price the shares actually trade at once the IPO-hype’s calmed down. WalMart, on the other hand, is publicly traded, and it’s stock prices will reflect information about how employees are compensated, how it’s supply chains are running, etc.

        Using EMH to discuss HobbyLobby is pretty much meaningless while using it to discuss WalMart is probably worthwhile. We don’t know HL’s actual value; we do know WalMart’s. We have some notion of how both companies might rent seek, how they might deliver value, etc., but WalMart’s successes and setbacks are both divulged in their SEC filings; HL’s are private information.Report

      • James Hanley in reply to zic says:

        I didn’t say that, did I? I said the fools were the people who used EMH to explain non-publicly-traded markets.

        Oh, so the only thing wrong with EMH is that non-economists didn’t understand it?

        Valuation’s a bitch, and doesn’t comport with EMH as far as I can tell,

        Oh, so the problem is economists after all, for supporting a model that doesn’t comport with valuation?

        I’m so confused now. How about you?Report

      • zic in reply to zic says:

        @james-hanley actually, you are confused. Pity.

        EMH talks about this (significant) but limited pie. It’s used to describe all the pies. And that’s foolish. I did not attack EMH anywhere on this thread, btw; I suggested that it’s used wrong (as proxy for free markets,) when it’s use should be constrained to publicly traded.

        Valuation, due-diligence, and a host of other things need to be don to replace the information about publicly-traded companies; they’re examples of why all the fruit in the basket aren’t berries with their freshness and blemishes easily observable. Some of the fruits have rinds and peels, and you have to crack them open. These not-so-obivous fruits also happen to be the companies that may prove to be entrepreneurial and grow quickly.Report

  23. 3dcheckers says:

    Do you really think that if Warren Buffett relinquished all his duties at Berkshire Hathaway and managed a portfolio of $1,000,000 he wouldn’t beat the market over the next 5 years? Wouldn’t that in itself disprove EMT? Also when you consider all the signs that someone is operating at an elite level for their field (for example he read every book on investing in the Omaha library by age 10) isn’t it hard to attribute his past performance to luck?Report