How did Krugman get it so wrong?
The title is a play on Paul Krugman’s recent NYT piece “How did economists get it so wrong?” Krugman has a lot to say, but because he dances around the two key issues his essay doesn’t add up to much.
Krugman thinks we need to bring back (old) Keynesian economics. But the original Keynesian model was discredited for good reason; it lacked an explanation for the trend rate of nominal GDP growth. Allan Meltzer was one of many who observed that the case for fiscal policy must lie with monetary policy ineffectiveness. But Meltzer pointed out back in 1988 that if monetary policy is ineffective at the zero rate bound, the obvious solution is not fiscal stimulus, it’s a positive trend rate of inflation. I.e., a rate of inflation high enough to eliminate the possibility of a liquidity trap. Paul Krugman knows this, indeed reached the same conclusion a decade later, but for some reason fails to draw the obvious implications from this reasoning:
But zero, it turned out, isn’t low enough to end this recession. And the Fed can’t push rates below zero, since at near-zero rates investors simply hoard cash rather than lending it out. So by late 2008, with interest rates basically at what macroeconomists call the “zero lower bound” even as the recession continued to deepen, conventional monetary policy had lost all traction.
Now what? This is the second time America has been up against the zero lower bound, the previous occasion being the Great Depression. And it was precisely the observation that there’s a lower bound to interest rates that led Keynes to advocate higher government spending: when monetary policy is ineffective and the private sector can’t be persuaded to spend more, the public sector must take its place in supporting the economy. Fiscal stimulus is the Keynesian answer to the kind of depression-type economic situation we’re currently in.
This can only be described as a giant non sequitor. In fact even Krugman doesn’t believe that when conventional monetary policy fails “the public sector must take its place.” He has repeated argued that inflation targeting can be highly effective in a liquidity trap. In other words, when conventional monetary policy has failed, the obvious solution is to adopt a more effective monetary policy regime. I can’t fathom how anyone could disagree.
[The original post had a typo, saying “private sector” rather than “public sector.” HT Bryan Caplan.]
Now I expect I will get Krugman supporters writing in and telling me that he knows all that, that he agrees with me, but that he doesn’t think inflation targeting is currently politically feasible. That explanation might work in an essay about what the Fed should do next week (though I doubt it) but it certainly won’t explain Krugman’s blind spot in an essay devoted to calling for macroeconomics, both theory and policy, to be re-worked from the ground up. In this essay he has literally nothing to say about what went wrong with monetary policy. Indeed at one point he bizarrely states that:
Neither side was prepared to cope with an economy that went off the rails despite the Fed’s best efforts.
This is despite the fact that Krugman has repeatedly criticized the Fed and other central banks for being too conservative, for being unwilling to set an explicit inflation target. It makes no sense, unless one assumes that he is so bent on rescuing old-style Keynesianism that he doesn’t want to publicize the potential of monetary stimulus. He knows full well that fiscal policy is an irrelevant 5th wheel in a world where monetary policy is always capable of boosting expected demand growth to the target level.
Some might argue that I put too much faith in the power of monetary policy to paper over the economy’s flaws. Arnold Kling argued that the crisis reflected a real problem with the economy, the need to reallocate labor and other resources. I do think there was a real problem in late 2007, but this problem was minor compared to the nominal shock experienced in late 2008. More importantly, Krugman clearly agrees with me. He ridicules the “freshwater economists” who don’t see demand shocks, and think everything is a real shock. If the problem of falling NGDP is a real shock that can’t be papered over by printing money, it is equally true that it can’t be papered over by running deficits.
[Although I don’t agree with Kling’s diagnosis, his conclusion that deficit spending doesn’t get at the root of the problem certainly does follow from his analysis.]
So if Krugman really does believe that the problem we face is inadequate demand, he needs to come up with a reason why this problem doesn’t call for a better monetary regime, and instead requires the rehabilitation of an old-fashioned Keynesianism that confused money and credit, that confused saving gluts and liquidity traps, that ignored policy expectations, and that had no model of the trend growth rate of NGDP.
The other big theme of his essay is the failure of the EMH. There are two problems with his analysis. First, since he doesn’t understand how monetary policy failed last year, he doesn’t understand why the financial markets have been so unstable. And whenever an anti-EMH economist sees a dramatic market movement that they can’t understand, they immediately assume that there is no rational explanation. Since I think I understand why the stock and commodity markets crashed last fall, so I am not persuaded by this argument. But even I am at a loss to explain why tech stocks got so high in 2000, or houses in Arizona got so expensive in 2006. So let’s give Krugman the benefit of the doubt. Then what?
The bigger problem, which I already discussed in an earlier essay, is that the anti-EMH crowd really can’t come up with any useful implications for their theory. Or at least any implications that are useful for macro policy. If the anti-EMH crowd wants to argue for regulations specifically targeting some market failure–say banks taking on too much risk, then I won’t strongly object. The obvious regulatory fix for the sub-prime fiasco would be to require all mortgages to be backed by at least a 20% down payments on the property. It wouldn’t eliminate future banking crises, but it might make them less severe. Now try convincing Krugman’s fellow Democrats in Congress to enact such a regulation.
But Krugman has a much bigger agenda:
So here’s what I think economists have to do. First, they have to face up to the inconvenient reality that financial markets fall far short of perfection, that they are subject to extraordinary delusions and the madness of crowds. Second, they have to admit “” and this will be very hard for the people who giggled and whispered over Keynes “” that Keynesian economics remains the best framework we have for making sense of recessions and depressions. Third, they’ll have to do their best to incorporate the realities of finance into macroeconomics.
I seem to recall Stigler once remarking something to the effect that in economics the problem isn’t in coming up with answers; it is coming up with interesting questions. Once you formulate an interesting question, the logic of economics quickly guides you to the obvious implications. I also recall constantly hearing highly technical macroeconomists talking endlessly about “research programs,” attempts to develop complex mathematical models of the economy. We kept hearing that “progress was being made” and they we were “close” to having models that could guide policymakers. One thing that Krugman and I would probably agree on is that much of this “progress” was a myth. We really don’t know much more about business cycles than Friedman did, or Irving Fisher, for that matter.
What’s my point? I completely distrust an economist who talks about a promising new area of research that will soon yield all sorts of insights. In my view when a new area is discovered, most of the really useful insights are almost immediately obvious. The only exception that comes to mind is the huge gap between the development of externalities theory, and the discovery of the Coase Theorem. Perhaps others can find some other exceptions. So one implication of this line of reasoning is that I will never live to see the day when behavioral economics and behavioral finance finally revolutionize economics and finance. I probably don’t know enough about these fields to have an intelligent opinion, but my hunch is that if the standard model that we teach in our textbooks has not yet been revolutionized by behavioral economics, it will never be revolutionized. That behavior economics will always be on the fringes, providing interesting anomalies.
And I think this is the problem with Krugman’s agenda. If we knew how to “incorporate the realities of finance into macroeconomics” we would have done so already. We haven’t done so, because we don’t know how. And I think there are good reasons why we don’t know how. Almost any proposal to do so must somehow, at least implicitly, assume the policymakers are smarter than highly paid investment bankers. And even with all the bizarre bubbles we have seen in recent decades, I think that assumption is just too much for most economists to swallow.
Krugman is a very persuasive writer. I keep telling right-wingers that if we have deflationary monetary policies, policies that reduce NGDP, the free market will be blamed. I don’t see any persuasive rebuttals to Krugman from those on the right. I don’t know if my arguments are persuasive, but at least I’m trying.
PS. Thanks to Dilip for the link.
Tags:
4. September 2009 at 02:54
Scott,
“But Krugman has a much bigger agenda”
Yes, he wants to expand government and now is the perfect opportunity to do it.
“I also recall constantly hearing highly technical macroeconomists talking endlessly about “research programs,” attempts to develop complex mathematical models of the economy. We kept hearing that “progress was being made” and they we were “close” to having models that could guide policymakers.”
I think physicists are closer to coming up with a unified theory of all forces or with cold fusion than what we are of coming close to a complete macro model. But this is not a failure. Imagine if every hydrogen atom had a mind of its own and decided to behave like a helium atom just to piss you off.
Alex.
4. September 2009 at 03:22
Scott, exactly what does he mean when he says markets are far from perfect, or wrong? does he mean risk premia evaporated in the face of excess leverage – or that mortgages were handed out to folks with low incomes who prob couldnt afford it but were a key voter block? or does he mean that treasuries continue to be mispriced, including TIPs – which means none of us could have seen expectations collapse in 2008? its a popular narrative just to blame the market. but what do they mean?
4. September 2009 at 03:49
This is off topic, but is there a reason that inside Google Reader, the text of all your blog entries now say “Purchase Cialis Online Purchase Cialis Online” in one huge run on sentence?
Not happening with any other feeds that I subscribe to…
4. September 2009 at 04:17
Scott,
I’m in awe of your blog and hardly dare comment, so I will restrict myself to pointing out that your RSS feed has been corrupted by spammers – all I get is “Purchase Cialis OnlineCialis From CanadaCialis PrescriptionOrder Cialis Without PrescriptionCialis Buy OnlineCialis Online PharmacyViagra Versus CialisCialis Rug Side EffectsViagra Cialis LevitraBuy Cialis Online CanadaBuy CialisLiquid CialisCialis And ViagraCialis 20 MgTadalafil CialisCialis Super ActiveNon Prescription CialisCialis Free SampleSuper CialisCialis DrugBuy CialisonlineGeneric Cialis CanadaCialis Online PurchaseCialis On-lineOvernight CialisBuy Cialis On LineCialis DoseCialis Online CanadaCheap CialisCialis Without PerscriptionDiscount Generic”
I’m sorry to be using the comments for this, but I couldn’t find any other way to let you know.
Cheers
4. September 2009 at 04:18
“Almost any proposal to do so must somehow, at least implicitly, assume the policymakers are smarter than highly paid investment bankers. ”
Or, as Buchanan pointed out, that they face better incentives or have better motivations.
Which they don’t.
4. September 2009 at 05:20
Alex, Yes, there’s lots more I could have said about the search for the macro Holy Grail, but I’ve already said it in other posts. I think in the end economists will give up, and just let the markets decide how much money is needed to stabilize NGDP expectations.
Van, Those are good questions. When I read recent pieces by Krugman I have the impression that he is going out of his way to look for flaws in the market. I recall the other day he said something about how efficient the Post Office is, and the Department of Motor Vehicles.
Mark and Sebastian, Thanks for pointing that out. I apologize to my readers. I know very little about computers, and don’t understand why this problem keeps coming back. The tech people keep fixing it, but the problem returns.
Tim, I agree. But I am willing to grant the possibility of regulators acting in the public interest, just for the sake of argument. But then I remember that I was once offered a job at the NY Fed, and I try to imagime myself trying to outsmart a bunch of investment bankers. If I was that smart I’d already be retired and living on a tropical island.
But I also agree that the government was cheerleading the sub-prime lending. More “regulation” would have meant more sub-prime loans, because that’s what the regulators wanted, or that’s what the Congressmen pressuring regulators wanted.
4. September 2009 at 05:21
Yes, please look into the RSS thing… something has definitely gone awry.
4. September 2009 at 05:38
oh that”s great thanks
4. September 2009 at 05:59
I think it’s a mistake to read Krugman’s NY Times writing as that of a Nobel winning economist. It makes more sense to read his writing as that of a progressive opinion maker. Thus his support for Keynesianism is a means to an end: a larger and more active government role in the economy. His constant bashing of the EMH can be seen as a progressive’s condemnation of “laissez faire”, Chicago school economics – which he partly blames for the recession.
4. September 2009 at 06:27
Perhaps i am too optimistic, but exactly how persuasive are Prof Krugman’s NYT views? The majority of Americans do not want a massive overhaul to healthcare. Americans view government officials with equal contempt to wallstreet types. And most importantly, Americans by and large dont read the NYT. Essentially, arent his views just singing to the choir…a very small, insulated, self absorbed choir?
4. September 2009 at 06:34
Van, what may be more important is that Krugman seems to be very influential to other economists and economic commentators.
Apart from that, I think you underestimate him, some of his articles are carried far and wide. My colleagues have mentioned him several times in their conversations about the Irish economy in the staff canteen (in Ireland).
4. September 2009 at 06:41
But why would his view carry any more or less weight than Ned Phelps or Gary Becker or Edward Prescott, among economists? His nobel was based on trade, certainly not what he writes about in his column. I can see non-economists spending a lot of time discussing his views, but professional economists?
4. September 2009 at 06:51
I think Krugman has written papers on macroeconomics too, Scott has mentioned them.
4. September 2009 at 08:13
I also have the RSS Cialis problem (not the problem that requires Cialis). Of course, only a problem if you are not getting paid by Eli Lilly.
4. September 2009 at 08:17
‘…I am at a loss to explain why tech stocks got so high in 2000…’
Meaning you didn’t take my suggestion of a few months ago to read Stan Liebowitz’s ‘Re-Thinking the Network Economy, where you will find an explanation.
Speaking of Stan (and his colleague Steve Margolis) he’d agree with this: ‘When I read recent pieces by Krugman I have the impression that he is going out of his way to look for flaws in the market.’
(Stop me if you’ve heard this before) It was a major embarrassment for Krugman when he thought he’d found such in the theory of ‘Path Dependence–promoted by Brian Arthur and Paul Davi–in , iirc, 1994. Krugman ran with it in a chapter of ‘Peddling Prosperity’, even coining the phrase ‘the economics of QWERTY’.
Then Stan and Steve pointed out to him that they’d demolished that theory four years earlier in a piece in The Journal of Law and Economics, titled ‘The Fable of the Keys’. Krugman didn’t respond well.
4. September 2009 at 08:51
I was just discussing the “Economics of QWERTY” on the Austrian Economics blog with Greg Ransom, it’s interesting, see towards the last two pages of this thread…
http://austrianeconomists.typepad.com/weblog/2009/08/hoover-and-the-great-depression-redux.html
4. September 2009 at 10:27
Scott,
Here’s a relevant set of questions:
1) if you believe in the EMH.
2) and if you believe that the gold price is solely a function of the future oversupply of money in circulation.*
3) then, what would a material rise in the gold price imply for the future cost of monetary stimulus?
*Please don’t argue that gold has a “safe haven” value. For every “safe haven” threat that does not involve Fed balance sheet expansion, currency is a better store of value than gold — and that includes war and civil unrest. Gold is ONLY superior to currency when the threat implies that the currency may decline in value.
4. September 2009 at 10:43
If we are not seeing a real, and big, problem with the economy; then perhaps you could include an actual real problem that we can distinguish. Aside from war making, there must be some technology shocks that do cause real problems.
4. September 2009 at 10:56
Scott,
BTW, according to your Output Gap thinking, there should be no reason for gold to rally ahead of a more aggressive Fed monetary stimulus.
And FWIW, I think the rally, if it continues, will tell us that the Fed is about to adopt your line of thinking — charge interest on reserves, formulate an inflation target, etc.
This stands to reason. Even if you believe in fiscal policy, the run rate of stimulus peaked in August, and a decline in stimulus is equal to a contractionary policy.
Further, the Fed’s attempts at “Credit Easing” have reduced spreads, but private-sector credit is still declining, as are incomes.
So time to launch “stage 2”: aggressive monetary policy.
4. September 2009 at 11:00
Bubbles form because almost all of the incentives are there to form them. Executives who earn bonuses and capital appreciation on their stocks and stock options love them. Stock and bond holders love them, for the same reasons (Soros makes money on the way up and the way down, he claims). Consumers love them, because employment increases, sometimes along with wages and credit and their standard of living increases. Producers of goods and services love them, due to higher revenues and profits. Politicians love them, because economically happy voters are more willing to elect incumbents. It is no wonder bubbles occur. The real wonder is why anyone wonders about their causes at all.
Usually, bubbles begin with some innovative financial products or new goods or services that just happen to strike a chord with consumers (unconditioned stimuli, paired with novel conditioned stimuli, in the latter case). The former draws in even the sophisticated investors initially, with the unsophisticated to get in later. The latter involves mostly less sophisticated ones, with some unsophisticated producers ending up with the shock of overhang after realizing sometimes long periods of supernormal revenue and profit growth.
You can add herd behavior and cognitive biases to this simple reality, but it is clear that the above explanation should suffice for anyone with half a brain.
But, there are productive things to be said about cognitive biases. Most people who claim to understand them don’t. Authors of papers in IEEE journals (electrical engineering) know far more about these biases that nearly any psychologists and certainly more than probably any economists. They’ve been building model learning systems and even brains for decades.
But, I am a psychologist, so I will have to couch this short introduction in the nomenclature of my field. There begins with what I term the “myopic bias” which is seen in every organism with a nervous system. It’s the implicit assumption that all that is known about a predictive stimulus and an outcome is all there is to be known. For example, a rat learns to press a lever for food toward an asymptotic rate when controlled for deprivation. Then, fail to load the food cartridge, and the rat attacks the experimental apparatus. Of course, the rat had no idea that researchers were filling the food cartridges and could choose to cease doing so at anytime. Likewise, human beings naturally make assumptions,a nd sometimes attach soda machines.
Now, let’s take a closer look at a long list of cognitive biases as enumerated by psychologists such as Kahneman. Almost his entire list can be explained by a simple model of learned associations, in which multiple memory systems in the brain interact in predictable ways. Yet, Kahneman treats these as separate phenomena. Little wonder few have confidence that these biases will ever matter much to macroeconomics.
So, how to people learn? To sum up briefly, sought after unconditioned stimuli can sometimes be predicted by certain conditioned stimuli. However, the dynamism of the relationships between variables, number of variables, etc. are usually, if not always unknown.
So, in this context I mention the fundamental principles of learning, including generalization, discrimination, symmetry, transitivity, and blocking, among others. To keep this short and simple, forgetting to carry your car keys with you to your car outside a house you’ve been living in for years represents a generalization failure. Seemingly irrelevant changes in that morning compared to most others represents a novel environment over which that previous learning has not taken place. These can be unexpected phone calls or even just novel thinking that particular morning. Most people call these “distractions,” but that term is no more than a label.
Similarly, discrimination failure would be the case of pulling out your own car keys when approaching the car door of someone else. Only the simiarities between this situation and the situation with opening your own car door out-competed claims of differences on your limited sensory and cognitive resources.
These two concepts alone go a long way to describing learning, but there’s also symmetry, which in short is recognizing similarities and transitivity which is recognizing that if a = b, and b = c, then a = c. Symmetry and transitivity require no explicit learning and are a natural by-product of the functioning of neural networks. The latter is the stuff of metaphor and imagination itself.
Now, the learning environment, and the challenges the brain faces, are much mroe complex than many give them credit for. There is a representative finding, for example, that school children tested in the same class in which they learned the material tested perform better on average than children tested in different rooms. So, not only are what some may consider irrelevant predictors playing a role, both in implicit memory systems and explicit ones, but the actual learning environment is also complex.
Consider the explosion of permutations of possible ways to complete certain tasks, especially when there are serial, multi-step procedures with different options at every step. The permutations can easily get into the millions, or much higher. The saying that the definition of insanity is doing the same thing over and over and expecting a different result couldn’t be more wrong. It’s that each trial is actually different in a learning environment far more complex than usually realized. We are fortunate that we make somewhat more better guesses than worse ones.
Now I haven’t mentioned blocking until now, the last property of learning mentioned above. There can be differences between perceived permutation spaces, as I like to call them, and actual ones. Sometimes the actual ones can never be known. When the number of perceived ones is less than the actual number, blocking may be occurring. This involves previously learned associations now stored in the implicit memory system directing attentional and behavioral resources in such a way that possible alternatives are missed. I think it’s easy to see this problem in human thinking, hopefully without resorting to experiments done with rats and apes.
Briefly, just as a note, the explicit memory system, or what we often call consiousness, handles novel stimuli in inverse proportion to the emotional importance of the novelty. The implicit system handles unexpected outcomes, such as those that result in net deficits or surpluses in net intakes of reinforcemnt based on previous expectations. To the degree that we are emotional, the implicit system dominates and older learning and instinct takes over.
There is much more to these ideas, but I’ve gone on much too long already. The point is that humans are far from rational financially and very simple experiments and highly synthesized models can easily demonstrate this. Rationality at the individual level only exists with respect to passing genes, not with respect to financial gains. There are sometimes just happy overlaps.
Now, what does this say about aggregate behavior? Well, we may not know much about that yet. Maybe aggregate behavior can often be more rational than indivdual behavior, but anyone who’s seen a riot at a soccer match or footage of Nazi rallies may have pause.
4. September 2009 at 11:49
Current, I notice that Barkley Rosser is very vague about his claims that L&M haven’t debunked QWERTY. I’ve never seen Paul David respond to them with anything but doubletalk and false promises to one day offer a real rebuttal.
David certainly was scarce (i.e., a no show) in 2001 at the EH.NET when Stan’s UoT colleague Peter Lewin and I debated David’s research assistant at the time he published his AER article, Doug Puffert. The result of that debate was a pretty thorough capitulation by Doug–who, came across as a nice guy, btw–who ended up apologizing to L&M for misunderstanding their argument and promising to correct an article he’d published.
At any rate, when EH.NET let him write their encyclopedia entry on ‘Path Dependence’ in 2008 (which is the most recent volley from the David camp I know of)
http://eh.net/encyclopedia/article/puffert.path.dependence
he had changed his tune considerably. There are still errors, both of commission and omission, but he’s clearly less confident in his argument now than he was before he debated us. For instance:
‘For Liebowitz and Margolis, it was most important to show that the costs of switching to an alternative keyboard would outweigh any benefits, so that there is no market failure in remaining with the QWERTY standard. ***This claim appears to stand.*** David had made no explicit claim for market failure, but Liebowitz and Margolis — as well, indeed, as some supporters of David’s account — took that as the main issue at stake in David’s argument.’
My emphasis in the above.
Nor did Paul David deign to take part in the debate among Hal Varian, David Friedman and myself around Thanksgiving 2002 on usenet’s sci.econ. I’m not aware of David having said anything about the issue in nearly ten years.
In that debate there was an interesting bit about the woman who claims to be the worlds fastest touch typist, typing on a Dvorak keyboard.
Turns out that she wasn’t ‘locked-out’ at all by network effects. She was introduced to the keyboard by a typewriter salesman at the business college she attended. When she discovered she could type really fast on it she bought one and toted it around with her to her various jobs, making an above living as a secretary because of her typing speed.
Which directly contradicts Paul David’s story, as well as a few paragraphs in Puffert’s encyclopedia article I linked to above. You might ask Barkley Rosser if he’s aware of that.
Finally, I apologize to Scott for taking off on this tangent and filling a macro blog with micro matters.
4. September 2009 at 12:30
Van writes:
“The majority of Americans do not want a massive overhaul to healthcare. Americans view government officials with equal contempt to wallstreet types.”
Actually, this is incorrect. Until only last month, a large majority of Americans wanted Obama’s healthcare plans (even though there was no such thing as Obama’s plan because he has so mismanaged Congress). Nonetheless, most Americans (in spite of a massive media blitz) still favor significant overhaul of the system.
http://www.pollingreport.com/health.htm
Also, in terms of trust in institutions, Wall Street (aka Big Business and Banks) ranks well below almost every other institution… This is more of a popularity contest than anything, but you’re claim here is simply incorrect. It reflects a persistent belief in a nonexistent “silent majority” that agrees with your own beliefs/positions. Sorry, the data speaks for itself.
4. September 2009 at 12:31
My apologies, here is the data on trust in institutions:
http://www.pollingreport.com/institut.htm
4. September 2009 at 12:37
Obviously the EMH is not perfectly true, but you don’t need psychological pseudo science to explain why. There is ample evidence that markets are efficient despite biases because the efficiency comes from the players at the margin who know what they are doing.
There was a great letter in the WSJ today from a Hedge fund manager:
One of my friends uses a pair-arbitrage strategy involving securities dual-listed in the US and Canada. He asserts that this strategy is barely cost-effective and earns a scant (and essentially irrelevant) $300/day. It takes many many strategies like that to put millions or billions of dollars to work. Some strategies aren’t worth their expense.
Ultimately that’s the rub: there are inefficiencies in the market but they get corrected when there is a profit to be made fixing them. If it costs too much to close the inefficiency, there it will remain. This applies to certain big inefficiencies (such as executive compensation). Yes, most people will agree its abstractly inefficient, but the cost of making it efficient out ways the benefit.
4. September 2009 at 13:05
Since I’m probably the only non-right-winger here:
Krugman’s views do seem to have shifted more from a sort of pragmatic New-Keynesianism to a populist progressive stance. The latter is inherently distrustful of big business, and represents a long American tradition of anti-big-business sentiment.
Having said that, I am struck by the fact that many anti-Krugmanites have responded to criticism by clinging ever more sharply to the artifacts of their faith rather than accepting the criticism and trying to improve things: the EMH being (I suppose) the ground on which so many have chosen to do battle. The broader points underlying this issue – and I am not defending Krugman’s representation of it so much as the point itself – are:
1) Chicago school economics rely heavily on a hyperational utility maximizing rational agent that is far removed from psychological and social reality. Almost everyone accepts this, but they keep building models using these assumptions. You have to ask, WHY? That’s perilously close to a working definition of insanity.
2) I suspect this has been perpetuated (strangely enough) by a social predilection in academic economics to value mathematical elegance over reality (something Scott has publicly agreed with).
3) Behavioral economics has been the red-headed stepchild of the field, but is finally getting the credit it deserves.
4) We all know about market failures, but in practice, market failures are more the norm than the exception. That’s called living in the real world. It’s just a matter of degree. Many times it’s not worth fixing the failures, since this can cause other problems – but the anti-intervention faction has developed a Pavlovian reflex to defend unregulated markets at all cost. Relying on faith based economics to justify deregulating markets without giving such action any practical consideration is just madness.
5) Capitalism doesn’t exist by itself. Locke’s notion that contracts exist without institutions that enforce them is plainly idealistic. Contracts (and by extension capitalism) exist within an institutional framework (largely provided by government), and the monetary system is part of that institutional framework – by necessity the existence of that framework creates distortions (e.g. bank lending, FDIC insurance, taxation, etc.) Deregulating one half of the market by itself has a word – it’s called a “subsidy”.
6) Mathematical economics loves continuity, but as a I would note that the dominant paradigm in empirical statistical analysis is not continuous – it uses sampling based approximation of distributions. Economic reality is sometimes like physics, and sometimes like computer science – a phenomenon of discrete things interacting.
I’d compare it to using Ordinary Least Squares in a massively non-linear world. OLS may often give you a general answer that is sort of right – but it’s sometimes massively wrong. Yet neoclassical orthodoxy thinks the whole world is Ordinary Least Squares.
OLS is great – subject to a whole host of assumptions (which are almost never true) we can show it’s a best linear unbiased estimate. Beautifully _efficient_.
But the world is not OLS… And pretending that it is OLS yields theoretically efficient outcomes that are in fact perilous.
I would submit that the new standard in economic institution building should not be efficiency, but robustness. Virtually every other field of structural design – architecture, computer programming, statistics, you name it – has accepted the practical importance of robustness as a critical design principle.
Except for mathematical economics.
So ask yourself… why?
4. September 2009 at 13:55
I’m not an Economist or Right-Winger, but I’ve often felt that Krugman claims to be much more liberal than he is. I used to post that on his blog as a comment. I’m not sure why a Sales-Tax Holiday or Dated Coupon, an idea I picked up from Becker and Quinn, wouldn’t be useful against a savings spree. The last post that I read on the VAT rate cut in Britain showed that it’s having a positive effect. Of course, that’s going to be controversial for years.
Also, it’s hard for me to know what any particular person means by Stimulus. Is Social Safety Net Spending part of the Stimulus if it necessitates borrowed money? Meltzer, in his recent WSJ post where he argues that our current crisis isn’t as bad as the Great Depression, says that one reason is that our SSN now isn’t just a soup line.
Here’s a quote from a recent B of E post:
“The idea behind providing such commitments is that they both pull down market interest rates further out along the yield curve and raise expected future inflation. Indeed, in the canonical New Keynesian/New Classical DSGE model, this is the only way of stimulating the economy at the zero interest rate lower bound, as the impact of monetary policy is completely summarised by the current and future path of the policy rate. A variation on this theme, in essence adopted by the European Central Bank, is to provide unlimited financing to the banking system at the policy rate at longer maturities than usual.”
Now, I thought that a Stimulus was supposed to help raise expectations of future inflation, and also help with unemployment. Indeed, that’s what I thought that the Chicago Plan meant by Fiscal Measures.
So, although I find the Monetary Response the main response, I’m still not certain why a Stimulus can’t help. Indeed, looking at this from one point of view, it seems to me that some people are more worried about a Ratcheting Effect than our current crisis. If I thought that the RE was a real problem, then I’d probably be more worried about our reactions to this crisis. But I see the size of our govt largely due to the competition of interests, and far less to do with ideology. That’s why we could have a banking policy of deregulation combined with implicit govt guarantees hardening into explicit govt guarantees over the last thirty years. That’s the Ratcheting Effect we should have been worrying about. Just the view of someone who sees himself as neither left-wing nor right-wing.
4. September 2009 at 14:29
Statsguy:
The EMH has little to do with Scott’s argument against Krugman here. Second, the EMH is a weak form of ratex. Its not a question of ratex versus the rest of Keynesian thought. And what is Keynesian thought anyways? Some gems have endured, but most Keynesian theories were eviscerated by the experience of the 70s. Krugman’s goal in rehabilitating the word Keynesian is largely political.
Third, its useful to know when a claim violates a ratex assumption or violates the EMH. Even if it is not your intention to apply those ideas. The reason for this is the problem of special pleading. i.e., you need to consider the consequences of abandoning the EMH in this instance.
Fourth, Scott advocates a very active Fed. It would be hard to claim that the he insists on unregulated markets.
Fifth, Krugman is being inconsistent in his public essays and his academic work. He isn’t making criticisms that should be listened to; he is purely trying to advance a political agenda under the guise of intellectualism.
Sixth, the GT consists of two things 1) a very clever, convoluted argument that unemployment is due to sticky-high wages–i.e., precisely the classical claim. Keynes argued (very obscurely) that the CB must induce an inflation to drive down real-wages. 2) He dresses the matter up by attacking several errors in reasoning present in the Marshall school in which he was raised–errors that in his own ignorance of the greater body of economic thought wrongly attributed to a ‘classical’ economics.
4. September 2009 at 14:31
‘We all know about market failures, but in practice, market failures are more the norm than the exception.’
I find that to be a stunning claim. Every day–indeed, every minute of every day–millions of people experience market success. I just walked to the grocery store and found something I wanted for dinner tonight, at prices I have no problem with paying. It happens every night.
I also find that most claims of market failure turn out to be the claimant’s unfamiliarity with the subtleties of a markets functioning. That is, not failure, but success.
4. September 2009 at 15:05
Mark, Yes, I contacted tech support—but it is a weekend. These problems always seem to crop up right after I attack Krugman . . .
(Just kidding.)
MBP, You may be right. I generally try not to attack motives. I don’t doubt that Krugman believes in the proposals that he puts forth, it’s just that they don’t seem consistent with his theoretical framework. (I.e. he seems too smart to really believe in the old Keynesian model.)
Van, Those are good arguments. However while only a small proportion of Americans read the NYT, a large proportion of Americans who write new legislation and who teach in elite institutions read the NYT. So I think he does have some influence.
Current, I agree. My non-economist colleagues always ask me about him. They thought it as a big deal that Krugman mentioned my name (only to dismiss me) but couldn’t care less that Cowen and Caplan and Mankiw have said lots of good things about my blog.
Van#2, If you want to see how influential Krugman is among other liberal economists, read Delong–he loves Krugman’s blog.
Thanks Gene.
More responses later . . .
4. September 2009 at 15:39
Patrick, I’ve been so busy recently that I’ve read virtually nothing. Tyler Cowen’s book is the only exception. Perhaps you could summarize the argument.
By the way, the fact that I can’t think of an explanation is meaningless, as I’ve argued. The market had its reasons, and no one has yet come up with any interesting implications for the anti-EMH position, so I was just throwing that assumption out for the sake of argument, only to show it didn’t lead anywhere interesting.
Yes, I do remember your QWERTY example. Like many of us, Krugman is better at criticizing others than accepting criticism.
Current, I can’t get Greg’s blog here in China. Any chance you could block and copy the passage over here?
David, Yes, I believe the EMH is a useful theory (albeit not literally true.) I don’t think the rise in gold prices is an indication that the market expects inflation, I think TIPS spreads are a far superior indicator of inflation fears. There is a big and growing demand for gold jewelry in India and China, which are developing fast. Both countries have cultures that traditionally covet gold.
Mattyoung. The fall in the real estate market from mid-2006 to mid-2008 was a real problem. The rise in oil prices in early 2008 was a real problem. Hurricane Katrina was a real problem. The sharp rise in the minimum wage was a real problem. Higher MTRs are a real problem. If we do something about global warming that might well be a real problem. The 1971-74 price controls were a real problem.
Technology generally doesn’t go backward in any important way, so “technology shocks” rarely cause recessions. But real shocks can get “entangled” with monetary policy, and cause recessions. I think that is one reason why recessions often follow oil shocks. The oil shocks cause money to be too tight, as the central bank doesn’t respond in the appropriate way.
David#2, You said;
“BTW, according to your Output Gap thinking, there should be no reason for gold to rally ahead of a more aggressive Fed monetary stimulus.”
Just to be clear, I oppose policy rules (such as the Taylor Rule) that utilize output gaps. I favor NGDP targeting. NGDP is a nominal variable, whereas output gaps are real variables.
You said;
“And FWIW, I think the rally, if it continues, will tell us that the Fed is about to adopt your line of thinking “” charge interest on reserves, formulate an inflation target, etc.”
I hope you are right about Fed policy, but I doubt it. Otherwise the TIPS spreads would show inflation fears.
4. September 2009 at 16:01
Mike, You said;
“You can add herd behavior and cognitive biases to this simple reality, but it is clear that the above explanation should suffice for anyone with half a brain.”
I must have only 1/4 a brain because I don’t understand your explanation at all. If you are right then why don’t Wall Street firms start up “Kahneman mutual funds” and outperform the market by taking advantage of investor irrationality?
Patrick, No need to apologize, I find this stuff very interesting. And your summary was helpful, as it is difficult for me to follow up on very long links people attach when there are lots of comments, although I try to at least look at all the links I can open. But I do appreciate when people summarize the arguments in any link, as you have done nicely here.
Stasguy, I don’t disagree with the specifics you mention, but the data on polling never speaks for itself. Indeed I think it is grossly misleading to even speak of “public opinion” as if it is some sort of objective reality. It turns out that simply asking a question will reshape public opinion. I recall a great example from a few years back, it went as follows.
1. People were asked if we should be spending more of Medicare. Many or most said “yes.” (I don’t recall the exact percentage.)
2. People were asked what the appropriate rate of increase in Medicare spending was . They were given several options; 2% , 4%, 6%, 8%, 10%. Only two percent of the population thought Medicare spending should rise as fast as IT WAS ACTUALLY RISING (which was quite rapid during the period when this poll occurred.) 98% favored a smaller rise. Thus whether people favored more Medicare spending depended completely on how the question was asked.
After that I lost all confidence in polling on complex policy issues. I still think polls have some validity in horse races like elcetions, or simple yes/no questions such as whether person X should be confirmed by the Supreme Court, but not complex issues. I support massive overhaul of the medical system, but I don’t view myself as a supporter of Obama’s proposal–although I might have supported Democratic Senator Wydan’s proposal.
More later . . .
4. September 2009 at 16:04
Everyone,
I’ve never read so much fail at one sitting. I’m sure Scott will reply to all of this silliness. Just to drive the point home so will I.
Scott,
I had this discussion on “The Austrian Economists” blog, not on Greg’s own blog. This is the gist of it. Greg pointed out that the “path-dependency” example of the QWERTY keyboard is, broadly speaking an myth. He pointed it out as an example of an illustrative myth, something told within a tradition to make a specific set of points. A story that isn’t necessarily true, but is useful. Those who tell it don’t necessarily know that it isn’t true, it’s something evolved rather than built by a human mind. (This is a fascinating idea, I’ve come across these stories many times in engineering without having the words to describe them. Greg is extending Hayek in an interesting way.)
Several posters on that blog mis-understood Greg to be talking about the QWERTY problem itself. I pointed out what he was saying. Also, I pointed out that the QWERTY myth is much more difficult to resolve than most people think. The naive form of it assumes that it takes the same time to learn one type of keyboard as it does to learn another. This isn’t true. The worlds fastest typist with a normal keyboard uses a dvorak keyboard, she can do ~150 words per minute. However, a legal stenographer can do ~300 words per minute, so can someone who writes teletext for UK news (I used to know someone who did this). Those folks don’t use normal keyboards, they use “chord” keyboards. These have five buttons each under a finger of one hand. The typist presses several keys at once, like a pianist. These combinations represent words or letters in a complex coding system.
The Dvorak vs QWERTY controversy is very complicated. I’ve known some people who use keyboards all day in their jobs who have changed to Dvorak because the keyboard shape reduces RSI, even though it slows their typing.
4. September 2009 at 16:14
Jon, I think that’s a very good summary of how and why markets are approximately efficient, but not perfectly efficient.
Stasguy, You said,
“1) Chicago school economics rely heavily on a hyperational utility maximizing rational agent that is far removed from psychological and social reality. Almost everyone accepts this, but they keep building models using these assumptions. You have to ask, WHY? That’s perilously close to a working definition of insanity.”
I use these models because they are very useful. My problem with the Chicago School is that they put too little weight on wage and price stickiness, and hence demand shocks.
You said;
“4) We all know about market failures, but in practice, market failures are more the norm than the exception. That’s called living in the real world. It’s just a matter of degree. Many times it’s not worth fixing the failures, since this can cause other problems – but the anti-intervention faction has developed a Pavlovian reflex to defend unregulated markets at all cost. Relying on faith based economics to justify deregulating markets without giving such action any practical consideration is just madness.”
I agree. Indeed I also agree with most of your other comments about deregulating half a market, and the importance of robustness. Does that mean I’m not a “right-winger?”
4. September 2009 at 16:20
Dr. Sumner,
You wrote:
“Mike, You said;
“You can add herd behavior and cognitive biases to this simple reality, but it is clear that the above explanation should suffice for anyone with half a brain.”
I must have only 1/4 a brain because I don’t understand your explanation at all. If you are right then why don’t Wall Street firms start up “Kahneman mutual funds” and outperform the market by taking advantage of investor irrationality?”
First, the half-a-brain statement referred to the four paragraph, sans the cognitive biases. The point is, that perhaps even from the perspective of rational expectations, the development of bubbles should be expected.
Second, I mention that Kahneman’s ideas are woefully incomplete and fail to recognize that even a simple synthesis of well-known learning theory can explain all the observations he treats as individual phenomena. Even this Nobel winner lacks a real synthesis in his overall perspective on brain and behavior.
I want to add a separate point here that there is a real problem with the idea that investment strategies are easily developed or even mimicked. Given what I wrote above about permutation spaces and other aspects of the nature of learning, such phenomena are more complex than they appear on the surface, which is just one of many reasons I’m highly suspicious of strong EMH theories.
4. September 2009 at 16:21
I meant to say that the half-a-brain statement refers only to the first paragraph in my original comments on this post.
4. September 2009 at 17:09
Statsguy, I don’t know a lot about mathematical economics, but I do know that there are highly technical macroeconomists who put great weight on robustness, such as Bennett McCallum. So it may not be quite as bad as you assume.
Don, Krugman favors the spending side of stimulus much more than the tax cut side. Of course tax cuts are now associated with (right-wing) Reaganomics. In my view even better than a sales tax holiday would be a payroll tax holiday. Singapore uses that sytem, and someone commented here that France does as well.
Automatic stabilizers are usually treated separately from discretionary stimulus, although you are correct that they have the same effect. The reason I oppose fiscal stimulus is that it increases the budget deficit. Yes, it could “help” as you say, but if monetary policy is targeting the forecast, as it clearly should, then it doesn’t need any help, because the expected growth in AD is exactly on target.
Jon, I agree with your comment. Let me add that without the liquidity trap, the GT would amount to little or nothing. The “classical economists” already knew about sticky wages and prices, and hence knew that anything that reduced nominal spending could have real effects. For instance they knew that if an increase in saving led to lower velocity, it would raise unemployment. Without the liquidity trap (which Keynes misunderstood, as even Krugman acknowledged) the GT wouldn’t amount to a hill of beans.
Patrick, I think you and statsguy may be debating semantics. If “market failure” is less than 100% perfect competition, then it is common–and I took that to be Statsguy’s point. If statsguy is claiming that most markets exhibit failure that calls for government regulation, then I agree with you 100%. Most market “failure” would be made far worse if the government tried to “fix” the problem.
Current, Just to be clear, when I mistakenly said I couldn’t get Greg’s blog over here, I meant I couldn’t open the link you attached (which I wrongly assumed was his blog.)
I don’t know enough about the issue you raise, but aren’t legal stenographers typing out shortcuts, and thus am I right that their system would not be appropriate for general users?
Patrick seemed to indicate that the other side of the debate had admitted that switching over was not worth the cost, so there was no market failure. Is that right? That would seem to be the key issue from my perspective.
Mike; You said;
“Second, I mention that Kahneman’s ideas are woefully incomplete and fail to recognize that even a simple synthesis of well-known learning theory can explain all the observations he treats as individual phenomena. Even this Nobel winner lacks a real synthesis in his overall perspective on brain and behavior.”
It is very likely that I have misundersttod your argument, but here is my reaction anyway:
I have always thought bubbles were defined as market prices that are “obviously wrong” ex ante, not just wrong ex post (which of course is always true.) I’m having trouble understanding how you think of bubbles. You seem to suggest on the one hand that a really “simple synthesis” can explain these phenomena, and that it is nevertheless too hard for a Nobel Prize winner like Kahneman to understand. Where am I misunderstanding your view?
Just to be clear, I don’t want to get bogged down in a debate about behavioral econ, which I know little about. I’d like to focus the discussion on the relevance of these ideas for bubbles. Can we predict them? Can public policymakers effective regulate them? These are the only issues relevant for this blog. I have no ability to debate cutting edge theory of learning.
4. September 2009 at 18:56
Well, how about we just stick to my first paragraph for the time being then, leaving my very sparse treatment of learning theory out for the time being.
I wrote:
“Bubbles form because almost all of the incentives are there to form them. Executives who earn bonuses and capital appreciation on their stocks and stock options love them. Stock and bond holders love them, for the same reasons (Soros makes money on the way up and the way down, he claims). Consumers love them, because employment increases, sometimes along with wages and credit and their standard of living increases. Producers of goods and services love them, due to higher revenues and profits. Politicians love them, because economically happy voters are more willing to elect incumbents. It is no wonder bubbles occur. The real wonder is why anyone wonders about their causes at all.”
With respect to Kahneman, many, many people understand the workings of the brain better than he does, but perhaps none of them, or very few of them work in the fields of finance, economics, or even psychology. Most of them might be in AI engineeering research.
4. September 2009 at 19:48
Mark –
I’ve noticed that, too. But, I assumed this was the efficient outcome.
4. September 2009 at 19:53
Patrick:
“I just walked to the grocery store and found something I wanted for dinner tonight, at prices I have no problem with paying. It happens every night.”
Scott already answered this – in many/most circumstances, “fixing” the market failure causes more harm than good. But, to follow your specific example (since my line of work occasionally intersects consumer package goods), the “free market” you happen to be referring to is _already_ heavily regulated.
You know the claims on that package of food you bought? That went through legal review by the manufacturer to comply with FDA truth in advertising claims. Indeed, the plant where it was manufactured was inspected by FDA agents, and probably local building code agents too. If you bought fresh food, it was batch tested for bacteria count. The nutritional panel on the box (which, if you’re a parent like me, you read religiously)? Mandated and standardized by the FDA and Congress.
But let’s go further… The interstate highways on which the food was transported? Built by federal money. The sidewalks you used? Built by local money. The cash register? Tapped an electronic network that was built using federally supported technology (developed vis a vis DARPA 40 years ago). The electricity in the store? You got it. The employees who sold you the stuff, and shelved it, and so forth? Yep. The port where the food or ingredients may have been imported?
All of this goes well beyond the bare minimum of basic security and property rights (which also depend on government). The “free” grocery markets that seem to work so magically and seemlessly every day are in fact supported by nearly invisible systems and institutions that are mostly noticed when they break down, or there is some health scare.
Likewise, the “non-market” instutitions we rely on are supported by markets. All of our politicians? Elected in public campaigns that relied on media market buys, and run by staffers who participate in a labor market. Those federally funded highways? Most of them built by private contractors. And the clipboards that FDA agent uses in his/her inspection? Yep.
Markets happen to have some beautiful properties, but as the NRA loves to remind us (rightly)… Freedom is Not Free, though I’m not sure they completely understand the meaning of that phrase. Well, Free Markets Are Not Free either.
(And, separately, Scott is 100% right about general concerns on polling on complex issues, particularly question wording effects.)
All of the above doesn’t defend Krugman here… That his writing has shown a marked shift from deficits-as-enablers-of-monetary-policy to deficits-good-in-themselves seems true, and troubling. (In other words, he seems to think that Debt is a better way out than Inflation, rather than debt spending is a way to stuff newly printed money into the system…) Krugman seems to spend many more posts arguing the benefits of pure deficit stimulus spending rather than the need for non-standard monetary expansion (which, when he does post about, he tends to support – just not as enthusiastically). For example, he’s pushed stimulus hard – very hard – but reacted with lukewarm support to “QE”:
http://krugman.blogs.nytimes.com/2009/03/20/fiscal-aspects-of-quantitative-easing-wonkish/
Why do we not have a post titled “Monetary aspects of stimulus spending”?
It’s not that Krugman would disagree with ssumner’s monetary proposals, but rather where Krugman puts his emphasis.
But Krugman has been right that we’ve had a much needed kick in the pants in the economics field. Suddenly, real-world economics is in vogue. On this point, I think SSumner and Krugman are actually aligned.
5. September 2009 at 02:08
Yes, chord keyboards use shortcuts. But, here is where one of the problems with this area comes up. Wouldn’t it perhaps be best if we all learnt those shortcuts and stopped using conventional keyboards? Because it takes more time to learn chord keyboards that is a very complicated question.
5. September 2009 at 07:36
Current, chord keyboards and court stenography are completely irrelevant to the QWERTY debate. Especially your remarks about devices that rely on electronics which were not available to 19th century typewriter manufacturers.
The only issue in the QWERTY debate is, did Paul David and Brian Arthur identify a market failure. As Scott correctly noted even David’s staunchest defender concedes they didn’t.
5. September 2009 at 07:52
Well, yes Stats Guy, there are goods that are non-rival and non-excludeable in consumption, for which one can make a case for government provision–though not always; radio and TV broadcasts fit the description, but entrepreneurs found ways to provide them. And, who provided the Oregon Trail? Not the DOT.
But, those are a small fraction of the gazillions of goods and services that change hands.
As for things like: ‘The employees who sold you the stuff, and shelved it, and so forth? Yep. The port where the food or ingredients may have been imported?’ That’s just baffling. You surely don’t think the store employees were government provided. Nor that ports–many of which preceded any formal government–were the creation of government.
5. September 2009 at 07:58
Scott, this makes you a real outlier as far as I can tell.
And let me suggest this — the fact that your macro model, and Krigman’s, has NO PLACE for even imagining real “problems” in the structure of production taking more or less time is exactly the reason you can’t imagine the unimaginably large
size of these effects. The poverty of you macro determines the poverty of your empirical understanding and your causal picture.
This is Klings case against Krugman, this is Hayek’s case against Keynes and the macroeconomic profession (see his Nobel lecture), and this is Garrison’s case against Friedman, Lucas, the New Keynesians, the Real cycle theorists, etc., etc.
Scott wrote,
I do think there was a real problem in late 2007, but this problem was minor compared to the nominal shock experienced in late 2008.
5. September 2009 at 08:07
For amusement purposes only,
http://www.youtube.com/watch?v=NndiiezGkNY
But her fingers don’t seem to be moving particularly fast.
5. September 2009 at 08:29
Regarding Statsguy’s criticisms of “Mathematics economics”….
I agree with Statsguy about mathematical models. I think most are hopelessly unrealistic and unhelpful. I disagree entirely though with his criticisms of capitalism. It is fashionable for modern leftists to say that Capitalism requires the justifications provided by “Chicago Economists” and their “mathematical models”. They imply that if the arguments of Chicago Economists are wrong then Socialism and Interventionism would clearly be the best courses. Have they not heard of the Austrian School? or the New Institutionalists? For that matter, have they not heard of the Classical economists or the first marginalists? Have they not heard of the Socialist calculation debate? Of course they have heard of all this, they ignore these arguments for rhetorical purposes.
The criticism of “unrealistic mathematical models” is especially funny coming from the mouths of New Keynesians and Old Keynesians. Models are just as important to these schools as they are to Chicago economists. All of these schools use aggregate capital, aggregate consumer goods output and aggregate investment, each of these are massive simplifications. Old Keynesian theory relies on barrowloads of outrageous assumptions, such as the liquidity preference theory of interest and the marginal-propensity-to-consume model of consumer behaviour. New Keynsian economics is not much better. Keynesians economists who deride unrealistic mathematical models are saying: Do as I say, don’t do as I do.
As Lord Harris said, criticising markets for not being perfect entails believing in perfect government. “Market failure” is an analytical concept, it is often defined as a situation where transactions occur that are not “Pareto optimal”. Described in this way all markets “fail”, and as Statsguy says “market failure is the norm.” All this demonstrates is the dubiousness of the concept. All transactions in a market economy involve “sticky prices” and asymmetric information. The problems occur in Socialist and Interventionist economies too, they reflect fundamental limitations of human knowledge. The most important question is: Can government intervention do better than the market? As Mises and Hayek showed in the Socialist calculation debate the answer is clearly no.
The questions around “Behavioural Economics” are similar. Certainly individuals base their economic decisions on general rules, this is not news, the first marginalists wrote about it. Certainly these general rules will have specific flaws in specific circumstances, which is what Tversky, Thaler and Kahneman mean by biases. This is not an argument for socialism of interventionism. Behavioural economists study the implications of behavioural biases in market situations. This doesn’t mean that such biases do not occur in other decisions or other forms of decision making. Congressmen, Presidents and state bureaucrats are not blessed with immunity from these forms of irrationality because of their positions. The Behavioural Economists have simply decided to discuss the biases in the context of market transactions because it suites their left-wing political agenda. They are hoping to apply the biases associated with the framing effect to their readers and make them artificially associate markets and biases. (This is a variant of an old ploy of Gramsci, though they may not know that).
The important question here is “Should we support governments that legislate to counteract these biases?” This requires firstly that it be possible to know when and where they occur and how to counteract them. In real life situations single biases do not arise in isolation, real market transactions involve several biases, hence it is not clear how to counteract them. Even if it were the political questions remain. If governments are given powers to counteract biases then why should they use such powers to that end? Why shouldn’t politicians and bureaucrats use these powers to enrich themselves and further increase their poltical power? If these biases are really unconcious then the voter will not be aware if they are being successfully counteracted or not. So, the voter is not able to stop abuse of these powers from the ballot box because he can’t distinguish it from their use. Even if politicians and bureaucrats are benevolent and paternalistic the problem of ignorance remains. How can the electorate distinguish well meaning but hapless use of these powers, which may be destructive, from competent use of them?
Furthermore, restrictions put in place by regulatory bodies to deal with biases further undercut the general nature of the law. The law must be general in order to allow many sorts of evolution within society. As Hayek wrote “Humiliating to human pride as it may be, we must recognize that the advance and even the preservation of civilisation are dependent upon a maximum of opportunity for accidents to happen. These accidents occur in the combination of knowledge and attitudes, skills and habits, acquired by individual men and also when qualified men are confronted with particular circumstances which they are not equipped to deal with.” (“The Constitution of Liberty” p.27). A regulation may counteract a bias for a time. But, the situation may change in the future to one where it is beneficial for individuals to act differently. That can only be discovered if acting differently is permitted, of course some schemes for “liberatarian paternalism” permit this, but not all.
5. September 2009 at 08:54
Scott,
What is your take on what the markets are currently telling us? Does monetary policy seem much better now?
5. September 2009 at 10:30
“The bigger problem, which I already discussed in an earlier essay, is that the anti-EMH crowd really can’t come up with any useful implications for their theory. Or at least any implications that are useful for macro policy. ”
Anti-EMH macro implications are very clear – volatility of macro variables will be higher than estimated by EMH theories.
5. September 2009 at 11:26
Regarding keyboards and path-dependency….
I haven’t been particularly clear about this. I agree with Patrick R. Sullivan’s response mostly.
The problem though is the way the issues are put. Now, Patrick, you may know about this sort of thing, but I think it’s worth talking about it for the benefit of everyone else.
What some critics of the market claims is that there was a sort of “market failure” when the QWERTY keyboard was designed. The keyboard designers of the time who worked for the successful typewriters companies chose an poor layout. Supposedly network effects have meant that this layout has persisted. Now, it is doubtful that this is true. Apart from this though it is doubtful if the questions associated with it are good ones. As an engineer I have made these sorts of decisions before, I’ve made decisions that have resulted in network effects within the company I work. In my view I have made decisions that are poor in the sense it was said QWERTY is poor, they have resulted in less than optimal situations that have persisted through time. But, this itself isn’t necessarily a criticism of the decision making process within the company I work, or that of the market more widely. The question that must be asked is – what is the alternative? The alternative is to examine each decision that could have network effects much more carefully. But doing this has huge costs in time and resources since it is difficult to know beforehand what decisions will become important through network effects. Every decision that could possibly fall into that category is huge. Indeed, if we could say what network effects will occur than rational humans would have little use for the networks themselves, they would just go to whomever knows what will happen and learn from them. This is related to the point Hayek makes which I quote above.
5. September 2009 at 11:30
Incidentally, although modern chord keyboards are electronic mechanical versions of them did exist in the 19th century, though they were rather different.
My point about this was to show that the decisions to use one keyboard or another are more complicated than just the WPM that on provides. There is the length of time to learn and some other costs as well, such as difference in RSI problems.
5. September 2009 at 16:21
Current:
“As Lord Harris said, criticising markets for not being perfect entails believing in perfect government.”
Hardly… One can criticize both markets and the government at the same time. Living in the real world means living in a world where perfection does not exist. It means choosing the lesser of evils on a case by case, empirical basis. As noted, many/most market failures are not worth fixing. But some are.
Alternatively, we could scrap 350 years of American pragmatism for ideological purism.
123 writes:
“Anti-EMH macro implications are very clear – volatility of macro variables will be higher than estimated by EMH theories.”
Precisely… Which means robustness becomes a worthwhile objective of system design. Not merely local efficiency.
5. September 2009 at 17:07
I agree that we can criticize both, and that in the real world perfection doesn’t exist. However, this doesn’t justify a “case by case, empirical basis”. The important question is which methods works best overall, as I said, that question is settled, and the answer most definitely isn’t socialism.
No such “case by case basis” can be constructed. Because in an interventionist economy the two are interwoven. The capitalist takes his queue from the prices provided by others, some of which may be state companies. No “market failure” of a particular market can be clearly distinguished. See what I wrote two posts above about the QWERTY keyboard.
As I said above the idea of “market failure” is misconcieved. That doesn’t mean that no criticisms of capitalism can be found. Rather those embodied by the normal idea of “market failure” or the lack of “perfect markets” are not correct. For example, consider a situation where you are doing work that involves nature, farming for example. In such a situation you trade your time and effort for results in the form of crops. In such a situation can you trade these things in a perfect manner for the outputs? No, since the exact relationship between input and output isn’t known you can’t. Markets are no more imperfect than this form of autarky, or most froms of autarky. Information is always imperfect and prices always sticky, even without a market involving others.
5. September 2009 at 18:21
“The important question is which methods works best overall, as I said, that question is settled, and the answer most definitely isn’t socialism.”
How do we get from regulated markets to socialism? That is the type of polemic leap that Rush Limbaugh would make.
The determination of whether to use the apparatus of the state to intervene in a market (or, alternatively, to create a market that otherwise would not exist) is entirely empirical… It is always a cost benefit analysis. Some orthodox neoclassic folks simply have really really high priors that the costs are always worse than the benefits.
Oddly, such folks seem to recant their moral high ground when THEY are the ones in need of state intervention. Greenspan and Hank Paulson come to mind.
So let me ask you this: WHY is it that there seems to be a strong correlation between neoclassical beliefs and predilection toward arcane mathematical “research” that (we, and now many others, agree) is often worthless? Is there a causal reason, or is it just chance?
5. September 2009 at 19:27
Scott:
The report of the Jackson Holes conference in The Economist (Economic focus: 29 Aug) mentions a paper by Carl Walsh which apparently proposes targeting of the price level rather than the inflation rate. Walsh’s paper prompted discussion of the consistency of current monetary policy objectives. The report states:
“Not surprisingly, Fed officials denied there was an inconsistency between keeping inflation stable and rates low for a long time. The point, they argued, was to stop inflation expectations falling, not to push them up.”
Does this mean that the Fed is now targeting inflation expectations?
5. September 2009 at 20:01
So, what would the
5. September 2009 at 21:16
Is advocating monetary intervention from the government consistent with the EMH?
More generally, how would someone who believes in the EMH think about bubbles –
A. Bubbles cannot form without govt intervention
B. Bubbles form naturally but the information required to predict & prevent them is impossible to obtain. Hence ‘markets might be inefficient but let’s have post-intervention via monetary means not pre-intervention via regulation’
or some completely different option?
Earlier in the thread, someone pointed out that irrational behaviour can be corrected by other market participants who take advantage of it. But what happens when following the herd is the rational thing to do? (selling shares of a sound company in a bear market, firms being pressured to buy into the subprime bubble due to profits competition makes). The ideal strategy would seem to be follow the bubble in the short term(with possible heavy losses in not doing so) and sell out at the very earliest sign that it is going to fade. There are people who take advantage of the bubble and the market does eventually return to the non-hype price, but with a lot of suffering in between.
Sometimes, portions of economics seem like thermodynamics, which only talks about equilibrium states and says little about the transitions between these equilibrium states which requires a more complicated theory like statistical mechanics where one cant merely measure simple macro things like temperature, volume but actually go into the details of the gas, figure out that it is made of atoms(a theory which didn’t have a good reputation as late as 1900), look at how atoms interact and then one can form models of inter-equilibrium states. In economics, this kind of theory might be vastly more difficult because the interactions are more complex and one would have to settle with crude approximations.
Anyway, sorry for rambling on, I am clueless about all this stuff and would be glad to see any responses/corrections.
6. September 2009 at 12:02
I’m going to give a brief answer to this, but you really need to study the past several hundred years of thought on the subject of banking.
Banker have long claimed that they should be permitted to mediate short-run fluctuations in the demand for money by issuing unlimited notes. History is replete with stories of government trying to get that policy right. Under the current regime, the Fed manages a monopoly of base money. Errors in the handling of the monopoly cause unintentional fluctuations which are then corrected.
Bubbles have very little to do with it.
6. September 2009 at 12:16
Earlier Scott asked me to summarize Stan Liebowit’s argument in his book ‘Re-Thinking the Network Economy’ which partly explains the ridiculous heights to which internet stocks rose in 2000.
In short, it was something on the order of ‘the madness of crowds’ of stock analysts, and even some prominent economists such as Carl Shapiro and Hal Varian in their ‘Information Rules’, who were promoting ‘first mover advantage’ in markets with ‘network effects’, in which it was important to give away your products to have an ‘installed base’. Unfortunately, without a very good grasp of what they were talking about.
Stan has a few chapters of his book up on his website
http://www.utdallas.edu/~liebowit/
In the concluding chapter, Stan talks about a special section of the WSJ’s January 3, 2000 paper, in which there was an article titled, “So Long Supply and Demand.” Stan writes:
—————-quote—————–
The supposed death of supply and demand was born of monumental hubris championed most fervently by those largely unschooled in economic analysis. For example, the author of that article provided several illustrations of how this belief had entered managerial thinking. He quoted Mark McElroy, a principal in International Business Machines Corp.’s Global Knowledge Management Practice as saying: “Conventional economics is dead. Deal with it!” He also quoted Danny Hillis, vice president of research and development at Walt Disney Co. as saying that the new economy “is actually much broader than technology alone. It is a new way of thinking.”
Thomas Petzinger, the author of that article [wrote]….
‘You can understand why economists throw cold water on the new-economy concept, since accepting it would require them to abandon many of their dearest tools and techniques. It has become cliched to cite the historian Thomas Kuhn’s 40-year-old concept of a “paradigm shift” — a revolution in knowledge that forces scientists to give up the beliefs on which they have staked their careers. But that’s exactly what economics and accountants could be facing.’
[snip some stuff, and back to Petzinger:]
‘On an economywide level, these accelerating improvements may now be entering a supercritical phase in which they compound exponentially. Inventories, which once triggered or prolonged recessions, are not just declining but in many places evaporating…”Economists fail to realize that these improvements are reducing costs so radically as to enable entirely new ways of doing business,” says telecom consultant David Isenberg of isen.com, Westfield, N.J…. Creativity is overtaking capital as the principal elixir of growth. And creativity, although precious, shares few of the constraints that limit the range and availability of capital and physical goods. “In a knowledge-based economy, there are no constraints to growth,” says Michael Mauboussin, CS First Boston’s managing director of equity research. “Man alive! That’s not something new?”
[snip Stan’s commentary on the above, and again return to Petzinger:]
‘In his classic undergraduate text “Economics,” Paul Samuelson noted that any second grader could figure out that increased supplies cause lower value. But that was before Windows 95, automatic teller machines and Nike shoes. Products used in networks — whether computing, financial or social — increase in unit value as the supply increases…Former Stanford economist W. Brian Arthur has popularized this more-begets-more concept under the banner of “increasing returns.” The timeless notion of diminishing returns isn’t dead, of course, but it applies to an ever-shrinking proportion of value-added activity, such as grain harvests and polyvinyl-chloride production… This explains why a seemingly insane strategy such as giving away your basic product has become a strategy of choice in the new economy…[T]he vendor collects revenue from another source, such as from selling upgrades, support or advertising. (Radio and television broadcasters — networks, after all — have always operated this way.) Another network, the cell-phone system, exploded when telecom companies began providing phones for practically nothing, even free of charge, and reaping increasing returns from air-time charges…
[snip more commentary]
‘Our 500-year-old system of accounting has grave limitations in this world.. But for now, according to the CS First Boston atoms-to-bits report, “there is a substantial and growing chasm between our accounting system and economic reality”…[I]n an economy awash in capital, the endgame, not the score at the end of each quarter, is all that counts…”Earnings are a decision variable, not a requirement,” says Prof. Arthur, the economist. “If everyone thinks you’re doing fine without earnings, why have them?”‘
——————-endquote——————-
That’s more than enough example of the crackpot theorizing and elementary misconception that was prevalent during the dotcom boom and bubble. But, it was accepted as valid by many investors.
6. September 2009 at 14:03
I get your point that the goverment is already heavily intervening in money supply. I am confused right now about what i really wanted to ask, so please ignore the previous post. Maybe i’ll post again sometime in the future.
6. September 2009 at 20:10
Mike, I have two problems with your paragraph.
1. I don’t see why bubbles are good for investors.
2. Even if people wanted bubbles, I don’t see why that would make it happen. In China everyone thought the government wouldn’t allow the stock market to crash before the Olympics, because that would look bad. The government may well have not wanted the stock market to crash before the Olympics, but it crashed anyway.
Statsguy,
You are right that the government is heavily involved in almost all aspects of the US economy—which proves the “big lie” of leftists, almost all of whom who claim the Anglo-Saxon economic model is some sort of laissez-faire or “market fundementalism.”
But you seem to infer that this involvment in beneficial or at least needed. That we wouldn’t have safe food or labeled food without regulation, or utilities and highways unless the government built them. Or no internet. I don’t know if that is your claim, but it isn’t so. Many economists who have studied the FDA claim it has cost more lives than it has saved. The main reason food companies didn’t label ingedients 50 years ago is that few people cared. The main reason they do so now is because many consumers care. They respond to the market. Yes, there are also regulations, but no evidence they do any good.
There are whole cities in places like Texas that are built by private enterprise. In Georgia there are cities where all the essential government services are performed by private enterprise. It’s true that government does those things in most places, but there is little evidence that government involvement is needed.
I am not saying that the optimal government involvement in fields like infrastructure is zero, I am saying two other things:
1. It is probably far lower than we now observe.
2. The fact that certain services have been provided by the government, or regulated by the governemtn, in no way shows that involvemnt was either necessary or beneficial. You’d hav eto look at each case on a case by case basis.
Because I am a pragmatic liberatarian, I believe that some of the government activities that you favor almost certainly should be provided by the goverment. I think it is very unlikely that the optimal governemtn involment is zero. So I am not making a knee jerk libertarian argument here, just suggesting that we need to be careful. Thus the optimal government involvement in highways might be procuring land, and then auctioning off the rights to build and operate with tolls or adjacent land leasing rights. And the optimal role might well change over time.
I certainly agree about the much needed kick in the pants, let’s hope that we don’t get kicked the wrong way, as in 1936.
More later . . .
7. September 2009 at 01:42
Greg, I don’t understand your point at all. I just said the problems of late 2007 were real problems.
And if my macro model is so inadequate, why do I have no trouble explaining the stylized facts of business cycles and inflation? I thought if a model was inadequate, it would have trouble explaining reality. But once again in late 2008 we have a sharp break in the rate at which NGDP rises, and once again we have a sharp rise in unemployment. All easily explicable with simply a sticky wage model.
Current; you said;
“The criticism of “unrealistic mathematical models” is especially funny coming from the mouths of New Keynesians and Old Keynesians. Models are just as important to these schools as they are to Chicago economists. All of these schools use aggregate capital, aggregate consumer goods output and aggregate investment, each of these are massive simplifications. Old Keynesian theory relies on barrowloads of outrageous assumptions, such as the liquidity preference theory of interest and the marginal-propensity-to-consume model of consumer behaviour. New Keynsian economics is not much better. Keynesians economists who deride unrealistic mathematical models are saying: Do as I say, don’t do as I do.”
Very well put.
rob, The short answer is that it is a bit better than in March, but only due to luck. The rapid recovery in Asia boosted the world equilibrium real interest rate slightly, and made monetary policy slightly more expansionary–but it was nothing the Fed did, the whole QE thing amounted to little or nothing.
7. September 2009 at 02:34
123, You said,
“Anti-EMH macro implications are very clear – volatility of macro variables will be higher than estimated by EMH theories.”
I have several problems with this. First, I’m not sure it is true. For instance the 1987 stock market crash did not cause even a tiny ripple in macro aggregates, and it was arguably the largest violation of market efficiency in history.
Second, even if it was true I don’t see any important macro policy implications. With or without the EMH I favor targeting NGDP futures prices. Indeed I’d still favor this target even if someone convinced me the anti-EMH position was true.
Third, I don’t think the EMH has anything useful to say about the size of macro fluctuations. Most economists (including me) think macro fluctuations are caused by sticky wages and prices, and the EMH has nothing to say about markets where prices are sticky.
Statsguy, You said;
“So let me ask you this: WHY is it that there seems to be a strong correlation between neoclassical beliefs and predilection toward arcane mathematical “research” that (we, and now many others, agree) is often worthless? Is there a causal reason, or is it just chance?”
When I was at Chicago in the 1970s, it was far less mathematical than the elite Ivy League universities, and also far more right wing. Today George Mason University has one of the best free market economics departments, and its also less technical than many of the left-leaning departments.
Most economists in elite departments vote Democratic, and most do highly technical research. So I’m not quite sure what point you are making. There are non-technical schools of thought on both the right (Austrian) and the left (New Keynesian.) Today even monetarism if the Friedman/Meltzer variety is relatively non-technical.
Winton, They do pay some attention to expectations, and have for a while. The problem is that they pay no where near enough attention to expectations. Your description of the Jackson’s Hole conference makes me want to despair. Here we have a Fed running a deflationary monetary policy, and the criticism they are getting is from economists arguing that it should be even more deflationary. So they have to argue in defense that their policy is not excessively inflationary! Pathetic.
And they say they aren’t trying to raise inflation expectations? That is outrageous. If so, why the %$#&%#*% are they asking for hundreds of billions of dollars in fiscal stimulus. I learned in EC101 that the whole point of fiscal stimulus was to raise AD; has something changed that I am unaware of? Is the government trying to boost AD, or not?
7. September 2009 at 05:03
“That we wouldn’t have safe food or labeled food without regulation, or utilities and highways unless the government built them. Or no internet. I don’t know if that is your claim, but it isn’t so…”
(We’re so far off of Krugman here… and I hesitate to write this just because it may not be worth your time to read it.)
I’m making no claim here, other than buying food isn’t a free market experience. So to point to buying food at US supermarkets and loudly proclaim how magically the free market works (without government interference) is an odd argument at best. (Likewise, you are correct that pointing to the US as a failure of laissez faire isn’t fair either.)
Is ALL of that regulation needed? Surely not. Could some of it be better implemented? Sure. I have no doubt we’d have some sort of privatized food regulation without the FDA, but no idea how it would have evolved in practice (over the last 80 years) or how well it would really work on the scale that we currently operate. “Privatized regulation” is only as good as the system design, and often path dependent just like government agencies. For example, the credit agencies are for all intents and purposes semi-privatized regulatory bodies. With personnel that are paid _much_ more than government officials. Not only are they hardly efficient at all, and often wrong, and difficult to correct (when you have an ID theft or another false claim)… But they recently slapped AAA ratings on mortgage backed CDOs which (probably) contributed massively to the recent credit crisis. Private doesn’t automatically mean better – it all depends on system design.
As to counterfactuals (FDA regulation costs lives, private services would have stepped in if govt. hadn’t and would have done a better job, etc.), I’ve always been struck by how strong these claims are and how impossible they are to prove. (I’m not saying you are making these claims…)
As to your argument that there are BETTER ways for government to involve itself (using contracting/monitoring systems instead of direct provision), who can argue with that? It all comes down to specifics. We have some cases of really bad contracting, but almost all of this can be traced to lousy oversight/performance metrics (e.g. the privatized prison system). And, the public option (e.g., prisons) has a mixed track record as well.
Back to the FDA – the size of the counterfactual that those studies purport to test is astounding. Moreover, many use older data (prior to innovations at FDA such as fast track and the orphan drugs program), and they do not give value to the impact of standardization and aggregation of information – in other words, providing an arbiter of extremely noisy technical data that private courts seem to have a great deal of trouble addressing.
As to the question of whether FDA’s existence is good or bad… again, such a grand counterfactual. Perhaps we can draw some comparisons to less regulated sectors of the industry (herbals). Herbal manufacturers used to make some ridiculous claims (and many still do), and the private court system entirely failed to limit them. (Welcome to limited liability corporations.) Individual legal action was not worth it, and simple bankruptcy (and low cost of entry) prevented class action suits from succeeding as a regulatory mechanism. The scale of the damage (compared to profits) meant there was nothing worthwhile to go after. We might hope that individual people could monitor the different companies, but they seemed to fail to do so…
It got to the point that the legitimate herbal companies clamored for some sort of regulation to prevent their reputation from being destroyed by the BAD herbal companies. And… they certainly had the option of seeking a private company to implement this sort of “voluntary” regulation, but this did not materialize? Why? Perhaps it had something to do with the inherent amount of noise in the herbal marketplace, the low cost of entry, slow courts, high damages, inability to recover damages retroactively, etc… We could write this off as industry capture (industry trying to create barriers to entry), but the facts are that FDA regulation isn’t that big of a barrier to entry for herbal companies even now… Existing FDA regulation of herbals is fairly weak compared to its drug regulation process. Assuming companies really are legitimate and making accurate claims, the new barriers to entry aren’t that severe at all. (The regulation is primarily that they put what they say in the package, and if they make a claim on package they have to back it up with evidence – there is no “approval” process.) And, without question, there really were a lot of bad actors in the herbal supplement industry (and probably still are).
I’m not arguing that the current FDA process is the best process, or that the FDA could not adopt a more privatized approach (charge fees for registration, contract out the registration/inspection process to private companies which they they oversee, etc.). Herbals are vastly less complicated than prescription drugs, but I have a hard time seeing how a purely private approach would have worked in the case of herbal supplements.
Which, btw, you can go buy at the supermarket…
7. September 2009 at 07:22
‘I have no doubt we’d have some sort of privatized food regulation without the FDA, but no idea how it would have evolved in practice (over the last 80 years) or how well it would really work on the scale that we currently operate.’
That’s easy, it’s called branding. I.e., it’s a counterproductive business practice to poison your customers.
7. September 2009 at 11:32
‘First, I’m not sure it is true. For instance the 1987 stock market crash did not cause even a tiny ripple in macro aggregates, and it was arguably the largest violation of market efficiency in history.”
Greenspan has managed the 1987 crash really well. Also the stockmarket for the whole year 1987 was up, albeit with huge volatility. Absolute level of stockmarket was too high for 6 months only in 1987, and this overvaluation had too limited duration to cause any serious macro distortions. The 1987 crash was more important for EMH pricing of options – it turned out that deep out of the money puts were significantly undervalued for many years until the crash.
“Second, even if it was true I don’t see any important macro policy implications. With or without the EMH I favor targeting NGDP futures prices. Indeed I’d still favor this target even if someone convinced me the anti-EMH position was true.”
Counter-cyclical capital/margin regulation is the best tool to deal with the bubbles, leaving conventional monetary tools free to focus to conventional objectives (inflation or NGDP targets). But Mr. Bean from BoE has said that in the absence of such regulation the second best solution is to include asset prices into the Taylor-like rules. So in practice policymakers are seeing macro policy implications.
“Third, I don’t think the EMH has anything useful to say about the size of macro fluctuations. Most economists (including me) think macro fluctuations are caused by sticky wages and prices, and the EMH has nothing to say about markets where prices are sticky.”
While EMH might have nothing to say about the nominal fluctuations, bubbles are very important source of real shocks. There are lots of examples – 2008 oil boom, credit bubble of last decade. Bubbles in forex markets are very important in international macro.
7. September 2009 at 15:57
Dr. Sumner, you responded:
“Mike, I have two problems with your paragraph.
1. I don’t see why bubbles are good for investors.
2. Even if people wanted bubbles, I don’t see why that would make it happen. In China everyone thought the government wouldn’t allow the stock market to crash before the Olympics, because that would look bad. The government may well have not wanted the stock market to crash before the Olympics, but it crashed anyway.”
The point about investors loving bubbles is that it’s so easy to make money on the way up, usually for years. Every portfolio I had between 2002 and 2007 weren’t just winners. They were big winners and very little research or analysis was required. And just a priori, if investors don’t like bubbles, then how did so many risky firms have such high demand for their securities? The fact is, there was a great deal of money to be made in many, many companies even with unsound business practices. The only people who didn’t like the bubbles were those who stayed in too long, or who’ve suffered the collateral damage. Even George Soros has said that for selfish reasons he loves bubbles, but wants them controlled for the benefit of greater society.
7. September 2009 at 16:02
Dr. Sumner,
With respect to your second point, China is a unique case. In most counries as I see it, politicians are largely ignorant about economics and often don’t look passed the next election cycle. Afterall, it was very clear there was a housing bubble forming, but no one in a position to make a difference did nearly enough to control it, if you believe it could’ve been controlled.
7. September 2009 at 16:23
newbie, You said;
“Earlier in the thread, someone pointed out that irrational behaviour can be corrected by other market participants who take advantage of it. But what happens when following the herd is the rational thing to do? (selling shares of a sound company in a bear market, firms being pressured to buy into the subprime bubble due to profits competition makes).”
This is a common misconception. There is no such thing as “rising stoick prices” and “falling stock prices.” There are only prices that have risen and prices that have fallen. Asset prices follow a random walk, just because they have recently fallen, doesn’t make them any more likely to fall in the future. So if you believe there are bubbles, the right thing to do in a “bear market” is clearly to buy stocks, not sell. Robert Shiller has studied this issue in depth, and claims that you should buy stocks when the P/E ratio is low, that is, when you are in a bear market.
I am an agnostic on Shiller’s theory, I don’t think we know enough to say whether it is right or wrong. But if it is right, then people need to know the implication of the theory. I don’t think they do, which might be part of the problem of market irrationality. Thus perhaps those people who don’t believe in the EMH might be precisely those people who make markets irrational.
I not sure what you mean by me advocating monetary intervention. I don’t favor more monetary intervention, just different monetary intervention. A monetary policy that is more stable.
Also see Jon’s reply, which I agree with.
I don’t think government intervention has much to do with whether bubbles exist or not. But it may create asset price movements that are falsely perceived as bubbles, or more likely negative bubbles.
Patrick, Thanks. It shows how we never learn. People have predicted the death of neoclassical econ for a long time. Marx, then Keynes, now the behavioralists, but it keeps coming back. Constraints, opportunity costs, etc, never go out of style.
Statsguy, I know little about herbal medicines; I guess I always assumed they were placebos. Which is fine, as placebos are highly effective, and so I presume herbals are also highly effective.
I agree with you that there is much uncertainty about what the government can accomplish. I guess I tend to err on the side of the free market, unless the government can be shown to make things better. You could argue trhat that is political bias, but you could also argue that it is a sort of inductive argument based on the fac that lots of government itnerventions that looked good initially turned out to look bad. The other test I would look for is a strong theoretical argument. Something like externalities. I am not impressed by theoretical arguments such as consumer ignorance, for instance. Colleges, cars and TVs are all far too complicated for the average consumer to be able to appraise for quality. How could they possibly know which products are best? And yet somehow even high school dropouts “know” that Sony is better than Vizio, Lexus is better than Ford, and Harvard is better than Bentley. So I have a stong prior that lack of information is not a good excuse for product regulation, unless someone can present some stong statistical evidence that it is.
You mention that the bigger herbal companies don’t bother with a private (voluntary) regulatory scheme, but want the government to come in and get rid of the “bad” companies. I think that attitude is very revealing.
I understand your argument was that people couldn’t cite groceries as a free market, but everything is a matter of degree. The grocery industry in the US is less regulated than in Germany, which is less regulated than in North Korea. So I still think Patrick’s argument has some merit. At the same time I think your response is reasonable, and there is nothing there that I strongly take issue with.
123, You said,
“‘First, I’m not sure it is true. For instance the 1987 stock market crash did not cause even a tiny ripple in macro aggregates, and it was arguably the largest violation of market efficiency in history.”
Greenspan has managed the 1987 crash really well. Also the stockmarket for the whole year 1987 was up, albeit with huge volatility. Absolute level of stockmarket was too high for 6 months only in 1987, and this overvaluation had too limited duration to cause any serious macro distortions. The 1987 crash was more important for EMH pricing of options – it turned out that deep out of the money puts were significantly undervalued for many years until the crash.”
Greenspan did only what the Fed is supposed to do, target nominal aggregates. But I don’t understand the argument that prices were only too high for six months, and that that somehow affected the results. The 1987 bubble actually looked a lot like 1929 and 2000. In all three cases prices of stocks had been in a major bull market for years, and then spiked sharply higher in the last year before crashing. I’m not saying they were identical, but I don’t thnik any differences were large enough to explain the massive difference in macro outcomes.
Regarding Mr Bean and Taylor type rules, that is exactly the problem. The Taylor rule cannot be “fixed” by adding asset prices, as we’d never agree on a formula, and indeed there is no correct formula because the model is flawed. Money was not “too easy” in 1928-29, despite the stock boom. Instead of trying to fix a broken Tayler Rule, we need a forward-looking monetary policy. If we had had one, the crash of late 2008 never would have happened, with or without the housing bubble.
I do agree that the perception of most economists will be exactly what Mr. bean suggests, but it won’t fix the problem. Just as Sarbannes-Oxley didn’t fix the problem of corporate misbehavior in banking. To get the proper regulation you first need to correctly diagnose the problem. And our current problems were not caused by asset price instability, they were caused by deflationary monetary policies.
You said;
“While EMH might have nothing to say about the nominal fluctuations, bubbles are very important source of real shocks. There are lots of examples – 2008 oil boom, credit bubble of last decade. Bubbles in forex markets are very important in international macro.”
The 2008 oil price boom was not a bubble, it reflected massive demand in developing countries bumping up against a relatively inelastic short run supply. (or possibly cartel-like action by producers, which is also different from a bubble.) Nor did the high oil prices have much of a macro impact on the economy, (although it was slightly contractionary for the US.) But let’s first solve the big (nominal) problems and then worry about the small real problems.
7. September 2009 at 17:23
Mike, It is never easy to make money ex ante. Ex post it might look easy when markets have risen. But no one knew in 2002 that asset markets would appreciate so strongly. Stock prices follow a random walk. There is no such thing as a rising stock market, only a stock market that has risen.
The reason the government did nothing about the housing bubble was not that they have a short run bias, but because they couldn’t agree that there was a bubble. The Fed looked at the situation, and decided that it wasn’t clear that a bubble had formed.
7. September 2009 at 19:47
Scott, I’m an engineer / scientist, not an economist. I have never-the-less spent considerable time to understand some of the fundamentals of classical economic theory and the foundations upon which it is based. Among many other things I’ve found, one of the foundations appears to be severely lacking in validity. EMH is indeed only a hypothesis… not a proven theory, nor certainly a fact. In science, when a hypothesis is described, it must be tested for validity. The EMH has been tested numerous times —- and found false on several counts. You call these failures in tests of the hypothesis to be “anomolies”. In science “anomolies” (even one) are a failure of the hypothesis. So… what I cannot fathom is why any economist would continue to use EMH as a foundation for the field of study being pursued when it has long since been shown to be disproven by the “anomolies”.
Moreover, the EMH is also dependant on another assertion — that is “rational behavior in economic decisions” (for the market as a whole — not as an individual). This is, in the first count a circular argument, and in the 2nd count the same “anomolies” describe irrational market behavior — whether or not you can find a rational explanation is immaterial, since the market as a whole behaved irrationally — you yourself said that you have no explanation for rational reasoning for the housing and dot.com bubbles.
Thus, as a scientist, and an engineer, I find it absurd that the field of economics (in academia) would continue to subscribe to the EMH and as well as its underlying assertion of rational behavior. It seems to me that, reading historical theories of the development of economics as a field of study, that the original premis is one that is self-serving rather than an objective foundation which hypothesis testing has shown to NOT having been proven false. It is well known that a theory may never be proven true, but it is accepted as true when all hypothesis used to test the theory’s assertion have been unable to show the theory false.
Since there have been so many instances over the years where the hypothesis has been tested and found false, I cannot but wonder why economists continue to cling to a theory that is false, and use it’s assertion as ‘truth’ to further the field of study or their own beliefs.
8. September 2009 at 05:55
‘People have predicted the death of neoclassical econ for a long time. Marx, then Keynes, now the behavioralists, but it keeps coming back. Constraints, opportunity costs, etc, never go out of style.’
And that was exactly the message Stan delivered to readers of his book.
8. September 2009 at 06:07
‘EMH is indeed only a hypothesis… not a proven theory, nor certainly a fact. In science, when a hypothesis is described, it must be tested for validity. The EMH has been tested numerous times “”- and found false on several counts.’
William, for the EMH to be found to be false, there has to have been an alternative theory shown to work better in explaining market performance. A theory that could be exploited. I.e. that would produce superior returns to a long term buy and hold strategy.
So far (as I know), no cigar. In fact, I just detailed an alternative (at least for internet commerce) offered about a decade ago–the ‘First Mover Advantage’–that had a lot of people buying into it, to their ultimate disadvantage.
8. September 2009 at 06:38
ssumner:
“I guess I always assumed [herbals] were placebos. Which is fine, as placebos are highly effective, and so I presume herbals are also highly effective.”
And since FDA approval increases their legitimacy in the eyes of many people, and credibility increases the placebo effect, we can conclude that even if FDA regulation is not getting us substantively better products it is still adding _tremendous_ value to our health and wellbeing. 🙂
… I look forward to your next post.
8. September 2009 at 07:26
Dr. Sumner,
The point is not anticipating bubbles as an investor, but merely investing in a bubble climate and realizing larger gains on a wider range of bubble assets than one would without the bubble.
With respect to recognizing bubbles, given the now famous graph displaying the historically steepest spike in home prices from 2003-2006 or so versus personal incomes, etc., how could anyone miss a bubble in housing?
My brother was a mortgage broker in Colorado and told me a huge washout in housing was coming, because of the liar, no down payment, teaser rate mortgages he was dealing.
Also, of course, there are many other metrics for determining whether one is in fact in an asset bubble. Earnings yield ratios can reveal implicit levels of speculation on stocks, for example. Cap rates in real estate, along with replacement costs versus market prices, adjusted for population and incomes…
I didn’t know a financial meltdown was coming, but I anticipated a sharp economic downturn on these bases alone, and I didn’t do much in the way of analysis.
I think the Soros view on markets is much closer to reality. They are inherently relatively inefficient with respect to strong EMH claims.
8. September 2009 at 07:42
Problems with Strong Versions of EMH from a non-economist:
1.) If this approach is so rational and well-supported, then why do so many reject it, at least implicitly and try in vain to beat the market? Does this suppose that a market can be relatively more rational in the aggregate than at the micro level?
2.) Today, with the influence of large, institutional investors, I’d think there are opportunities for smaller investors who can invest and divest much more nimbly.
3.) As Milton Friedman argued, if strong versions of EMH led even the brightest stock pickers to fail to beat the market on average, might this not cause many of them to give up and do other things? Then, markets might become even less efficient, again providing more opportunities for more talented investors such that a cycle occurs.
4.) Non-financial risk aversion (in 80% of any large population) and emotional decision-making at times, even if rarely, favor implicit memory system processing, which relies on more well-established learning and instincts, many of which are antithetical to rational analysis. Known personality differences, such as those with risk-neutral or risk-seeking behavior can then outperform the broader market, in principle.
5.) Of course, insider trading and market manipulation may undermine the strength of EMH.
6.) Markets have natural incentives to form bubbles and create temporary, but consequential disequilibria, as most participants benefit handsomely from booms.
7.) Mimickry of trading strategies is not as straight-forward or as successful as some strong EMH adherents claim. The brain faces vast numbers of permutations at every stage of decision making to work from one stage to the next in successfully working toward an ultimate goal.
A former investment banking employee who was charged with teaching trading strategies to what he describes as very bright traders said his training failed to take hold. Old, suboptimal habits and instincts could not be eradicated. There is a link to a discussion is which he discusses this in the context of neuroeconomics:
http://www.youtube.com/watch?v=A_B0rvSxUqY
8.) Investors employ a large number of different trading strategies, which is related to the permutation problem mentioned above, many of which are dubious at best, even with respect to large investors. How does one come to the conclusion that a market with such heterogenous approaches happens to be more efficient than any traders or investors on average?
9.) EMH is not testable in the most rigorous sense and some ofthe same data, as in much of economics, can be interpreted in diametrically opposing ways.
10.) Some strong EMH adherents have de-emphasized even fundamental analysis, trusting markets to get valuations as close to correct as currently possible, even when the fundamentals versus the market are clearly out of whack, as is the case with bubbles.
11.) What about problems of assumptions of normality in the distributions of stock prices? What about fat tails?
12.) How does one overcome the fact that if many investors adopt buy and hold strategies, it leaves more opportunities for short-term traders, especially during booms in which even apes throwing darts can pick short-term winners more often than losers? During the last boom, I was often long with options, as the risk/reward was great. Even if I lost out on 70% of my positions, I could still profit handsomely, and beat the markets by a long, long shot. Sure, maybe I was just lucky, on a long thin tail, but this wasn’t uncommon among those I knew, anyway.
13.) Trading without a sound macroecnomic context can be disasterous, and yet even many professional economists don’t seem to understand macroeconomics that well. Just look at Krugman on old Keynesians, the number of liquidationists out there, and those like Feldstein who don’t understand that interest rates don’t rise when government borrowing doesn’t exceed the collapse in private borrowing.
14.) There are always opportunities for at least short-term arbitrage, allowing some to make supernormal profits.
15.) Agent-principal problems abound.
16.) There often aren’t incentives for good investment performance from securities brokers and oher financial services professionals, as profits depend on transaction volumes and in some cases, pump and dump strategies, or those to help investment banking clients.
17.) Portfolio distributions are often influenced by emotional considerations other than financial ones, leading to suboptimal results and opportunities for sounder investors.
18.) Other explanations can account for observations, such as that 70% or more of professional money managers fail to consistently beat markets over the long term. These include some items mentioned above, along with the complexity of markets and sometimes vastly different approaches to dealing with uncertainty, often without the use of explicit models such as CAPM and its related successors.
19.) What about asymmetric information in general?
I think I have some more potential problems with strong EMH theories. Perhaps I’ll add some more later.
8. September 2009 at 09:07
You wrote,
” Greenspan did only what the Fed is supposed to do, target nominal aggregates. But I don’t understand the argument that prices were only too high for six months, and that that somehow affected the results. The 1987 bubble actually looked a lot like 1929 and 2000. In all three cases prices of stocks had been in a major bull market for years, and then spiked sharply higher in the last year before crashing. I’m not saying they were identical, but I don’t thnik any differences were large enough to explain the massive difference in macro outcomes. ”
There is a big difference. 1987 crash is invisible in annual data – Dow was up for the year.
“Regarding Mr Bean and Taylor type rules, that is exactly the problem. The Taylor rule cannot be “fixed” by adding asset prices, as we’d never agree on a formula, and indeed there is no correct formula because the model is flawed. Money was not “too easy” in 1928-29, despite the stock boom. Instead of trying to fix a broken Tayler Rule, we need a forward-looking monetary policy. If we had had one, the crash of late 2008 never would have happened, with or without the housing bubble.”
I mostly agree with you here, but asset prices are a forward looking indicators. So anti-EMH types might say that it makes sense to include asset prices together with the short term market-based inflation or NGDP forecasts when setting the policy variables.
“I do agree that the perception of most economists will be exactly what Mr. bean suggests, but it won’t fix the problem. Just as Sarbannes-Oxley didn’t fix the problem of corporate misbehavior in banking. To get the proper regulation you first need to correctly diagnose the problem. And our current problems were not caused by asset price instability, they were caused by deflationary monetary policies.”
Yes, most of the damage was done by bad monetary policies. But calling Fed deflationary is the same as calling Obama a socialist – 100% true but not effective in persuading opponents.
”
The 2008 oil price boom was not a bubble, it reflected massive demand in developing countries bumping up against a relatively inelastic short run supply. (or possibly cartel-like action by producers, which is also different from a bubble.)”
Oil futures markets were in contango, so high prices reflected long run supply concerns. I think that last summer market’s long run supply forecast was a bit crazy and bubble-like.
” Nor did the high oil prices have much of a macro impact on the economy, (although it was slightly contractionary for the US.)
”
One way high oil prices were contractionary is that they created expectations of Fed tightening.
But this talk about oil detracts form the main bubble – the credit bubble that did the most damage. Have you seen any studies that prove that EMH is correct in credit markets?
“But let’s first solve the big (nominal) problems and then worry about the small real problems.”
Real problems are important too. Japan is a good example, as so much time has passed from the crash that nominal problems are no longer limiting growth there.
8. September 2009 at 09:17
This is timely three USC marketing guys have a paper showing not only that network effects aren’t strong enough to lock-out superior products, but may even speed the process of the late-comer with a superior product displacing the prior leader.
8. September 2009 at 13:33
William W. Childers writes:
You’ve got this wrong. Rational expectations is the notion of complete omniscient efficiency (“everything is known”). The EMH is a much weaker form of this idea.
Although we’re sometimes sloppy in our language, the efficiency in the EMH is not in finding the “right” price or the abstractly optimal solution as would befit an omniscient entity. No, the efficient market hypothesis is that all known information is incorporated into the market. “All known information is used, not wasted.” i.e., markets use the available information efficiently.
Yes market behavior can be irrational (in hindsight), but the EMH is a statement that markets do not simultaneous behave irrationally and everyone KNOWs that this is so.
8. September 2009 at 18:00
Entertaining post, Scott.
Ironic that the simplest solution to the Baby Sitters club parable that Krugman describes is printing more scrip 🙂
8. September 2009 at 20:03
Patrick – you wrote: “for the EMH to be found to be false, there has to have been an alternative theory shown to work better in explaining market performance”
This is not the case in science and engineering. If found false and nobody has an alternative to replace the false theory, then work ensues to come up with a theory that can be supported by tests of alternative hypothesis. In short the answer is “gee, sorry ’bout that… but I don’t YET know of another theory to replace the one that’s now been found in error.”
There may be some aspects of the failed theory that, if appropriately broken down into sub-sets, may lead to the theory that can actually be supported by tests of the hypothesis used to test the theory.
8. September 2009 at 20:21
Jon, since its the capital markets to which EMH applies, and the entire purpose of hypothesis is to explain why capital markets in a free enterprise system “work” to best distribute capital efficiently — that is, efficiency of capital, which means it is distributed with minimal waste, in a timely manner, and with a net positive return… these are the elements of “efficiency”.
There are thus two predicates (assumptions) required to assert EMH applies to serve the purpose of efficient capital distribution … 1) that rational market behavior applies, and 2) that all available information is applied. If either of those assumptions is false, then EMH is shown to be false — that is, it does NOT serve to provide for the “best” means for the useful distribution of capital.
8. September 2009 at 20:29
By the way, an “anomoly” is by definition an event which cannot be explained by or within the confines of the existing theory or hypothesis — hence, in science, an “anomoly” requires either that the theory or hypothesis 1) be rejected, and 2) then modified in such a manner as to include the “anomolous” event or events within their scope.
Very often, in science, an anomoly is asserted to be due to “measurement error” or “experimental lapse”… and if this is accepted, then the operating theory / hypothesis is still in force — until several other experiments to test various aspects of the theory / hypothesis also show “anomolies” in the results. At which point, the hypothesis is not only suspect, but for most objective observers, it’s been shown false.
8. September 2009 at 20:56
William:
#2 subsumes #1. These the most direct statements of the EMH that I am familiar with:
A market in which prices always “fully reflect” available information is called “efficient.”
“A market is efficient with respect to information set θt if it is impossible to make economic profits by trading on the basis of information set θt.”
Notice now that you’ve attached conditions quite apart from the EMH as a term-of-art.
So you write:
These are not the EMH. Although efficient markets–in the technical sense–may quite often deliver those attributes as well.
You apparently agree with Patrick. As long as theories are useful they get used until something more useful comes along.
8. September 2009 at 21:07
William:
Please see, FAMA, Eugene F., 1970. Efficient Capital Markets: A Review of Theory and Empirical Work, Journal of Finance, 25(2), 383-417.
9. September 2009 at 03:11
William, The laws of supply and demand have also been tested, and found false on many occasions. Does that mean that model is useless? Clearly not. S&D is a very useful approximation of reality. Since you are a scientist I would use the following analogy. Newton’s laws of motion have been tested and anomalies have been found, such as the famous eclipse experiment from around 1919. It was shown that Einstein’s Theory of Relativity better explained the data. But engineers still find Newton’s laws a very useful approximation of reality. That’s how I feel about the EMH. It certainly is not precisely true, but I have found the hypothesis to be extremely useful in helping me to understand economic phenomena.
In addition, the anomalies you refer to do not disprove the EMH. The are mostly a product of data mining, and hence a misuse of the classical principles of statistical significance testing. Since you seem to be a stickler for following accepted principles, I don’t think you would want to put any weight on the studies showing anomalies, as most did not use proper procedures.
The fact that I can’t find a rational reason for some particular market movement doesn’t imply that there isn’t one. In addition, I would argue that even if the tech and housing bubbles did involve some irrationality, that would not change my view that the EMH is very useful.
Patrick, Thanks. I agree.
Statsguy, You might be right, but what worries me is those highly effective drugs that save 100,000s of lives that are held back for many years by the FDA. There is more to the pharmaceutical industry than placebos. Plus, I think most people get the placebo effect from reassurance by their doctor, not the FDA.
I don’t know when my next post will be, in the last couple weeks it has been a nightmare trying to keep up this blog with a Chinese DSL line. I often lose large blocks of what I write. I am not able to keep up with other blogs. Some days I can’t even get Yahoo. It is a real stuggle. I can’t open most links. Often I cannot block and move quotations.
Mike; You asked;
“With respect to recognizing bubbles, given the now famous graph displaying the historically steepest spike in home prices from 2003-2006 or so versus personal incomes, etc., how could anyone miss a bubble in housing?”
The government regulators missed it. And also all the highly intelligent people who bought homes in 2006 because they believed we were in a bull market, and who believed investors make money in bull markets. Except when they don’t. I have people post here who tell me that bubbles are a good time to buy–are they also stupid?
Mike#2, I just answered a bunch of your questions on another comment thread, where you asked a similar question. You’ll want to check that out. If you have follow-ups feel free to put them here, as it is easier to keep up with just one thread going.
123, I don’t see what’s so magical about one year. The stock market was regarded as absurdly overvalued in early 1928, and finished 1929 even higher. By April 1930 it was even higher. I admit 1987 isn’t eactly the same, but was the difference big enough to explain the difference between no effect and a catastrophic collapse in AD? I doubt it.
I don’t think we are that far apart on asset prices and monetary policy. Here’s how I would state my argument. NGDP futures prices would contain everything about asset prices that is relevant to monetary policy.
Regarding nominal and real problems, I should have been more specific. I meant nominal problems are more important in business cycles. I certainly agree that real problems are more an issue with long run growth, such as in Japan.
I see no reason why oil prices wouldn’t still be high if the global boom had continued. Even with the worst global collapse since the Depression oil prices are $70, vs $10-20 back in 1998. That’s impressive. Isn’t it clear we have a supply issue (not necessarily peak oil, but something close.) BTW, I think it quite possible oil could fall sharply in a few years if electric cars pan out. These things are extremely hard to predict.
9. September 2009 at 03:19
Jon, I agree with you about the EMH, but even the Ratex theory doesn’t assume omniscience.
Thruth, Good point. Krugman never seems to give that point the emphasis it deserves.
William, You said;
“If either of those assumptions is false, then EMH is shown to be false “” that is, it does NOT serve to provide for the “best” means for the useful distribution of capital.”
This is wrong. It is the best way of allocating capital until something better is found. The fact that one particular theory doesn’t work perfectly certainly doesn’t imply that some alternative theory is better. I doubt whether we will ever find a better way of allocating capital. We certainly haven’t so far, even if the EMH is false.
9. September 2009 at 06:04
Dr. Sumner,
You posted:
“Mike; You asked;
“With respect to recognizing bubbles, given the now famous graph displaying the historically steepest spike in home prices from 2003-2006 or so versus personal incomes, etc., how could anyone miss a bubble in housing?”
The government regulators missed it. And also all the highly intelligent people who bought homes in 2006 because they believed we were in a bull market, and who believed investors make money in bull markets. Except when they don’t. I have people post here who tell me that bubbles are a good time to buy-are they also stupid?”
The point I’ve been making all along is that many people are actually smart to take advantage of housing bubbles or any bubbles for that matter. Sure, it caused pain for many, but look at the equity gains leading up to this crisis and the other great investment opportunities. Those who bought their homes to actually live in, because they love their homes just lost some equity, but will get some of it back. Many who wanted short-term investments in houses did very well. Subprime borrowers took hits, but they had little to lose in the first place and some of them actually are managing to keep their homes. Of course, unemployment is a problem for some homeowners how, but this could have mostly been avoided with better monetary policy.
Those flippers or converters to condos are the ones that took the really big hits, and so what? The bubble was very good to many, many people.
9. September 2009 at 06:17
Dr. Sumner,
Thank you for taking the time to respond to some of my many points on strong EMH. Sine you say you’re not an adherent to strong EMH anyway, perhaps the discussion is losing importance. Clearly, I’d think EMG is true to a degree.
I just want to ask at this point, you didn’t address the following points against strong EMH:
1.) If this approach is so rational and well-supported, then why do so many reject it, at least implicitly and try in vain to beat the market? Does this suppose that a market can be relatively more rational in the aggregate than at the micro level?
2.) Today, with the influence of large, institutional investors, I’d think there are opportunities for smaller investors who can invest and divest much more nimbly.
3.) As Milton Friedman argued, if strong versions of EMH led even the brightest stock pickers to fail to beat the market on average, might this not cause many of them to give up and do other things? Then, markets might become even less efficient, again providing more opportunities for more talented investors such that a cycle occurs.
4.) Non-financial risk aversion (in 80% of any large population) and emotional decision-making at times, even if rarely, favor implicit memory system processing, which relies on more well-established learning and instincts, many of which are antithetical to rational analysis. Known personality differences, such as those with risk-neutral or risk-seeking behavior can then outperform the broader market.
14.) There are always opportunities for at least short-term arbitrage, allowing some to make supernormal profits.
16.) There often aren’t incentives for good investment performance from securities brokers, as profits depend on transaction volumes and in some cases, pump and dump strategies.
17.) Portfolio distributions are often influenced by emotional considerations other than financial ones, leading to suboptimal results and opportunities for sounder investors.
18.) Other explanations can account for observations, such as that 70% or more of professional money managers fail to consistently beat markets over the long term. These include some items mentioned above, along with the complexity of markets and sometimes vastly different approaches to dealing with uncertainty, often without the use of explicit models such as CAPM and its related successors.
19.) Even if all information taken into account by the market increases its effiency, the proliferation of it is uneven and there are often large amounts of information missed by the majority, but exploited by a minority of investors. These are dynamic opportunties that once missed, cannot be mimicked, as opportunities dry up before competition heats up.
Am I to take it that you have fewer problems with these points than those you specifically pointed out?
9. September 2009 at 06:47
I’ll try one more time to explain why William is not grasping the implications of EMH.
If one has money to invest, i.e. one wishes to defer consumption today to enhance his consumption in the future, he has to choose how to invest. Options are:
Bury the money in a hole in his backyard, or put it into a safe deposit box in a bank.
Buy Krugerands, and do the above with them.
Go to Las Vegas and play the roulette wheel or slot machines.
Buy corporate or government bonds.
Buy equities.
If the latter, choose from the available theories as to what will maximize his returns.
Now, the EMH can be ‘tested’. And, in fact, has been. AFAIK, no one has found a superior alternative to simply buying into indexed mutual funds (that is, a long term, buy and hold strategy of diversified stocks).
Even if the ‘investor’ does go the Las Vegas route, AND hits his lucky number ONCE, that doesn’t invalidate the superiority of EMH. It remains the ‘best’ strategy.
In investing, you can’t bet something with nothing.
9. September 2009 at 06:49
Oops, should be ‘beat’, not ‘bet’.
9. September 2009 at 13:53
Scott,
often lose large blocks of what I write.
I have a suggestion that you need to learn by heart and practice regularly. (These instructions assume you use a PC. If you use a Mac, they have similar shortcuts.) Control A, followed by Control C.
If the post is lost in the intertube ether, then use Control V.
This is a lesson I have learned the hard way, after making more than a couple of posts (elsewhere) that were inspired, and lengthy, responses to someone else, only to see them dissapear, and to not have the will or time to try to rewrite them.
9. September 2009 at 14:10
“I don’t see what’s so magical about one year. The stock market was regarded as absurdly overvalued in early 1928, and finished 1929 even higher. By April 1930 it was even higher. I admit 1987 isn’t eactly the same, but was the difference big enough to explain the difference between no effect and a catastrophic collapse in AD? I doubt it. ”
According to Shiller’s P/E, stockmarket was much more expensive in 1929 than in 1987. In 1987 stockmarket got a little bit overvalued for six months, in 1920s there was a huge overvaluation lasting four-five years. So it does not surprise me that 1987 had no macro effects, but it might make sense to look to asset prices as a secondary cause of crisis in 1929. I have a question for you, is the asset price overvaluation in 1929 more comparable to 2000 or to 2008? In 2000 a narrow asset class (dot com and telco stocks) was extremely overvalued, but the financial sector together with the rest of the economy was in a reasonable condition. In 2008 credit and real estate was extremely overvalued, so financial sector was in a deep trouble.
“I don’t think we are that far apart on asset prices and monetary policy. Here’s how I would state my argument. NGDP futures prices would contain everything about asset prices that is relevant to monetary policy.”
Anti-EMH types would say that if during asset bubbles NGDP growth would be reduced a little bit, stability of real output would be greater. Or from the practical perspective, Fed would be able say to financial sector lobby that if they make big noise about proposed tightening of capital requirements, Fed will increase interest rates instead.
“Regarding nominal and real problems, I should have been more specific. I meant nominal problems are more important in business cycles. I certainly agree that real problems are more an issue with long run growth, such as in Japan.”
I guess that in Japan current real problems are the product of the 80s bubble. So if Japan had good monetary policy in the 90s, the result would be stagflation.
“I see no reason why oil prices wouldn’t still be high if the global boom had continued. Even with the worst global collapse since the Depression oil prices are $70, vs $10-20 back in 1998. That’s impressive. Isn’t it clear we have a supply issue (not necessarily peak oil, but something close.) BTW, I think it quite possible oil could fall sharply in a few years if electric cars pan out. These things are extremely hard to predict.”
Last summer oil market did not agree with you. Last summer oil market was saying that there is no chance for electric cars in a few years, that it is very easy to predict that peak oil scenario is the only way. Last summer oil market forecast was extremely biased. Crude at 100 was reasonable last summer, but not at 147.
But the real problem with EMH is credit markets. There are no studies that support EMH for credit. Instead of efficient price of credit we get moral hazard and bubbles.
9. September 2009 at 15:55
Mike, If you think bubbles are a good time to buy, then you have a very different definition of “bubble” than almost anyone else. The standard view is that bubles are when assets are overpriced, and hence bubbles are a good time to sell. I can’t respond to your comments on bubbles unless I know how you define them.
Regarding the list of questions:
1. The EMH doesn’t say the average person won’t try to beat the market, it says the average person won’t succeed. That is very different. Why do they try? Maybe they don’t believe the EMH.
2. I don’t see any obvious advantages to being small.
3. I agree.
4. I agree, but the advantage is probably so miniscule that it isn’t very important.
14. I agree, but these are rare and the profits are usually very small.
17, I agree, the the advantage is probably miniscule.
18. I don’t follow this argument. The failure of managed funds to beat the market is an incredibly powerful argument for the EMH, if you are a pragmatist like me. It says, “even if the market is not perfectly efficient, it is efficient enough that most Wall Street experts aren’t able to beat it.”
19. I agree but the experts then move the price to the right level, making potential profits for the average expert fairly small.
happyjuggler0, I was wondering why I didn’t see more inspired posts from you. (Just kidding.) Seriously, I do the same thing, the next time through my comment is much more prefunctory. So I know how you feel. And thanks for the tip on getting the stuff back.
More later . . .
9. September 2009 at 16:52
Patrick, I agree.
123, You said;
“in 1920s there was a huge overvaluation lasting four-five years.”
I saw a recent academic paper arguing that stocks weren’t even overvalued in 1929. It seemed just as convincing to me as Shiller’s argument. But if what you say here is true, then stocks would have also been massively overvalued in late 1929 and early 1930. Hence that could not have provided any reason for the sharp plunge in AD in late 1929 and early 1930. I am convinced that if you wrote down a mathemetical model linking stock prices and AD, you couldn’t come close to fitting both 1929 and 1987, the differences simply aren’t big enough. Plus any change in AD should be related to the change in wealth, not the fact of whether stocks were “overvalued” according to Shiller’s formula.
And of course the 1929 was not some exogenous shock, but rather reflected the stock market’s rational response to the US sliding into a deep slump.
you said;
“Anti-EMH types would say that if during asset bubbles NGDP growth would be reduced a little bit, stability of real output would be greater.”
I don’t see any evidence for this view. Real GDP growth tends to be faster during asset price bubbles, so even NGDP targeting would mean countercyclical movements in inflation. In any case, one step at a time. Even NGDP targeting would have called for much tighter policy during the asset bubbles of 1999-2000 and 2004-06, than what actually occurred. Let’s see how that works first, before taking the next step.
You said;
“I guess that in Japan current real problems are the product of the 80s bubble. So if Japan had good monetary policy in the 90s, the result would be stagflation.”
I don’t see how a bubble can create a long term real problem. In my view the “real” problem is Japanese economic policy, which is too pro-business, or pro-special intrerest.
You misunderstood my comment about electric cars. I agree that electric cars won’t be a big factor for a few years, and hence I think the $147 price was fully rational. If the recession had not occurred, then I think prices would still be around $147. Why not? Consumers were willingly paying that much and auto sales in the 3rd world were booming. China recently overtook the US as the world’s biggest car market, and will be twice the size of the US car market before too long. That’s a lot of demand for gasoline. I don’t see any problem with $147 oil.
Regarding electric cars, I should have said a couple decades. But again, that wasn’t a prediction, just an observation as to why oil prices are so hard to predict.
10. September 2009 at 06:19
“So if Krugman really does believe that the problem we face is inadequate demand, he needs to come up with a reason why this problem doesn’t call for a better monetary regime, and instead requires the rehabilitation of an old-fashioned Keynesianism that confused money and credit, that confused saving gluts and liquidity traps, that ignored policy expectations, and that had no model of the trend growth rate of NGDP.”
Scott, your views on “Old Keynesianism” deserves its own post.
10. September 2009 at 07:22
Scott: in the wsj today, boe considering negative reserve rate
10. September 2009 at 07:33
I agree with 123 that the stock market was over-valued in the late 1920s. I think the distortions that come about through the Austrian business cycle theory were in play then.
However, I don’t think this is a major criticism of EMH. Because, entrepreneurs and investors could not consider the whole structure of production. We can now write-off plans as unsustainable that investors couldn’t tell were unsustainable at the time.
I’ll write more about this EMH controversy when I have time.
10. September 2009 at 07:38
Dr. Sumner,
You posted:
“Mike, If you think bubbles are a good time to buy, then you have a very different definition of “bubble” than almost anyone else. The standard view is that bubles are when assets are overpriced, and hence bubbles are a good time to sell. I can’t respond to your comments on bubbles unless I know how you define them.”
I consider a bubble to be a situation in which asset prices reflect a demand that is based on increasing speculation. For example, during the tech bubble there were web-based companies with extraordinarily high p/e ratios that ended up failing miserably. I haven’t looked it up, but how many failed for every Google?
Or, to take the recent housing run up, home valuations and the costs of borrowing versus buyer income(sometimes merely stated), replacement costs, etc. very much suggested wild speculation in terms of expected increases in the collateral for the mortgages.
Tulipomania would be another example.
However, the point that bubbles are a good time to buy is easy to support when you consider that bubbles can last for sometimes several years. Even if one gets in during the last year or two, there can be easy, signicant gains to be made. I still had very profitable long positions in early 2007. House flippers made bundles during the recent housing boom and some homeowners had larger home equity lines of credit to start businesses, among other ways to invest their money.
In the tech bubble in particular, a chart of the S&P 500 shows a significant increase in the rate growth of the index versus a decades old baseline. This was even more true of Nasdaq. So, even some index fund investors easily profited generously during this particular boom on the way up.
10. September 2009 at 08:51
Tulipmania is an interesting case. Read Doug French on the subject
http://mises.org/story/2564
11. September 2009 at 01:25
I know the discussion about alternative medicine is almost over, anyway…
In the US the reason courts don’t prosecute companies for making dangerous “herbal” drugs is because of legislation making it difficult for them. It’s not the case that because something isn’t a drug normal consumer laws apply. Rather there are special and rather lax laws for alternative medicine. Unsurprisingly these came about because of pressure from the alternative medicine industry.
I would like to give you a link, but Scott’s blog is refusing my post if I do.
Besides, does any really think that anything can be done to help those stupid enough to take such remedies.
11. September 2009 at 01:26
The link is
www dot sciencebasedmedicine dot org slash questionmark p=530
11. September 2009 at 04:07
knapp, I will do one someday. But back in February I did do a post that criticized the General Theory. It contains some specifics.
Jon, Thanks, I’ll try to find it.
Current, Have you read any of the academic studies that argue stocks were not overvalued in 1929? And if so, what do you see as the flaws in those studies?
Mike, You may think they are a good time to buy, but the evidence suggests just the reverse. Asset prices follow a random walk. Just because they have been rising, doesn’t make them more likely to rise than fall in the future. Thus if an asset is overvalued, it is likely (by definition) to be worth less in the future. So you should sell.
If someone tells me “buy X, because it’s soared 75% in the last two years,” I really don’t know what to say. That seems like a crazy reason to buy something. I’d be more likely to run out and sell it short. (If I was a market timer.)
Current, I read the Garber book a while ago, and thought his presentation seemed fair. As I recall he said the boom was mostly based on fundamentals, although prices may have overshot at the end. That is the sort of pragmatic view of the EMH that I have. A useful theory, but not perfect.
Thanks for the info on herbal remedies, but I don’t know enough to have an intelligent opinion. Maybe others can contribute.
11. September 2009 at 08:41
Dr. Sumner,
You added:
“Mike, You may think they are a good time to buy, but the evidence suggests just the reverse. Asset prices follow a random walk. Just because they have been rising, doesn’t make them more likely to rise than fall in the future. Thus if an asset is overvalued, it is likely (by definition) to be worth less in the future. So you should sell.
If someone tells me “buy X, because it’s soared 75% in the last two years,” I really don’t know what to say. That seems like a crazy reason to buy something. I’d be more likely to run out and sell it short. (If I was a market timer.)”
Has the fact that higher prices/fundamentals prevented bubbles from continuing and even accelerating for years? Some people are doing the buying. The demand is coming from somewhere. , and anecdotally, at least, I’ve heard plenty of people say that bought certain stocks or even houses because of the price appreciations in recent history.
You might think the market is in a random walk and that buying into something over-valued is ludicrous, but the vast majority of people don’t approach your sophistication. You probably wouldn’t buy a house with a teaser rate mortgage, speculating that increased future income and equity gains would help you pay future higher rates, but millions of people did.
Likewise with the tech boom in the late 90s. early 2000s, the market was over-valued by every metric I can think of, yet there wasn’t as much flattening of demand for index funds as one would expect with increasing realizations of price unsustainability. “Corrections” tend to be punctuated, not gradual.
11. September 2009 at 09:16
Scott,
If someone tells me “buy X, because it’s soared 75% in the last two years,” I really don’t know what to say. That seems like a crazy reason to buy something. I’d be more likely to run out and sell it short.
Actually either side of that trade is exceptionally dangerous. If the proverbial “everyone” is buying based on the rear view mirror, then they are irrational. If the market is irrational, then it can get more irrational, i.e. prices can get more out of whack with fundamentals. Shorting such a market is a an easy way to go broke.
At the same time, most definitely not everyone who bought on the way up could get out in time to save them from losses. Some of them suffered horrible losses when the tech bubble burst in 2000-2002, buying more on the way down “because it is cheap”, or so it seemed, based on relatively recent price history.
But if you can’t safely play a bubble from either side, what do you do? You get on the sidelines. Hence there’s less money willing to short the mispriced the assets than otherwise would have been the case. This in turn means the market in question is composed not of rational players acting on dispersed knowledge about fundamentals leading to something that looks like EMH, but is instead composed of irrational players buying based on the previous price, thus giving prositive feedback to the bubble, with little or nothing to pull them back to a rational equilibrium.
Thus bubbles either bust when:
A) there are no more irrational folks with money they are willing to commit to the bubble, hence no more buyers at the previous price, leading to a falling price.
or
B) Enough new supply of assets (e.g. IPO’s or newly built homes) come to market which leads to more sellers than buyers at the previous price, leading to falling prices.
C) When they are “violently” popped by something or someone finally getting through to “them” that the emperor has no clothes on. This then turns into a stampede for the exits which only a few can get out “in time”.
Anyway, that’s my take on “bubble theory”.
P.S. I’m not so sure how inspired that was, but it took more time to write than I am willing to recommit if it disappears into the ether, so I just used Control A (select all), Control C (copy highlighted text). 🙂
11. September 2009 at 09:49
I haven’t, I plead guilty to biased reading. I’ll read them in the future though.
11. September 2009 at 10:49
I want to reiterate that using the inept results of managed mutual funds to support the EMH is a huge mistake, because mutual funds are in the marketing business not the speculation business. Evidence of this is the window dressing phenomenon as well as the fact that mutual fund companies open way more funds than they should if they themselves believed they were in the active management business.
11. September 2009 at 11:08
it seems rarely mentioned that the emh does not imply markets are rational, only that they exhibit a random walk. Who is to say that this random walk is rational? Maybe it is just… random.
11. September 2009 at 13:18
Scott,
you wrote,
“I saw a recent academic paper arguing that stocks weren’t even overvalued in 1929. It seemed just as convincing to me as Shiller’s argument. But if what you say here is true, then stocks would have also been massively overvalued in late 1929 and early 1930. Hence that could not have provided any reason for the sharp plunge in AD in late 1929 and early 1930. I am convinced that if you wrote down a mathemetical model linking stock prices and AD, you couldn’t come close to fitting both 1929 and 1987, the differences simply aren’t big enough. Plus any change in AD should be related to the change in wealth, not the fact of whether stocks were “overvalued” according to Shiller’s formula.”
It would be interesting to read paper about 1929 that you mentioned. I hope that the theory that 1929 crash has caused AD shock was completely discredited by Friedman, and I hope that the EMH macro debate is all about the real supply side, correct me if I am wrong. And overvaluation according to Shiller’s formula might be relevant for the supply side of the economy.
you said,
“And of course the 1929 was not some exogenous shock, but rather reflected the stock market’s rational response to the US sliding into a deep slump.”
Well, stockmarket performance in 1929-1932 is mostly the result of bad monetary policy, but a part of it is the result of overvaluation.
you said,
”
I don’t see any evidence for this view. Real GDP growth tends to be faster during asset price bubbles, so even NGDP targeting would mean countercyclical movements in inflation. In any case, one step at a time. Even NGDP targeting would have called for much tighter policy during the asset bubbles of 1999-2000 and 2004-06, than what actually occurred. Let’s see how that works first, before taking the next step.
”
For public choice reasons (as I don’t know any case in history where the authorities were systematically stopping bubbles) I think that NGDP targeting is the best we can get in terms of leaning against asset bubbles.
You said,
“I don’t see how a bubble can create a long term real problem. In my view the “real” problem is Japanese economic policy, which is too pro-business, or pro-special intrerest.”
Isn’t it true that pro-special interest policy in Japan is a direct result of the bubble?
“You misunderstood my comment about electric cars. I agree that electric cars won’t be a big factor for a few years, and hence I think the $147 price was fully rational.
If the recession had not occurred, then I think prices would still be around $147. Why not? Consumers were willingly paying that much and auto sales in the 3rd world were booming. China recently overtook the US as the world’s biggest car market, and will be twice the size of the US car market before too long. That’s a lot of demand for gasoline. I don’t see any problem with $147 oil.”
Yes, if there was no recession, I also think that oil would still be around $147. The problem is that $147 oil last summer was consistent with zero percent probability of a recession in 2009-2011. I think that last summer you could find zero percent probability of recession estimate only in the bubble markets. About those booming auto sales in the 3rd world – I think it is more correct to use forward looking estimates, for example stock prices of automakers last summer were pricing in quite a big slump of global auto sales.
The main problem with EMH is the credit markets. I see no support for EMH neither in theory nor in practice. EMH view is that subprime CDOs should not be a problem, because if their price is too high rational speculators would short them and drive their prices down. This did not happen. Even after house prices started deflating, there was quite a long period of time when subprime securites were trading at par. When short sellers started to appear, Lehman and Bear Stearns were very happy, because they suddenly had new raw material to package into syntethic CDOs. Short sellers were no match for European banks and pension funds who were ready to purchase unlimited amounts of synthetic CDOs.
12. September 2009 at 04:48
I leave for American tomorrow, so it will be a while before I catch up here. But I will.
13. September 2009 at 08:37
John Cochrane’s response to Krugman’s article.
http://faculty.chicagobooth.edu/john.cochrane/research/Papers/#news
14. September 2009 at 07:05
Mike, You said;
“The demand is coming from somewhere. , and anecdotally, at least, I’ve heard plenty of people say that bought certain stocks or even houses because of the price appreciations in recent history.
You might think the market is in a random walk and that buying into something over-valued is ludicrous, but the vast majority of people don’t approach your sophistication. You probably wouldn’t buy a house with a teaser rate mortgage, speculating that increased future income and equity gains would help you pay future higher rates, but millions of people did.”
First, I am not a sophisticated investor. I have no more ability to predict asset prices than the next guy. Second, my point was in response to your claim that buying assets during a bubble was a good investment strategy. I don’t at all deny that some people like to invest in assets that have recently appreciated. I’ve met them. I just think it’s a silly reason.
happyjuggler0; You said;
“But if you can’t safely play a bubble from either side, what do you do? You get on the sidelines. Hence there’s less money willing to short the mispriced the assets than otherwise would have been the case. This in turn means the market in question is composed not of rational players acting on dispersed knowledge about fundamentals leading to something that looks like EMH, but is instead composed of irrational players buying based on the previous price, thus giving prositive feedback to the bubble, with little or nothing to pull them back to a rational equilibrium.”
But if this were true then smart investors would sell short. In the real world they might not sell short, but that’s because I doubt that in the real world you ever reach a point where only fools hold stocks.
rob, There is little doubt that stocks respond at least somewhat to fundamentals. So asset prices are not completely random. There could be a random component, as you say.
If the mutual funds are not the “experts,” then who is? I used to read how the Harvard endowment was able to beat the market–until it got killed in the past 12 months.
Mutual funds market themselves on their rate of return, so I’d think they’d have an incentive to try to beat the market.
123, You said;
“Isn’t it true that pro-special interest policy in Japan is a direct result of the bubble?:
I think it pre-dated the bubble.
The EMH doesn’t imply that smart people can’t make mistakes. Obviously the big banks made some foolish investments in CDOs. No one would deny that. But how do we know markets were inefficient? Maybe they just made a mistake.
PKJ, Thanks, I enjoyed that article. The title was similar to my post.
14. September 2009 at 13:07
Special interest policy in Japan predated the bubble, but after the crash nothing else was left, and fiscal stimulus lobby spread like a cancer.
Scott, You wrote:
“The EMH doesn’t imply that smart people can’t make mistakes.”
Weak version of EMH says that stocks follow random path, and mistakes are just a random noise. Well, then we can say that Soviet agriculture was efficient, because harvest tonnage followed a random path, with mistakes of smart people in one collective farm offseting mistakes of another.
Semi-strong version of EMH is more interesting for macro – it says that markets quickly incorporate all publicly available information. But look, last summer most markets were telling you that there is a significant probability of a recession, except the crude oil futures market that forgot to incroporate this information.
There is also strong version of EMH – markets incorporate all public and private information. But during whole 2006 smart funds were shorting subprime assets, but they were no match for the whole banking system that was happily buying CDOs.
Scott, which version of EMH do you find most useful?
14. September 2009 at 13:49
Scott, you said:
“If the mutual funds are not the “experts,” then who is? I used to read how the Harvard endowment was able to beat the market-until it got killed in the past 12 months.
Mutual funds market themselves on their rate of return, so I’d think they’d have an incentive to try to beat the market.”
I’d say GS is full of experts, much much more than, say, Fidelity. Mutual fund companies use survivorship bias to their advantage, launching enough funds so that some of them are likely to look like stars. That may sound like good evidence itself in favor of the EMH, but all I think it shows is that marketing mutual funds is easier than beating the market. I agree they have an incentive to beat the market, but the easiest way to get one fund to beat the market is to launch five (Why take a risk?). Keep in mind also that most (if not all) mutual funds are “long only”, limiting their speculative powers by half. With this business model, mutual fund managers probably obtain their positions mainly on the basis of their resume, charisma and corporate-political smarts than on their sophistication as speculators. They just need to look the part. GS, on the other hand, probably has a lot more autistic-types sitting at computer screens, getting bonuses based on returns not funds-under-management.
I’d also say hedge funds in which the owners have a large share of their money in their own funds are probably the real “experts”.
Yes, I’m calling mutual fund managers phony “experts”, barely a step above a guy giving a seminar at your local hotel this weekend about how to beat the market with his flashy-looking software.
14. September 2009 at 14:04
What we’re discussing here is how the EMH relates to macro. I think the most important point about that relationship is in prediction. Economists shouldn’t endeavour to create models that predict asset prices better than the market.
The essential point here is this: EMH critics, where are your Yachts? If you’re actually right about this then why aren’t you writing papers from it from your yachts between martinis?
We can look at the things you mention in retrospect. I think it’s useful to note that the banks are backed by the fed and FDIC. That is to look for hidden forms of insurance.
But I think that neither of these errors you mention say anything very useful about economics.
Now, just because we can’t predict a path doesn’t mean it will be random in any particular way. Since there are important things that are unknown to market actors as well as us I don’t think we can predict that prices will follow some particular distribution.
However, just because prices aren’t a gaussian doesn’t mean that the important aspect of the EMH is wrong. Does anyone really believe these schemes that predict some other distribution and promise to make money from it?
The point isn’t that the noise-like nature of markets shows the EMH is true. Rather counter-factual reasoning shows that its reasonable.
14. September 2009 at 14:10
Does anyone read the cartoon Alex? It often talks about this sort of thing…
http://www.telegraph.co.uk/finance/alex/?cartoon=6171626&cc=6119770
14. September 2009 at 23:38
I personally think it’s naive to think that the banks and other institutions that packaged, bought, and sold junk mortgages and their derivatives was largely based upon mistakes. Almost all of the incentives were in place to motivate the intentional bad lending, regardless of the longer term consequences.
15. September 2009 at 00:49
Mike, the problem is with the “bought” bit. Certainly those who have junk have an incentive to sell it. The question is: why should anyone have an incentive to buy it?
The most reasonable answer here is asymmetric information. The Alex cartoon I mention leans on that heavily to make jokes. So, the bank management had little incentive to be good agents for the bank owners.
The problem with this idea is that the same setup of ownership has ensured for many years. If bank managers could rip-off bank shareholders by getting big bonuses from buying toxic asset then why didn’t this happen decades ago?
Bank shareholders are not idiots either, they read Alex and laugh at it, just like others in finance. If you buy shares in an investment bank you know that it will be full of devious scheming little gits, just like most public companies.
15. September 2009 at 13:00
Current,
I agree that asymmetric information would seem to be a large factor and that the moral hazard argument is weak at bst, however I take issue with the idea that shareholders are not idiots idiots. A fair proportion certainly are. How bright were those holding onto Bear Sterns or Lehman stocks? Or, how about the vote to keep Ken Lewis running Bank of America? And is it bright to invest in companies with such opaque balance sheets in some respects, at least when the bubble bursts?
At least once a generation or so there are failures to generalize lessons learned from previous economic downturns, just as every generation since the Korean war have been fooled into supporting wars that offer most Americans no net benefit.
15. September 2009 at 14:22
Perhaps I ought to rephrase that. I don’t think that bank stockholders are, in general, any stupider than the owners of any other asset.
The comparison to voters doesn’t stand up to scrutiny. Investors lose money in poor investments in direct correlation to the amount they put into those poor investments. Voters do not get fewer votes if they used them badly in the past.
If it was so obvious that the bank shareholders were idiots then why didn’t you short bank shares and hedge against the market? In other words: Where is your Yacht?
15. September 2009 at 16:37
Scott,
You should post a link to your Cato Unbound lead essay. I’m sure your regular readers would want to see it, as would those who stumble their way here. Also I suspect everyone would love to see the give and take with the response essays.
16. September 2009 at 02:52
There are also three replies by James D. Hamilton, George Selgin and Jeffrey Hummel. I don’t think we can see them until the next issue of Cato Unbound is out though.
16. September 2009 at 03:56
123, You said;
“Weak version of EMH says that stocks follow random path, and mistakes are just a random noise. Well, then we can say that Soviet agriculture was efficient, because harvest tonnage followed a random path, with mistakes of smart people in one collective farm offsetting mistakes of another.”
This is a non sequitor. The EMH says that, ex post, people can make mistakes, but ex ante, the mistakes were not obvious. The EMH does not predict that people have perfect foresight. Since they lack perfect foresight, they will make mistakes, ex post. This definition is beyond dispute–every economics and finance textbook will tell you that rational expectations doesn’t imply perfect foresight.
In contrast, Soviet agriculture was known to be inefficient (by Western economists at least) even ex ante.
You said;
“Semi-strong version of EMH is more interesting for macro – it says that markets quickly incorporate all publicly available information. But look, last summer most markets were telling you that there is a significant probability of a recession, except the crude oil futures market that forgot to incorporate this information.”
Oil trades in world markets, and the developing countries were booming in 2008. In addition, the US stock market did not expect a deep recession in mid-2008. The US stock crash (which discounted a steep recession) occurred after oil prices had already fallen sharply. Stocks crashed between September 2008 and March 2009.
You said;
“There is also strong version of EMH – markets incorporate all public and private information. But during whole 2006 smart funds were shorting subprime assets, but they were no match for the whole banking system that was happily buying CDOs.”
The point is that no one knows which funds are “smart funds” ex ante. We assign the label “smart” to those funds that, ex post, got lucky. The Harvard endowment used to be viewed as smarter than the average investor. Yet even I beat them over the past 12 months.
I believe market prices incorporate all public info, and some price info.
rob, So mutual funds are not the experts, but investment banks are? Didn’t Bear Stearns and Lehman just go bankrupt? And didn’t other investment banks suffer huge losses? I don’t think the hedge funds saw the meltdown coming either. GS has done better than average, but doesn’t someone have to be above average by definition? BTW, I have never denied that the EMH is only approximately true. I think GS probably really is a bit smarter that its rivals. My point is that the EMH is close enough to being true for purposes of macro modeling.
Current; You said;
“What we’re discussing here is how the EMH relates to macro. I think the most important point about that relationship is in prediction. Economists shouldn’t endeavour to create models that predict asset prices better than the market.”
Bingo. That is exactly my point. Well put.
Mike, You said;
“I personally think it’s naive to think that the banks and other institutions that packaged, bought, and sold junk mortgages and their derivatives was largely based upon mistakes. Almost all of the incentives were in place to motivate the intentional bad lending, regardless of the longer term consequences.”
Just the opposite. The fact that banks bought back huge amounts of their own junk clearly shows that they did not think it was junk, they thought the bonds were good investments.
Happyjuggler0 and Current; Yes, I will put up the link.
16. September 2009 at 12:24
Dr. Sumner,
“Just the opposite. The fact that banks bought back huge amounts of their own junk clearly shows that they did not think it was junk, they thought the bonds were good investments.”
It is not about the banks, but those who stand to gain from bad loans through bonuses, etc. Why should we think they care a thing about the long-term interests of theinstitutions they run?
16. September 2009 at 12:33
I think this EMH is getting silly…
In my opinion some strong versions of the EMH are wrong, others are built on faulty reasoning. I’ll write about that another day. However, I think that the criticism brought up against the EMH in this thread are incorrect.
The core of the (weak form) EMH is that economic theories should not predict short-run variations in prices. Any theory that does so is supposing that an economist can outperform a speculator at the speculator’s job. This is extremely unlikeley. Especially given that the speculator may have access to the same theories and indeed may be an economist. Now, there seem to be a number of misconceptions here…
Firstly, the theory doesn’t tell us that the prices of assets will follow some particular distribution. Stronger forms of the EMH do, but that’s not what we’re talking about here.
Secondly, if this theory is true it doesn’t mean that large, simultaneous falls in the price of many assets cannot occur. What all of these theories concern is known information. It is perfectly reasonable that important information should be unknown. As Scott never tires of telling us future cerntral bank policy isn’t really known. Austrian business cycle theory points to significant unknowns that affect both specific prices and macro-aggreagtes. Depending on the “flavour” of ABCT it is often not possible to gather the information that could tell you if you were in an unsustainable boom. The same is true of external shocks, if we could predict them then they wouldn’t be shocks.
Thirdly, particular asset price indices do not tell us anything direct about macro aggregates. As Warren Buffett often points out there have been long periods in history when there has been economic growth without an increase in stock prices. Income theory tells us that this is entirely expected. Interest is what you get for waiting. Profit above that interest is what you get for investing well. If companies listed on the stock market don’t invest well then their prices may drop during periods of growth. The FTSE100 first reach the 5000 mark, where it is now, in 1998. There are many other capital investments outside of the indices that may do well and so economic growth may continue. In some countries the main stock index isn’t at all representative, in Ireland for example, it is >90% bank stocks.
Fourthly, we should give up this theory even if it has some problems. It is not good scientific practice to entirely abandon a theory on the basis of various small and uncertain pieces of evidence against it, such as someone claim that “of course *I* could have predicted *that*”
Lastly, momentum theory isn’t a good argument against EMH. Let’s say stocks have “momentum”, but we agree that they cannot remain above levels determined by fundamentals. So, the momentum must stop, and a lot of people who have bought too late must lose. When? Momentum theory doesn’t tell us. The whole idea is that there is a “smart time to sell”, but that is obvious. Since this sort of thing makes such little sense why does anyone think it’s right?
16. September 2009 at 12:47
I should add that if the management of even one major lending institution feels pressure to boost short-term performance and shareholder value, and thus starts offering/buying bad loans, every other major lending institution may have pressure to do likewise to avoid losing market share in the short-term.
16. September 2009 at 12:51
Current, clearly EMH is true to a degree. I just don’t by strong EMH models for reasons I’ve listed.
Don’t underestimate the challenges human brains face in dealing with very complex questions with well-documented biases in thought. Also, don’t miss the point about the net incentives to create bubbles. Weaker EMH theories can certainly allow for these.
16. September 2009 at 14:56
Mike,
Certainly there are principle-agent problems. However, there are always agency problems, everywhere at all times. Adam Smith talked about them, they’re as ubiquitous as human greed itself.
I work in the electronics industry. In that industry occasionally one companies management get desperate and start undercutting everyone else regardless of the long term effects on their company. Does that mean that the whole industry collapses, certainly not. That only happens in finance, that’s why I don’t think failure of EMH is a good explanation.
I don’t know what you mean by “strong” and “weak” EMH theories. Part of the problem with this whole debate is that the strong/weak split is not clear.
Anyway, I disagree with you about biases, at least for financial markets. If biases played any significant role then the richest men on earth would be Tversky, Kahneman and the other behavioural economists. Or those who applied their ideas to finance. “Nudge” would have been written from a Yacht. The idea of a market having biases doesn’t follow directly from individual humans having biases.
What is your argument for the “incentive to create bubbles”?
16. September 2009 at 18:16
Current, if you read my posts further up, you’ll see why I think bubbles are inevitable. They are the result of confluent interests, even from the rational actor perspective.
Management gets increased compensation, share and bondholders get appreciation, consumers get a better job market and more credit, producers and sellers of goods and services get more revenue and higher profits, politicians get credit for low unemployment and igher economic growth(and more campaign donations),
That is what I mean. Bubbles are an inevitable feature of under-regulated capital markets.
16. September 2009 at 18:18
A relatively weaker EMH would involve less efficiency, to state the obvious.
17. September 2009 at 12:46
Mike, let’s unpick what you have said.
Firstly, let’s be clear. Those who make bad loans stand to lose unless they are protected by deposit insurance. Those who borrow may also stand to lose, similarly, unless they are protected by the state. In the short run managers may get increased compensation for making more loans. This may then have a knock-on effect for the rest of the economy in the short-run.
For a time until the toxicity is found everybody wins except the creditors. However, this doesn’t mean that “everybody else wins” over a longer term. Because, just as other agents benefit from the investment spending, so other agents lose because of the loss of what the savers would have spent.
It is implausible though to label this by itself as a cause of the business cycle. Consider rents for example, huge amounts of assets are rented out, houses, land, machinery. The same thing is possible here as it is with loans. There may be hidden problems in the contract. It’s similarly possible with almost any sort of service.
Given the above how would regulation of capital markets prevent the problem? Think about your car, you don’t know how long it will last until it breaks down unrepairably. Clearly people can make buying decisions that span time. Also, clearly there are always salesmen and agents who will say anything for a sale. That goes back to the animal kingdom. We can live with that, we are used to it.
The difficulty comes though if there is bailout involved, as there is when institutions are implicitly protected or when savers are protected. The first act of “regulation” should be to ensure that these parties are never bailed out, in order to make them act responsibly.
It seem to me though that even this is unlikely to cause a major business cycle episode. The amount of state-provided cover has expanded over the years. But, the major sort, deposit-insurance, is very old. I don’t think it can be to blame. If it were that dangerous the danger would have showed up earlier. It may have exacerbated something else though.
The major flaw with your idea is that you don’t see that for business cycle theory we must identify something that changes. Principle-agent problems, assymetric information, problems of long running contract and so on are all permanent. You need more explanation to show why they lead to a crisis at a particular time.
17. September 2009 at 13:03
Scott, You said;
“This is a non sequitor. The EMH says that, ex post, people can make mistakes, but ex ante, the mistakes were not obvious. The EMH does not predict that people have perfect foresight. Since they lack perfect foresight, they will make mistakes, ex post. This definition is beyond dispute-every economics and finance textbook will tell you that rational expectations doesn’t imply perfect foresight.
In contrast, Soviet agriculture was known to be inefficient (by Western economists at least) even ex ante.”
EMH is not about the perfect foresight, but about the absence of systemic biases. Moscow university used to say that Soviet agriculture is the most efficient agriculture possible (and mistakes in Soviet agriculture are only ex-post non-obvious ones), Chicago pointed out that soviet agriculture faced huge agency issues. Now Chicago is saying that credit markets dominated by smart bankers are efficient, but Moscow university is saying that banks operate under severe agency and moral hazard pressures.
You said;
“Oil trades in world markets, and the developing countries were booming in 2008. In addition, the US stock market did not expect a deep recession in mid-2008. The US stock crash (which discounted a steep recession) occurred after oil prices had already fallen sharply. Stocks crashed between September 2008 and March 2009.”
What has actually happened in September 2008 and later is irrelevant for the accuracy of what markets were forecasting last summer, as we should look to ex-ante information. Last summer global leading indicators were showing a heightened probability (not certainty) of a global recession, and almost all global markets were were consistent with that, except crude oil futures market that was saying that the probability of a global recession is zero. In January 2008 crude oil futures and shares of oil companies were predicting the same thing, but last summer they were pointing to completely different directions. Last summer either oil futures market was efficient, or energy company stock market was efficient, but not both of them together.
You said;
“The point is that no one knows which funds are “smart funds” ex ante. We assign the label “smart” to those funds that, ex post, got lucky. The Harvard endowment used to be viewed as smarter than the average investor. Yet even I beat them over the past 12 months.”
No, we assign the label “smart” to those funds that found
errors of logic in the prevailing pricing of subprime CDOs, and we assign the label “dumb” to too-big-to-fail banks that had weak incentives to search for such errors.
Current; You said;
“What we’re discussing here is how the EMH relates to macro. I think the most important point about that relationship is in prediction. Economists shouldn’t endeavour to create models that predict asset prices better than the market.”
No, it is important to study the degree of market efficiency for macro purposes. See the sad story of Iceland.
You said,
“The essential point here is this: EMH critics, where are your Yachts? If you’re actually right about this then why aren’t you writing papers from it from your yachts between martinis?”
Well, bankers who were buying subprime trash like to spend their time in yachts drinking martinis, and they can continue their habit courtesy of TARP.
17. September 2009 at 13:15
I hate to say it but I liked what DeLong wrote about EMH in his blog:
http://delong.typepad.com/sdj/2009/09/more-musings-on-the-total-intellectual-train-wreck-that-is-todays-chicago-school.html:
I actually don’t mind that people spend their time developing real business cycle models or investigating the consequences of the efficient market hypothesis. I do, however, wish that they wouldn’t write things like:
“It is true and very well documented that asset prices move more than reasonable expectations of future cashflows. This might be because people are prey to bursts of irrational optimism and pessimism. It might also be because people’s willingness to take on risk varies over time, and is sharply lower in bad economic times. As Gene Fama pointed out in 1972, these are observationally equivalent explanations at the superficial level of staring at prices and writing magazine articles and opeds…”
because such things are simply not true.
They are not observationally equivalent.
When major rises in stock prices–like that of 1995-2000–are due to increases in market risk tolerance or decreases in risk, then at their end investors and commentators tend to think that future returns are likely to be low by historical benchmarks, but that stocks are nevertheless worth holding because (a) their risks are small or (b) we have lowered the covariance between stock returns and our consumption. When major rises in stock prices are due to a wave of optimism, then at their end we find that investors and commentators are thinking that future returns are likely to be high by historical standards–Dow 36000 anyone?–but that risks are not especially low.
And to that I object, because they would be better finance economists if they did not write things that weren’t true.
17. September 2009 at 17:44
Mike, You said;
“It is not about the banks, but those who stand to gain from bad loans through bonuses, etc. Why should we think they care a thing about the long-term interests of theinstitutions they run?”
Lots of banks were very well run. In the future investors may be inclined to put their money in banks that don’t let the loan officers run wild. Once burnt twice shy. People learn form their mistakes. If the banks are making mistakes that hurt bank profits, it’s not a systemic problem.
Current, i agree.
Mike, you said;
“I should add that if the management of even one major lending institution feels pressure to boost short-term performance and shareholder value, and thus starts offering/buying bad loans, every other major lending institution may have pressure to do likewise to avoid losing market share in the short-term.”
Firms don’t care about short term profits, they care about stock value. And stock values are determined by long run expected profits, not next quarter’s profits.
You said;
“Don’t underestimate the challenges human brains face in dealing with very complex questions with well-documented biases in thought. Also, don’t miss the point about the net incentives to create bubbles. Weaker EMH theories can certainly allow for these.”
I agree. I believe that brains are programmed to see patterns where they don’t really exist. Indeed I have argued that this bias causes people to want to reject the EMH. They see patterns, and think the stock market is more predictable than it really is. This is ironic. The one thing that makes the EMH less than perfectly true, also biases people into thinking it is even less true than it really is.
123, I’m not sure I follow your analogy. Both Soviet agriculture and American banking have problems. Neither is free market. I’d say Soviet agriculture is even worse. How does this affect the EMH?
You said;
“No, we assign the label “smart” to those funds that found
errors of logic in the prevailing pricing of subprime CDOs, and we assign the label “dumb” to too-big-to-fail banks that had weak incentives to search for such errors.”
“logic” is a strong term. Were there actually logical errors? Or forecast errors” Would the CDOs still have been mis-priced if housing hadn’t crashed?
I still am not convinced about oil–weren’t most of the other commodities also high in mid-2008?
I don’t get Delong’s point. We can’t directly observe why people are optimistic. Just because someone writes a Dow 36,000 op ed, doesn’t mean most investors look at things that way. Is there some unambiguous market indicator that shows what Fama claims we can’t know, but Delong claims we can? And if so, what is it?
Would the 30 year TIPS yield work? That is the long run real interest rate. But perhaps Fama would respond that it isn’t the rate that is relevant for discounting risky stocks
18. September 2009 at 09:18
Dr. Sumner,
You replied:
“Lots of banks were very well run. In the future investors may be inclined to put their money in banks that don’t let the loan officers run wild. Once burnt twice shy. People learn form their mistakes. If the banks are making mistakes that hurt bank profits, it’s not a systemic problem.”
While most banks may have been very well run, the largest banks representing most of the lending supply were not. And, if investors learn enough relevant lessons from banks failures and financial crises, then wouldn’t one think they’d be less common?
“Firms don’t care about short term profits, they care about stock value. And stock values are determined by long run expected profits, not next quarter’s profits.”
Is this really true, given the structured incentives enjoyed by management and commissioned employees at major financial firms? Sheila Bair has pointed out the warped incentives that stock options often represent.
18. September 2009 at 09:46
What does Iceland have to do with anything?
Yes, but where is your Yacht? You seem to think that the EMH and agency problems have a relationship which they don’t have.
What you are saying would not be unreasonable if you were talking about the market for reserves. There are only a relatively small number of players in that market, and many agency issues.
But they are silly when talking about things like oil prices and bank stocks. We can all speculate in those. So, if all this stuff was obvious then where is your Yacht?
18. September 2009 at 09:56
I know how to program computers. When I first read about behavioural economics I thought: “I’m going to write a computer program to use the behavioural biases to find arbitrage opportunities”. A little later I thought “I bet someone has already done that”. That has probably been going on in a non-systematic way for centuries by some traders discovering the effects empirically.
Centralized asset markets aren’t like restaurants or shops. There are huge benefits for being able to understand them. Although biases may play a large part in the investments of many novices, I think that professionals have more complex ideas about how investment should be done.
18. September 2009 at 10:09
In a sense, I think you’re both wrong.
That all depends on what you mean by “firms”. Employees and boards of directors have all sorts of odd incentives. However, stock investors know this, it isn’t secret, they account for it in the price.
18. September 2009 at 15:05
Current,
You stated:
In a sense, I think you’re both wrong.
“That all depends on what you mean by “firms”. Employees and boards of directors have all sorts of odd incentives. However, stock investors know this, it isn’t secret, they account for it in the price.”
If the distorting incentives are priced in by investors, then why were so many caught unprepared by the financial crisis? Why would so much money have flowed into these companies in the first place?
I think my argument makes much more sense. There was much money to be made on the way up and the dumb money was caught with its pants down. And there was a lot of dumb money.
18. September 2009 at 15:09
Current,
I am also reminded that there was a lack of transparency with regard to mortgage derivatives meant that valuation on the basis of fundamentals was impossible, making it clear that invstors had other motives. It was fruitful speculation on the way up for many, and disaster for those caught in too late.
19. September 2009 at 07:58
Mike, I don’t think banking crises occur that often. yes, we have had two in the past 20 years, but before that it’s hard to find any examples going many decades. Even the crisis in the Depression wasn’t the banks fault, it was the Fed’s.
If you don’t think investors care about long run profits, look at the biotech industry–last time I looked most companies had no profits at all.
Current, I agree with you on efficient markets. Regarding long term profits I was talking about investors. I agree that there are some agency problems within the firm. One benefit of tying bonuses to stock performance is that it leads managers to take a more long term perspective.
19. September 2009 at 08:17
I heard a proposal earlier this week to pay those responsible for creating mortgage-backed securities and related derivatives, or indeed any new innovative finance product, be compensated in terms of those products.
20. September 2009 at 05:36
Dr. Sumner,
You replied:
“If you don’t think investors care about long run profits, look at the biotech industry-last time I looked most companies had no profits at all.”
This is not incompatible with my perspective at all. We also saw plenty of speculation on internet stocks that were never profitable during the tech bubble. Yet, plenty of people made money during that bubble, even on stocks that later went south.
Anecdotally, me and the people I know are usually in positions for weeks or months, not years. And we beat the market regularly, except for the meltdown. I was actually in some I-bonds at the time, in addition to increased cash and commodities, but my losses weren’t too great as of now. Still, I was concerned about inflation when the greatest deflationary period since the Great Depression started. So, macro isn’t my strong suit.
20. September 2009 at 05:39
I would add that during bubbles, the fundamentals of companies can change quite rapidly and if one uses a screener and finds better values, then those better values replace the least undervalued in the portfolio. There is nothing unnatural or inherently wrong with this strategy as I see it, and even just after the meltdown, I was still up 300% over five years. I took a lot of long, and then some short positions, though not enough of the latter.
20. September 2009 at 11:20
Scott, you wrote:
“I’m not sure I follow your analogy. Both Soviet agriculture and American banking have problems. Neither is free market. I’d say Soviet agriculture is even worse. How does this affect the EMH?”
The belief in EMH contradicts the belief that American banking is not free market, as there are arbitrage processes operating between credit market and stock market. This all reminds me of the socialist calculation debate of the thirties, with Basel capital standards and faulty models of
credit rating agencies playing the role of Trofim Lysenko’s agricultural pseudoscience. EMH theorists are just modern day equivalents of Oscar Lange who in the 30s argued that socialist economies could use auctions a la Robin Hanson to beat capitalism. EMH says that credit markets should produce unbiased estimates of default risk and recovery risk by using public information, this is no more likely than Soviet agricultural farms producing consistently good harvests. BTW this is not an issue of socialism vs. capitalism, as any big corporation could be compared to a mini socialist dictatorship (this applies to mutual fund management companies too).
You said;
“”logic” is a strong term. Were there actually logical errors? Or forecast errors” Would the CDOs still have been mis-priced if housing hadn’t crashed?”
Housing crash is an ex-post event and is irrelevant. Even in 2005 it was hard to agree that the statement “The probability that housing prices will decline by 5% nationwide is zero” is reasonable.
You said:
“I still am not convinced about oil-weren’t most of the other commodities also high in mid-2008?”
Other industrial commodities (ex energy) peaked in April, and in summer their price incorporated some global recession risk.
You said:
“I don’t get Delong’s point. We can’t directly observe why people are optimistic. Just because someone writes a Dow 36,000 op ed, doesn’t mean most investors look at things that way. Is there some unambiguous market indicator that shows what Fama claims we can’t know, but Delong claims we can? And if so, what is it?
Would the 30 year TIPS yield work? That is the long run real interest rate. But perhaps Fama would respond that it isn’t the rate that is relevant for discounting risky stocks”
There are two approaches. Survey data (including data collected by Shiller here http://icf.som.yale.edu/Confidence.Index/) and pension accounting assumptions used by pension trustees is one approach. Long term interest rates are another source of information but they must be combined with an estimate of stock market risk premium.
21. September 2009 at 02:17
I don’t think you get it. The EMH is about information that is widely available.
Do you think that you or I would be able to make an investment strategy by recognizing distorted incentives? Let’s say one of us had sat down in 2004 with a pile of company reports and picked out the companies in the S&P 500 with the worst setup incentive schemes. We had then invested in the rest and shorted the bad ones. Do you think that such an investment strategy would produce a better return than the S&P 500? I seriously doubt it because other investors are just as aware of these problems as we are.
Even if such a strategy did work for a limited time do you think it would work for successive decades? Of course it wouldn’t because once one investor makes a fortune by practicing it it becomes common knowledge as a strategy. This is the problem we deal with when discussing the EMH as it refers to economics. Since in economics we are concerned with embedded assumptions in economic theories that either concur with the EMH or don’t.
I agree that if you or I knew something private about incentives. For example if we knew that a scheme that is commonly supposed to be good is bad in practice, then in that case there is money to be made.
Well why didn’t you make a fortune from it if it was so easy? Where’s your Yacht?
What on earth is “dumb money” anyway?
21. September 2009 at 02:46
Current,
You stated:
“Even if such a strategy did work for a limited time do you think it would work for successive decades? Of course it wouldn’t because once one investor makes a fortune by practicing it it becomes common knowledge as a strategy. This is the problem we deal with when discussing the EMH as it refers to economics. Since in economics we are concerned with embedded assumptions in economic theories that either concur with the EMH or don’t.”
Where is the evidence that successful investment strategies are common? In fact, doesn’t EMH assume that consistently beating the market in the long run isn’t possible, or is at least extremely unlikely?
I’ve heard advocates of strong EMH say that people like Buffett or Soros were just very lucky to a degree, and in any case, had more influence on the management of firms than most investors(Of course, neither had this influence when first creating their wealth). However, given that their strategies have been enormoously successful, why haven’t they become standard practice, while each continues to grow his fortune substantially? If many investors believe in sufficiently strong EMH, then why would they even try to beat the markets? And, if they don’t, then how can they be said to be rational actors? I know this rationality assumption isn’t necessarily required in some versions of strong EMH, but it does bring up another inconsistency. So, there is that inconsistency in your reply, plus the one latter mentioned.
“Well why didn’t you make a fortune from it if it was so easy? Where’s your Yacht?
What on earth is “dumb money” anyway?”
Well, as I mentioned, my return over the last few years has been in the 300% range to date, even after suffering almost a 70% loss during the bottom. My intital investment was low, given my modest income, but I have more money working for me than most people my age, though not enough to comfortably retire yet. I earned my money largely by investing in long positions during the bubble, which is of course very easy to do and says little or nothing about my investmentn skills. I ignored the Black-Scholes model, by the way, and focused on fundamentals. Ironically to some, this strategy was highly successful, though you could say I was just lucky.
Dumb money would be money like mine just before the crisis last fall, which was having too many long positions and not enough short ones, with some money in I-bonds right before the most serious deflationary period since the Great Depression. I had simply grown lazy and paid little attention to macro-fundamentals in retrospect, and was quite ignorant in this area as well.
21. September 2009 at 02:47
Current,
I should have asked, where is the evidence that successful investment strategies are often copied?
21. September 2009 at 04:42
Is this some sort of highly sophisticated joke?
If not, I don’t think you understand what this sort of debate is about. It is not about the overall efficiency of a particular sort of economic order. It’s about pricing within particular markets.
Suppose we have a set of bank stocks. The banking system is very corrupt and entry to the market is mainly due to corruption. The EMH applies to stocks in these banks just as it would to a more free market banking system.
It’s not about institutional comparisons.
As I said in an earlier discussion: We could legislate so that Stout is sold on a centralized market from 9am to 5pm every weekday in Dublin. To obtain Stout a tippler phones the Stout market and puts in a bid for a quantity specifying Guinness or Murphys, etc. The Stout is then couriered to the customers the next working day when they are invoiced. Such a market allows market forces to play a role. However, it may well be better to allow the shape of the market to be determined by market forces too.
21. September 2009 at 05:36
Strong EMH doesn’t compare well with ideas such as the uncertainty principle in physics. The latter is based on a much larger collection of sound experiments, preditive validity, and much sounder logic than EMH. Many couldn’t imagine for decades the uncertainty principle could be incorrect, yet a more fundamental theory, like perhaps string theory, may indeed demonstrate it is, and bring about a sort of neoclassical model of the universe.
I’m not saying string theory is correct, by the way, as it is merely highly speculative at this point, but it is easy to imagine how it can undermine the uncertainty principle.
Strong EMH isn’t ultimately testable and hence should always be looked upon with skepticism, as is the case with any macro social theory.
21. September 2009 at 05:38
Current,
You offered:
“Is this some sort of highly sophisticated joke?
If not, I don’t think you understand what this sort of debate is about. It is not about the overall efficiency of a particular sort of economic order. It’s about pricing within particular markets.
Suppose we have a set of bank stocks. The banking system is very corrupt and entry to the market is mainly due to corruption. The EMH applies to stocks in these banks just as it would to a more free market banking system.”
Do you absolutely deny that the nature of an economic system cannot elminate some randomness in investment valutions? Does the random walk always apply?
21. September 2009 at 12:52
I didn’t say that successful investments are common.
That doesn’t imply that I think market beating investment strategies are common.
There are a few very successful investors. However, this doesn’t imply that the EMH is not true. It is quite possible that they are successful because they know things that are not available to the public. Such as particular facts about companies or individuals, or strategies that work in particular cases.
There have been many attempts to understand how these investors work. I have one of the several books published about Warren Buffett’s investment strategies.
Who is to say that these successful investors haven’t altered their strategies? I think they have. I think their earlier methods are now well known and consequently not that useful.
Isn’t this obvious, people teach each other strategies. If you don’t think this happens then why are you investing at all, since the entire market process depends upon it happening. If people didn’t learn from each other we would still live in caves.
Besides, the EMH in any reasonable form is about publically available information. How is an economist supposed to write an economic theory that deviates from the EMH and then publish it?! How can you make an economic theory that only works if no-one important knows it?
This criticism comes from the normal way the EMH is theoretically justified. I think there is a problem here, as I’ve explained elsewhere in the comments on this blog. But it would take too long to explain here.
To summarize, all actors don’t have the same information. If an actor occasionally receives information about a stock before the market he can use that to make a profit. The point is that it is not rational for those who don’t know anything special to behave as if they do.
I see. I don’t think that has much bearing on our discussion though.
No. The random walk theory doesn’t follow directly from the EMH. It isn’t certain that facts unknown to the market are necessarily random. There are numerous ways through which the unknown factors may be non-random. The idea of an “environment of increased uncertainty” is reasonable, and has obviously historically occurred.
21. September 2009 at 16:05
As for inflation targeting as a means to escape a liquidity trap, I want to point out that over the last four months inflation has run at a 1.8% per annum rate while income growth has been flat. Thus, real income has dropped.
Further, given that the unemployment rate is now near 10%, real income is unlikely to rise until demand is stimulated and the output gap is closed.
My question to you is given that consumers are the driving force of our economy (and that businesses are unlikely to invest absent a pickup of consumer demand), how can a monetary policy which is purposely attempting to drive prices higher stimulate demand where (1) real income is declining and (2) consumer balances sheets were in bad shape to begin with?
Perhaps to find equilibrium and to close the output gap and help employment a little deflation might not be such a bad thing. Also, Japan had deflation, but did it have unemployment to the extent that we are and may continue to suffer with? Did Japan’s deflation help keep people employed? Perhaps in a few years Japan’s experience may not look so bad.
Of course, deflation hurts the big banks that are leveraged, but it helps people who live week to week on a paycheck and are not leveraged. Conversely, inflation hurts these people.
Anyway, I am interested in your views as to the above.
Thank you,
25. September 2009 at 10:25
Krugman is all politics. If it supports the Dems, he’s basically for it. If it supports Conservatives, he’s against it. Economics (selectively) is just the language of his argument now.
The timing of his Nobel prize was basically just a snub to W from the international community. Maybe some of his early work is good but he has morphed into a partisan talking head. Unfortunately, now people give him more credibility to his political arguments than he deserves.
26. September 2009 at 12:29
Mike, You said;
“Anecdotally, me and the people I know are usually in positions for weeks or months, not years. And we beat the market regularly, except for the meltdown.”
I assume you and your friends have been lucky. If not, please tell me when you buy and sell, so that I can get rich too.
123, You said;
“The belief in EMH contradicts the belief that American banking is not free market, as there are arbitrage processes operating between credit market and stock market.”
Sorry, I don’t follow this. If I believe in the EMH does that mean I cannot believe a certain industry is not a free market? I must even believe that regulated utilities with no competition are a free market? I am sure I misunderstood your point, please explain again.
You said;
“Housing crash is an ex-post event and is irrelevant. Even in 2005 it was hard to agree that the statement “The probability that housing prices will decline by 5% nationwide is zero” is reasonable.”
I agree, but lots of very smart investment bankers seem to have thought it was very reasonable to think that. I have no idea what they were thinking.
You said;
“Would the 30 year TIPS yield work? That is the long run real interest rate. But perhaps Fama would respond that it isn’t the rate that is relevant for discounting risky stocks””
But how do we know what investors expect the future cash flows to look like?
Clifford; You said;
“My question to you is given that consumers are the driving force of our economy (and that businesses are unlikely to invest absent a pickup of consumer demand), how can a monetary policy which is purposely attempting to drive prices higher stimulate demand where (1) real income is declining and (2) consumer balances sheets were in bad shape to begin with?”
This is why I hate inflation, and like NGDP. Inflation causes endless confusion. First a few points: Both fiscal and monetary stimulus aims to boost NGDP. In both cases it will also raise the rate of inflation if successful. But the goal is to raise NGDP, or aggregate demand. I certainly do not favor policies that would boost prices but not real output.
Regarding the 1.8% inflation over the past 4 months, there are several problems. First, the CPI is horribly inaccurate. And second, 4 months is too short to be statistically significant. It represents merely a bump in oil prices. People should just ignore inflation, and look at NGDP. I don’t blame you for your confusion, as I do talk about inflation, but only because others do. I wish we could just talk about NGDP.
Steve, Yes, he seems partisan to me. He criticizes Bush for protectionism but not the Dems, for instance.
26. September 2009 at 13:09
Scott,
you said:
“Sorry, I don’t follow this. If I believe in the EMH does that mean I cannot believe a certain industry is not a free market? I must even believe that regulated utilities with no competition are a free market? I am sure I misunderstood your point, please explain again.”
Regulated utilities are irrelevant for EMH, as there is no arbitrage between the inefficient price you pay for your water consumption and the stock prices. Step by step:
1. Banks are not a free market, so the pricing of their output does not fully incorporate public information.
2. Output of banking system is the price of credit, so the price of credit does not fully incorporate public information.
3. There are arbitrageurs, who change equity prices and align them with the price of credit. After this equity prices no longer fully incorporate public information, and EMH no longer applies to the stockmarket.
You said;
“”Even in 2005 it was hard to agree that the statement “The probability that housing prices will decline by 5% nationwide is zero” is reasonable.”
I agree, but lots of very smart investment bankers seem to have thought it was very reasonable to think that. I have no idea what they were thinking.”
Look at this 2005 article from the Economist:
http://www.economist.com/opinion/displaystory.cfm?story_id=4079458
If this is not a public information then I don’t know what it is. The fact remains that the price of credit instruments did not incorporate the information from the Economist 2005 article until 2007.
You said;
“But how do we know what investors expect the future cash flows to look like?”
Usually we can assume that long term growth rates of the economy will be close to long term historical averages. Sometimes, for example in 1999 when singularity type thinking was very popular we have to use investor surveys.
26. September 2009 at 13:20
Scott,
you wrote:
“I agree, but lots of very smart investment bankers seem to have thought it was very reasonable to think that. I have no idea what they were thinking.”
Kling has the answer here:
“My view of securitization is that it is a 7-layer cake of moral hazard. The investment bankers get to eat the icing, and taxpayers get to clean up the mess. ”
http://econlog.econlib.org/archives/2009/09/felix_salmon_vs.html
26. September 2009 at 17:39
123,
I think you misunderstand the point of this debate. Of course the price of credit has nothing to do with the EMH. The price of credit is set by the Federal Reserve and associated agencies, though the private sector have a role too.
The point of the EMH is to illustrate the futility of economic theories that give results so precise that trading could be done against them.
27. September 2009 at 06:43
Current,
You replied:
“Isn’t this obvious, people teach each other strategies. If you don’t think this happens then why are you investing at all, since the entire market process depends upon it happening. If people didn’t learn from each other we would still live in caves.”
No, how mimicry occurs with respect to comples absract strategies isn’t obvious at all. The brain faces challenges much more daunting than many seem to realize. Just apply the counting rules in probability to see this. If you don’t think about it this way, even common three step procedures may sem simple, but the reality is that there may be many, many more possible ways to try to acheive the ultimate goal. And this is the simple case. In the case of abstract, dynamic markets with a vastly broader problem spaces, you end up with perceivalbe and actual numbers of ways to approach each step, or combinations of steps that can be in the millions, billions, or even more. The competing predictive contexts are many and successful paths few. Worse, the contexts are ever changing, leading to increased generalization and discrimination failures.
So, some tasks are easy for humans to mimick, like many motor tasks. This isn’t the case with complex, abstract perspectives in dynamic markets.
27. September 2009 at 06:47
Dr. Sumner,
You replied,
“Mike, You said;
“Anecdotally, me and the people I know are usually in positions for weeks or months, not years. And we beat the market regularly, except for the meltdown.”
I assume you and your friends have been lucky. If not, please tell me when you buy and sell, so that I can get rich too.”
I agree that I may have largely been lucky as you see it. But, I suspect I was lucky in another way. Winners seem easy to pick during a bubble. So, I am quite skeptical about my investment strategies. I haven’t done too badly this year, but last year’s losses still haven’t been made up, even though the three year return is still excellent.
27. September 2009 at 08:43
I see what you mean. You are getting now into some of the reasons why I’m suspicious of the normal justification given for the EMH. But, you haven’t really criticised the EMH.
The explanations I’ve read of the EMH refer to pieces of information. A piece of information is “publically available” or it is not. This idea is dubious to begin with there will always be a more continuous gradiation. Each market participant will then be able to make an estimation of the value of the stock.
I think that this sort of thing should be fleshed out more. I expect others have already done this, but I just haven’t read their work. In my view each participant must employ some sort of strategy or other. Even the simplest acts of trading have behind them simple strategies, even if the trader is unaware of them. So, each trader has a “background” that he draws upon. Each fact is processed through that background knowledge. That happens both knowingly and unknowingly, that is some of the background is conscious knowledge and some is unconscious. It happens both objectively and subjectively, that is sometimes people think in terms of what future prices will be numerically and associate numerical probabilities, and sometimes not.
I think that it may be the case that some stock market experts have a very good background in their work. That means that they can process the publically available knowledge better than other participants. Though this is hypothetically possible I don’t know if it actually occurs.
Although passing on information is very difficult. It is by no means impossible. Look at the computer you are using, have you any idea of the amount of knowledge that has gone into it? Despite it’s complexity though, it is unlikely that any significant part of that knowledge will be lost. There is too much financial incentive to retain it.
But, as I said earlier this sort of thing has no real implications for the EMH because. Because, as I said earlier, any reasonable version of the EMH concerns itself with what is widely known. Even if such expert background exists it is certainly not in the area of economics.
Scott Sumner or Brad DeLong or I cannot predict better than the market how specific market situations will evolve. I don’t have a Yacht and I doubt I ever will.
As Hermann Hans-Hoppe wrote about this in the context of a strict part of Austrian Economics – Praxeology – but I think his words apply more generally to economics:
27. September 2009 at 10:26
123, You said;
“1. Banks are not a free market, so the pricing of their output does not fully incorporate public information.”
But the EMH says that prices incorporate all publicly available information, so how does this contradict the EMH? I still think stock prices reflect all available information.
Obviously these very smart investment bankers did not believe what was in The Economist. It’s a shame, if that had they might not have lost most of their wealth.
Yes, the growth rate of GDP is fairly predictable, but haven’t corporate earnings been a rising share of GDP, until recently?
Kling is wrong, plenty of investment bankers (including the most important decision-makers), lost a fortune. I am pretty sure that the heads of Lehman and Bear Stearns lost money, as did most of the top people in those firms.
Mike, I don’t think the phrase “during a bubble” has any meaning. It implies that asset prices have momentum, and they don’t. They follow a random walk. It is not easy to make money during a bubble, because by the time a bubble is identified, there is a 50-50 chance prices will go up or down from that bubble position. And no one knows when a bubble will form, which would be the idea time to invest in one. If we are in a bubble, you should see, not buy.
27. September 2009 at 11:02
I think there is a difference between bank stocks and the loans banks make. My point earlier was that bank credit isn’t a free market. Obviously bank stocks are.
Perhaps the prices of bank stocks have not been so unreasonable. In Britain the price of banking stocks was in the doldrums long before the sub-prime crisis. Northern Rock had a P/E of 10 before the crisis.
27. September 2009 at 16:04
“Mike, I don’t think the phrase “during a bubble” has any meaning. It implies that asset prices have momentum, and they don’t. They follow a random walk. It is not easy to make money during a bubble, because by the time a bubble is identified, there is a 50-50 chance prices will go up or down from that bubble position. And no one knows when a bubble will form, which would be the idea time to invest in one. If we are in a bubble, you should see, not buy.”
I don’t know much about economics, so I acknowledge I can be wrong, but that statement doesn’t make sense to me. It seems so out of touch with the reality of the markets that I can’t reconcile it with reality. Maybe I’m just being close-minded and ignorant.
I don’t understand why one cannot come up with at least some rule of thumb metrics concerning the vunerability of a market, or even a whole economy to both likely and unlikely, endogenous or exogenous shocks. Even if one just went by the history of bubbles, one could have made money in the Tech bubble that ended in 2000-2001 the recent housing bubble.
The fact is, if you take the housing bubble into consideration, for example, many people did make a lot of money. Those who knew what to do with their gains are much better off than they would have been.
The housing market is less efficient than the stock market, of course, but it’s not as if the risk of default of a company in the macroeconomic context is impossible to get a rough handle on. You look at the lending institutions, overall leverage in stocks(risk increases exponentially with net debt and servicing), compare prices to discounted cash flow, and other fundamentals, etc.
The idea that no one can beat the market consistently… If a very talented medical researcher is good at spotting bargains in companiew with promising research, how can others in the market compete? That is, if the researcher is good at evaluating new proprietary medical treatments. How many of these researchers are there who actively trade in the market? Or take electrical engineers and tech stocks, etc.
27. September 2009 at 17:39
I should also add that the idea that there is a 50% chance the market will go up or down is not necessarily the view of investors and hence doesn’t necessarily have much, if anything to do with their behavior. In fact, if this were a common conception, then why would there be sustianed bubbles at all? Wouldn’t there perhaps be bounces around a lower equilibrium?
About investing during bubbles, companies have more margin for error. Consumption is up. Employment is up. Credit is loose. Sometimes wages are increasing. This gives investors more margin for error as well, as more companies have rising stock values and hence winners are easier to pick.
28. September 2009 at 06:59
Current, If UK banks stocks were in the doldrums, doesn’t that mean the crisis was partially anticipated by the market? (Obviously not fully anticipated.)
Mike, You need to look at the studies of stock market movements. Stocks follow a roughly random walk. I have other posts (Like my one on “clutch hitters”) that explain why people think there is momentum. It’s like sports fans who think athletes have hot streaks, when in fact the statistical evidence shows they do not. Just because you’ve made 4 baskets in a row, doesn’t mean you are any more likely to make the next shot than if you’d missed 4 in a row. Sports fans are convinced there are hot streaks, but they’re wrong, it’s all a random walk.
The same is true of investing. You may be convinced that stocks have momentum, but they don’t. I agree that a scientist may have inside information that could help in picking out biotech stocks. But even in that case I think their advantage is only very slight.
Mike#2, If you just flip a coin and map out plus or minus one on graph paper for each heads or tails, you will get bubble-like formations, despite the procedure being completely random.
28. September 2009 at 07:38
Yes, that was my point, perhaps some noticed problems. Or at least the banks were being treated with a healthy suspicion. One headline called Northern Rock “Big and Unpopular”.
28. September 2009 at 11:01
Scott, you wrote:
“I don’t think the phrase “during a bubble” has any meaning. It implies that asset prices have momentum, and they don’t”.
Sometimes options are priced in a way that implies momentum of the underlying asset. If you believe that asset prices never have momentum, you can try your luck in the options market.
You wrote:
“”1. Banks are not a free market, so the pricing of their output does not fully incorporate public information.”
But the EMH says that prices incorporate all publicly available information, so how does this contradict the EMH? I still think stock prices reflect all available information.”
Let’s focus on two groups of traders:
– arbitrageurs who use mispriced credit to buy equities and cause equities to deviate from EMH price
– EMH says that there should be a second group of traders who notice that equities have become mispriced (e.g. in 1989 Japan equity risk premium was clearly negative). Second group of traders short equities and bring equity prices down to their EMH levels.
Unsurprisingly first group of traders have much better access to credit than the second group, so first group prevails in setting equity prices, and EMH is violated. You can see it in the data that the second group is no match for the first one, for example private equity funds have much more firepower than short-bias equity hedge funds.
You said:
“Obviously these very smart investment bankers did not believe what was in The Economist. It’s a shame, if that had they might not have lost most of their wealth.”
Perhaps they were listening to Ben “housing prices have reached a permanently high plateau” Bernanke.
You said:
“Yes, the growth rate of GDP is fairly predictable, but haven’t corporate earnings been a rising share of GDP, until recently?”
Long term GDP share of corporate earnings is a legitimate question for debate. In June 2007 investors believed that high share of financial profits in the GDP is sustainable, in June 2008 they no longer believed that.
You said:
“Kling is wrong, plenty of investment bankers (including the most important decision-makers), lost a fortune. I am pretty sure that the heads of Lehman and Bear Stearns lost money, as did most of the top people in those firms.”
Kling is right. It is not OK to pick specific bankers (Lehman) and a specific year (2008). You should look at an aggregate compensation of industry managers over the full economic cycle. During boom years managers pocketed increasing share of GDP (via explicit FDIC subsidy and implicit Fannie-Freddie and TBTF subsidies). In 2008 most of the banks received bailouts so most managers could continue their rent seeking activities.
28. September 2009 at 11:31
Dr. Sumner,
Yes, I do need to look at the evidence, but perhaps using strong EMH (even if you support a weaker version) as the null hypothesis. I really do need evidence to make any argument here.
28. September 2009 at 11:39
I think 123 makes a good point here. Debtors, Employees, Managers and Shareholders are “stakeholders” in businesses. Only shareholders really own them. However, the others keeps a check on them by being able to withdraw their input.
The situation isn’t the same though when there is something like deposit insurance and a lender of last resort. Deposit insurance means creditors no longer have a reason to carefully vet their investments.
Banks are in a unique position. If the downside risk of their investments is high they have a simple solution. They can borrow from savers who are behind the wall of moral hazard. They can then invest this money to enhance their upside potential. The extra downside risk that borrowing creates is not important for the bank or the investors because it’s borne by the deposit insurance scheme.
29. September 2009 at 17:55
123, There may be a tiny bit of momentum, but surely not what bubble proponents like Mike are assuming. I’ve done better by assuming the reverse, I like to buy stock markets that have recently done very poorly. That also produces uneven results, but I doubt any worse than rising momentum.
I still disagree about Kling’s comment. In any case, the owners of these firms are ultimately responsible for their behavior. And the owners obviously lost a fortune, even taking account of the big profits in earlier years. It is my impression that a lot of top executives get much of their income from capital gains. Is that wrong? In any case, I don’t disagree with the implied policy advice, we absolutely should stop bailing out big banks. Monetary expansion is a much fairer and more efficient way of preventing the banking crisis from devastating the economy.
3. October 2009 at 12:25
Scott, you wrote:
“123, There may be a tiny bit of momentum, but surely not what bubble proponents like Mike are assuming.”
Usually there is a tiny bit of momentum that studies find on some time horizons, perhaps due to assymetric payoff structure. But sometimes there is huge momentum (Japan 1989, dotcoms 1999, housing 2005) and in these rare cases it might make sense to try capturing the up leg of the bubble.
” I’ve done better by assuming the reverse, I like to buy stock markets that have recently done very poorly. That also produces uneven results, but I doubt any worse than rising momentum.”
I remember your best trade was buying Asia in 1998. This trade worked so well because Asian stocks were fairly valued in 1998, and became overvalued afterwards because of underpricing of credit.
“I still disagree about Kling’s comment. In any case, the owners of these firms are ultimately responsible for their behavior. And the owners obviously lost a fortune, even taking account of the big profits in earlier years.”
Owners have unlimited upside and limited downside. Financial stocks outperformed the market from 1994 to 2006 and ex-ante expectations in 2006 indicated that financials will perform no better and no worse than whole stockmarket.
” It is my impression that a lot of top executives get much of their income from capital gains. Is that wrong? ”
It is right. Stock option grants have assymetric payoffs that encourage risk taking. Executives who diversified their portfolios did very well (Paulson sold his Goldman stake tax free before investing in blind trust).
“In any case, I don’t disagree with the implied policy advice, we absolutely should stop bailing out big banks. Monetary expansion is a much fairer and more efficient way of preventing the banking crisis from devastating the economy.”
I agree with that, assuming that lender of last resort function is preserved.
4. October 2009 at 06:38
123, You said;
“But sometimes there is huge momentum (Japan 1989, dotcoms 1999, housing 2005) and in these rare cases it might make sense to try capturing the up leg of the bubble.”
I don’t understand this claim at all. Especially for the dotcoms. Where do you get evidence that huge gains were expected in the dotcom market? If so, why didn’t arbitragers take advantage of those futures prices, as nominal interest rates were only about 6%. You could borrow money at 6%, buy dotcoms, and simultaneously sell them in the futures market.
You said;
“This trade worked so well because Asian stocks were fairly valued in 1998, and became overvalued afterwards because of underpricing of credit.”
I sold all my American stocks and went 100% into Asia in 1998. I think it is more likely they were undervalued in 1998, and fairly valued later. The HongKong market was around 8000, as I recall. Now it is over 20,000, and that’s considered a huge drop from the 31,000 in 2007. If Hong Kong is fairly valued now, it was undervalued in 1998. Japan is another story, I have never liked that market.
It’s probably not worth me debating the issue of moral hazard, as i agree with the basic point that bailouts have created a moral hazard problem, and in particular have encouraged too much risk taking. But I am no expert on banking. Perhaps more bankers have struck it rich than I assumed. I admit I was relying on press reports that suggest Wall Street is really suffering, and of course press reports may be highly unreliable.
5. October 2009 at 12:56
I have begun examining some of the literature on stock price momentum. It seems that many papers have been published concluding that the phenomenon exists. Even Fama has stated that it is the greatest threat to strong versions of EMH.
Below are links to the papers I’m examining:
http://www.e-m-h.org/FaFr88.pdf
http://www.qass.org.uk/2008/vol2_3/paper2.pdf
http://www.lse.ac.uk/collections/paulWoolleyCentre/publications/PWC1.htm
http://docs.google.com/gview?a=v&q=cache%3AFT6R0mRn7x8J%3Amba.tuck.dartmouth.edu%2Fpages%2Ffaculty%2Fjon.lewellen%2Fpublications%2FMnReversion.pdf+Temporary+movements+in+stock+prices&hl=en&gl=us&sig=AFQjCNF25TihSfOFatfXmrKAZJIptt4TGw&pli=1
And now, below is a link to an article with another paper by Fama on the subject:
http://www.voxeu.org/index.php?q=node/4052
5. October 2009 at 14:49
I think the question about momentum is how it can be used. It doesn’t really affect weak forms of the EMH, only the stronger ones.
There are several quite reasonable theories that suggest reasons that variations in stock price should have momentum.
5. October 2009 at 15:50
Mike and Current: you guys need to read Victor Niederhoffer on speculation, specifically on one of his pet concepts called “Ever Changing Cycles”. It is borrowed from horse-handicapping. It goes something like this:
Assume first that a certain strategy in handicapping horses works. Say, a familiar strategy such as “bet on the beaten favorite”. (The theory behind this strategy is that a good horse that lost his most recent race will be under-priced, because the loss will be over-accounted for.) So this strategy works for a while, but it can’t work forever because the inefficiency is bound to be picked up on and popularized by others. Eventually “bet on the beaten favorite” becomes the meme of the track. At that point the beaten favorite becomes overpriced, since it is the preferred strategy. The old best-strategy soon becomes the new worst-strategy.
Under the “ever changing cycles” theory no strategy will work for long. It is similar to EMH, with the difference that it allows for windows of opportunity. Perhaps it doesn’t violate the weaker EMH version.
According to Ever Changing Cycles, the moment buy-and-hold becomes the mantra, it no longer works, because if the masses are convinced buy-and-hold is the best idea for equities, equities will become overpriced.
Momentum strategies won’t work –except by chance — simply because they are a favorite strategy of the public. You don’t want to try this old favorite until it proves itself beaten…
6. October 2009 at 17:34
Mike, I admit that the studies are conflicting, but I am not convinced. The Fama and French article says their model works for 1926-85, but doesn’t find any significant effect for 1941-85. That means two things:
The result is very fragile, and may be due to data mining.
Or, the effect may have been real in the Great Depression, but then disappeared as markets got more efficient.
In general, I am very skeptical of these anomaly studies. The first studies showed prices were a random walk, then as they looked closer they found patterns. But if you look at any random set of numbers long enough you will find patterns. You may be right, but I think the effect is slight, and might disappear any moment. It is certainly not to be relied upon when making investments
Also, F&F did confirm one of my arguments. If you plan to hold stocks for 2 to 5 years, there is negative autocorrelation, so you should go against momentum. But I don’t even trust that finding.
Current, Maybe, But I’m not convinced.
Rob, Yes, I agree.
7. October 2009 at 13:57
Well, my point is that I don’t use momentum as a strategy, but I’ve heard many people say that so-and-so is investing in XYZ is has made a small fortune. I’ve had friends and relatives come to me wanting to mirror my portfolio, even though there’d been significant price appreciation and me having explained this.
However, my impression, and I’m still looking for papers on this, is that there is a higher number of profitable stocks during booms than under other circumstances. This makes sense from an a priori point of view, as greater perceived or real wealth will allow for the expansion of portfolios into stocks that otherwise would not yield expected profit.
8. October 2009 at 07:42
“Where do you get evidence that huge gains were expected in the dotcom market?”
Two pieces of evidence – peak of borrowed money in margin accounts, survey data.
” If so, why didn’t arbitragers take advantage of those futures prices, as nominal interest rates were only about 6%. You could borrow money at 6%, buy dotcoms, and simultaneously sell them in the futures market.”
The reason is that a combination of an asset and a short sale in a futures market is a risk free combination and earns a T-bill rate. Nobody borrows money at 6% to use it in a complicated way to earn a T-bill return.
“It’s probably not worth me debating the issue of moral hazard, as i agree with the basic point that bailouts have created a moral hazard problem, and in particular have encouraged too much risk taking. But I am no expert on banking. Perhaps more bankers have struck it rich than I assumed. I admit I was relying on press reports that suggest Wall Street is really suffering, and of course press reports may be highly unreliable.”
Not everebody has preserved his finance bubble fortune, so press reports really have substance. Indeed, one of the factors that limits bubbles is the inability to preserve the gain for everybody. For example, during dot com bubble Arnold Kling got out early enough, not everebody was so lucky. Were there any press reports about the suffering in the Sillicon Valley in 2001-3?
8. October 2009 at 08:20
I guess I’ll end my part of this discussion by providing a link to a short video in which George Soros says that he profits greatly from momentum bubbles provide. Now, one could argue that he’s just been very lucky, but he’ll have had a Hell of a run of luck.
http://bluematter.blogspot.com/2009/10/george-soros-on-m-reflexivity-and.html
8. October 2009 at 08:49
I should mention that though the above video is over 30 minutes in length, Soros’ position on bubbles and is offered beginning after the first few minutes.
8. October 2009 at 09:11
There were a couple of points forgot to add in my post above. For one, do you consider real estate markets to highly efficient. The conventional and a priori answer is no. Yet, knowledge of lending practices in the residential mortgage market was either not widely proliferated, or was irrationally ignored.
I strongly suspected the recent housing boom was a bubble, given the steep increase in home values versus previous trends and real per-capita income growth, the ratio of mortgage payments to rents, the replacement values of homes, and my brother’s input from his position as a mortgage broker. My brother told me about “liar loans,” negative amortization, and loan to home values of greater than 100%.
In fact, just about everyone I know suspected there was a housing bubble, but the magnitude of consequences for the financial sector was unknown to me.
And correct me if I’m wrong, but doesn’t strong EMH predict that fundamental analysis shouldn’t matter? Well, if this is the case, here’s a link to some charts that seem to indicate a rather strong inverse correlation between low P/E ratios and higher rates of return.
http://delong.typepad.com/sdj/2009/10/updated-campbell-shiller-regressions.html
Invert the P/E ratio and you get a simple, if incomplete rule of thumb expected return. Hence, When this inverted P/E ratio is low, it gives a quick estimate of the level of speculation in a stock, or index and hence makes spotting bubbles rather easy. Just compare the bubble data to trend and relevant macro fundamentals.
8. October 2009 at 09:13
To clarify my point above, the stock market behaved as if there was either wasn’t sufficient information within the market or that it was insufficently appreciated.
8. October 2009 at 13:35
Mike, I don’t know if stocks do better during booms. But I very much doubt the return is correlated with expected movements in output.
123, I don’t quite follow this:
“The reason is that a combination of an asset and a short sale in a futures market is a risk free combination and earns a T-bill rate.”
When you said investors expected huge gains I assumed there was a 20% or 30% premium in futures prices. If not, I don’t understand how you can say investors expected huge gains. I pay little attention to survey data as I don’t find it reliable. I am much more interested in asset prices as indicators of expectations, such as futures or forward prices.
My hunch is that if investors really expected huge gains then they would have sold all their T-bonds and poured into the stock market. But just as many T-bonds were held during the tech boom as before the boom. That makes me think the expected rates of return weren’t that different.
Mike, In a country of 300 million people wouldn’t you expect a few people to be consistently lucky? People like Soros and Buffett. And of course even Buffett lost a fortune last year.
Mike, There is no question that everyone knew about the lending practices in the real estate market. Tens of millions of people were buying houses on very easy terms, how could that be kept secret? Back in 2006 every single person I talked to knew about the zero down, no income verification mortgages. How could they not know, many of their friends were getting them?
Mike, I’ve said many times that anomaly studies showing market inefficiences are meaningless. It is easy to find anomalies looking at past data, it is much harder predicting where prices are going. I could do exactly the same given historical data from a roulette wheel.
Remember the predictions of Nouriel Roubini this past March? He said something to the effect of don’t be suckered into the stock market. And he has the best track record of anyone in the entire world during this crisis. So it shows you how hard it is to predict the market.
9. October 2009 at 01:13
I don’t agree with the implicit assumption both of you are making that Soros and Buffett know the same things as the rest of the market.
I expect they have information that is not available to others. I’m not saying that they are using insider information, only that they are in a position to learn more because of the large organizations they have behind them.
The only investor in Britain to consistently give returns of the sort Buffett gets is Anthony Bolton. He has said a number of times that he meets the CEOs of each of the companies he invests in and attempts to assess their personal integrity. This sort of information is just not available to smaller players.
11. October 2009 at 07:29
Current, I’m a bit skeptical about Buffett having that advantage. I’m not saying it has never happened, but I think his long term success can’t be explained by one or two tips.
11. October 2009 at 10:55
Buffett has an organization dedicated to gathering information about stocks. I think this can generate a stream of insights that are not widely recognized by the market.
As I’ve said previously I think Buffett (and his organization) probably also has unique background in interpreting market facts.
11. October 2009 at 13:55
Current,
I don’t buy the theory that CEO’s have any special insight into the future of their own businesses. A CEO is one individual, limited by individual biases, influenced by fads and fashion. My bet is that they are as surprised as anyone by how their stock prices perform (unless they are outright criminals cooking the books). The last company I worked for the CEO was obsessed with what the stock analysts were projecting for the company. Shouldn’t it be the other way around? Anyway, it is public knowledge whenever an insider buys or sells their own shares.
A few years ago some studies suggested that buying with the insiders gave you a slight edge… but I’m willing to bet the cycle has changed since those studies were released.
RE Buffett: his success over the past 30 years is mainly due to owning a single insurance company (Geico) and a lot of his edge has come from tax strategies. He is astute, but that doesn’t mean his market insight amounts to anything very special.
I do believe Soros has special insight, but he is a macro-strategist and doesn’t worry much about individual companies. The fact he predicted OPEC would lose their grip on oil prices or that the pound would bust had nothing to do with insider information. But if some of his big bets had gone the wrong way early in his career we wouldn’t be talking about him.
12. October 2009 at 00:18
I think it depends on the people involved. Where I work I think the CEO knows a great deal about the company.
I think “cooking the books” has a lot to do with it too. What Anthony Bolton said once was that he needed to look a CEO in the eye. What an analysts can get from knowing the CEO is not necessarily information about the business, but perhaps information about the CEOs personal competence and honesty.
12. October 2009 at 11:08
“When you said investors expected huge gains I assumed there was a 20% or 30% premium in futures prices. If not, I don’t understand how you can say investors expected huge gains. I pay little attention to survey data as I don’t find it reliable. I am much more interested in asset prices as indicators of expectations, such as futures or forward prices.”
Black-Scholes proved that options do not provide the information about expected return, as option prices depend on the risk free rate and volatility of the underlying asset. It gets even worse as dot com stocks were hard to borrow and short, so in practice combo of dot com stock hedged by short sale has expected return lower than risk free rate. On the other hand distribution of option prices at different strike prices do provide the information about the shape of the expected return distributions, and sometimes they indicate that there is a momentum effect expected (i.e. most probable scenario is a small gain, and there is a slight probability of huge loss).
“My hunch is that if investors really expected huge gains then they would have sold all their T-bonds and poured into the stock market. But just as many T-bonds were held during the tech boom as before the boom. That makes me think the expected rates of return weren’t that different.”
You should focus not on quantity of T-bonds (it is determined by Treasury when it picks the maturity profile of government debt) but on price of T-bonds. Oops – price of T-bonds is determined not by libertarian NGDP futures market, but by expectations of what wise men from FOMC will do in future. This all means that we have to look elsewhere for return differential between T-bonds and dot com stocks. Rising quantity of stocks held on margin during the tech boom gives us the information that return expectations of dot com stocks were very high (at least by the opinion of margin account holders).
14. October 2009 at 15:51
Current, rob, and 123, I will defer to all three of you, who know more about the details than I do.
123, I think you are probably right about the subset of investors who bought tech stocks, they probably did expect big returns, as you indicate. But here is a question (and I honestly don’t know the answer), doesn’t there have to be some margin at which certain investors are indifferent between holding tech stocks and T-bonds? And is that marginal investor the “market expectation.” I should know this stuff, but my knowledge of fiance is sketchy. How do finance-types characterize market expections if they are not normal, but double-humped? Do they go with a mean? Or the marginal investor?
15. October 2009 at 08:59
EMH/CAPM models that assume unlimited leverage and no short selling restrictions have marginal investors and “market expectations”. Other models relax these assumptions, they usually have two classes of investors – noise traders and informed investors. These models try to model the size of deviations from the efficient outcome. By the way it was very hard to borrow tech stocks for shorting, and it was very easy to access leverage via margin debt.
17. October 2009 at 08:45
123, Why was it hard to borrow tech stocks?
17. October 2009 at 11:55
Free floats were low, and not every investor in the free float participates in lending of stocks.
17. October 2009 at 12:20
Scott, you said:
“It’s probably not worth me debating the issue of moral hazard, as i agree with the basic point that bailouts have created a moral hazard problem, and in particular have encouraged too much risk taking. But I am no expert on banking. Perhaps more bankers have struck it rich than I assumed. I admit I was relying on press reports that suggest Wall Street is really suffering, and of course press reports may be highly unreliable.”
Gary Stern, former President of the Minneapolis Fed addreses all your doubts pretty well in his interview with Russ Roberts here:
http://www.econtalk.org/archives/2009/10/gary_stern_on_t.html#highlights
18. October 2009 at 10:34
123, I think we are going to hit 200 comments, as now you’ll have to explain free floats to me (I warned you I’ve never taken a finance course.)
Thanks for the Stern article, I read the whole thing and learned a lot. Let me back up to the point I thought I was defending (although it’s been a while so my memory may be faulty.)
I was trying to argue that investors, including elite bankers, did not expect the crash. So the market wasn’t irrational in that sense. Here’s where I’ll agree with you. Let’s suppose that without all the moral hazard discussed by Stern, and also faulty corporate governance (which I suppose is non-governmental moral hazard) then perhaps investors never would have piled into the subprime mortgage bonds unless there was less that a 5% chance that the housing market would blow up. With all the moral hazard perhaps they were willing to pile in with up to a 15% chance of it all blowing up. If that’s what we are debating, then I agree. And I suppose reading you and Stern has made me shift the numbers in your direction a bit. But I never thought these guys actually expected the fiasco (in the sense of a 50% chance) and still don’t. That was probably my orignal point. I would certainly never argue that moral hazard wasn’t a problem, as I have done blog posts complaining about moral hazard from FDIC and too big to fail. I have criticized the LTCM bailout. 1998 would have been a great time for a financial crisis, as the real economy was so strong. It would have been like taking a flu shot when healthy. With Lehman we took a flu shot when we were already sick–not a good idea.
24. October 2009 at 01:54
Free float is just a percentage of company shares owned by minority shareholders. Usually majority shareholders do not make shares available for lending. If a free float is owned mostly by individual investors, they usually don’t participate in stock lending market. This all meant that many dotcoms were very expensive to borrow and short. Whereas if you want to short Wall Mart, many funds that own shares will happily lend you stock at attractive rates.
Investors and elite bankers did not expect the market to crash. But this has very little to do with the efficiency, as most managers of soviet collective farms did not expect communism to crash too.
Combination of moral hazard and subprime CDOs is very dangerous, as these securities have enormous leverage and they blow up when market expectations of housing crash increase from 5% to 6%.
Expectations of the fiasco (in the sense of a 50% chance) is not the appropriate criterion for the efficiency of the banking system. For example, difference between the expected default rates of AAA and AA bonds is much less than 1 percentage point.
I agree that both governmental and non-governmental moral hazards are important. I believe that good corporate governance cannot solve the efficiency problems of big banks, as they are too big to be efficient.
I’m much less opposed to LTCM bailout, as private sector took all the losses, and government coordination simply gave time for orderly unwinding of the LTCP positions, thus preventing the spread of liquidity crisis. After LTCM banks learned the lesson and no longer gave hedge funds 40x leverage, and during this crisis hedge fund problems did not have any systemic implications.
24. October 2009 at 04:59
123, That’s a good point about LTCM. Indeed before anti-trust laws a private consortium of banks might have done the same. Perhaps even in 1998 they might have, but I think anti-trust might limit that sort of collusion. In any case I think you’re right.
Regarding the Soviet analogy, it is equally true that few Western experts saw the collapse coming. My point about the anti-EMH position is that it’s hard to come up with policy implications, as many things that seem silly in retrospect were not well understood by either investors or regulators. It is impossible, ex ante, to know who the next Nouriel Roubini’s will be and who is just one of those bearish doomsters that is always predicting that the dollar will crash or that Treasury bond market will collapse, etc. They also seem to have good arguments, but the day of reckoning for America never seems to come (in an international accounts sense, or in a federal debt sense.) Maybe someday it will come, and optimists like me will seem stupid in retrospect.
24. October 2009 at 09:38
Continuing with the LTCM analogy, private person J.P. Morgan did much better in 1907 than Bernanke and Paulson in 2008.
Anti-EMH view has lots of policy implications. The most important is that bailouts of bondholders of too big to fail institutions are harmful. There are plenty of others. For example, one source of market inefficiency is incomplete markets, so Shiller is trying to create new financial instruments that could help to hedge the house price risk.
Anti-EMH view says that Fannie and Freddie did lots of harm.
The summer of 2008 was a day of reckoning for America in terms of international accounts (just look at the import prices).
70s were problematic for the Treasury bond market.
You said:
“Regarding the Soviet analogy, it is equally true that few Western experts saw the collapse coming.”
I fully support the EMH view that it is impossible to forecast the year when North Korea will collapse. Indeed, North Korean economy is so efficient, it is even impossible to predict if more than 100 000 people will die of famine in 2015.
25. October 2009 at 06:46
I also thought of the JP Morgan example.
I don’t follow those anti-EMH implications, as I believe in both the EMH and those views. I oppose bailouts and I oppose government sponsored entities like Fannie Mae, etc. as they interfere with the free market. So they aren’t anything that distinguishes pro and anti-EMH views.
Both Shiller and I support he creation of NGDP futures markets.
I don’t follow your mid-2008 argument, I think the problem was falling NGDP after mid-2008. If NGDP had held up, the adjustment is international balances would have been far less painful.
26. October 2009 at 12:26
Policies suggested by anti-EMH can also be supported on other grounds (small government, distributional justice, etc.). The difference is that anti-EMH attaches much greater importance on such policies, because these policies make prices less wrong in auction type markets, increase information content in prices, increase real growth and reduce real volatility. EMH is just a simplification that is useful in some contexts, but it does great harm when evaluating the impact of abovementioned policies. EMH also discourages research that tries to measure the degree of market efficiency (try to imagine that Newtonians tried to stop the research of the theory of relativity). It also steers reality-based people away from Chicago school.
Regarding mid-2008 and 70s I just wanted to say that opinion of bears is justified by historical record. I agree that they will be less frequently right if monetary policy will be improved.
26. October 2009 at 13:58
123, you can’t be as general about it as that. Most of the anti-EMH stuff I’ve read have been attacks on the market.
The point of opposing the EMH is to construct an underhand way of saying that bureaucrats can make better decisions than markets.
26. October 2009 at 14:51
Current,
“The point of opposing the EMH is to construct an underhand way of saying that bureaucrats can make better decisions than markets.”
I think that is about 85% of the story. The other 15% is financial professionals trying to convince you that you should pay for their expensive services, instead of just dollar cost averaging into index funds. Clearly hedge funds, for instance, do not want you to believe in the EMH, but they do not want more regulation of markets either.
26. October 2009 at 15:05
In fact, as long as the meme out there is that anti-EMH implies, as you say: “bureaucrats can make better decisions than markets”, it puts Wall Street in a tough position. What side of this argument does it benefit a financial firm to take? Oh — who am I kidding? It will be win-win for Wall Street. The more bureaucracy, the more people WILL need to pay for professional financial services, to steer them through the maze. Don’t most people on Wall Street already vote Democratic these days?
27. October 2009 at 11:46
Current,
I agree that some people oppose EMH because they like socialism. But it is much better to measure the efficiency of markets instead of assuming away various frictions. Properly measured, inefficiencies of markets are much smaller than ineffciencies of bureaucracies.
rob,
just because anti-EMH is useful for Wall St., it doesn’t make EMH true. On the other hand, I agree that the most useful result of EMH is that most people should not pay for expensive asset management.
27. October 2009 at 11:59
Comparing the two is what institutional analysis is about, and what the socialist calculation debate is about. Comparing these sort of things empirically is extremely difficult.
The point of the EMH is, as we have said before, that Economists should not presume think they can do better than speculators.
27. October 2009 at 16:26
123, I think it is the EMH that emphasizes the importance of getting prices right. For instance, it is often EMH types who oppose laws against insider trading, because those laws prevent price sfrom moving to their equilibrium level.’
You said;
“Regarding mid-2008 and 70s I just wanted to say that opinion of bears is justified by historical record. I agree that they will be less frequently right if monetary policy will be improved.”
Not sure what this means. When stocks prices go down the bears like Roubini are right. When prices go up (such as the period since March) the bears like Roubini are completely wrong. Bears are sometimes right and sometimes wrong. What can we infer from this?
Current, That was also my impression. It’s a bit odd to use market friendly as the measure of success in a post where one also denigrates the Chicago school.
rob, I agree about investment advisors. I can’t believe how many smart people I know pay people to manage their money. Just buy indexed funds.
123#2, I completely agree.
BTW, Current’s first comment was #200, a first for this blog. Although without 123 I never would have made it. So thanks to everyone.
30. October 2009 at 09:41
Support for laws against insider trading is independent from EMH/anti-EMH views. Friedman supported EMH and criticized insider trading laws, typical college professor supports both EMH and insider trading laws, Galleon founder Raj Rajaratnam opposes EMH and will probably argue in court that insider trading laws are unjust. I oppose EMH and support some forms of insider trading laws, although it is difficult to draw a bright line which information is material, who has a fiduciary duty etc.
Lawyers are much less invested in EMH dogma and are more capable of a balanced discussion. Professor Bainbridge says “Given the inefficiency of derivatively informed trading, the market efficiency justification for insider trading loses much of its force.”, his whole blog post is worth a read:
http://www.professorbainbridge.com/professorbainbridgecom/2009/10/learning-to-love-insider-trading-lets-not.html
Scott, you said:
“Not sure what this means. When stocks prices go down the bears like Roubini are right. When prices go up (such as the period since March) the bears like Roubini are completely wrong. Bears are sometimes right and sometimes wrong. What can we infer from this?”
you also earlier said:
“It is impossible, ex ante, to know who the next Nouriel Roubini’s will be and who is just one of those bearish doomsters that is always predicting that the dollar will crash or that Treasury bond market will collapse, etc. They also seem to have good arguments, but the day of reckoning for America never seems to come (in an international accounts sense, or in a federal debt sense.) Maybe someday it will come, and optimists like me will seem stupid in retrospect.”
I just wanted to say that events in 70s and 2008 imply that it is worth listening to Roubini and Shiller, as extreme volatility is quite common. Saying that bears are sometimes right and sometimes wrong is too simplistic, as Roubini and Shiller can help you to measure the degree of ex-ante risk.
You said:
“I can’t believe how many smart people I know pay people to manage their money. ”
How do they select mutual funds they like? I think that EMH is correct in that most people are incapable of selecting a goood mutual fund. On the other hand, I can imagine myself investing in some low cost mutual fund that exploits Shiller effect and some other inefficiencies, and I would recomend such funds for some people.
30. October 2009 at 09:52
The strongest part of Krugman article is “PANGLOSSIAN FINANCE”, and it is about EMH myth. It is a sad fact that EMH has blinded many free marketers who were clueless during the crisis, and Krugman is able to use the truth about EMH to further his socialist agenda. By the way Krugman called the housing bubble in 2005 and this fact adds to his ability to persuade people.
31. October 2009 at 08:00
123, I only read part of the Bainbridge piece, but I wasn’t impressed by what I read. In the Texas example he argues that laws against insider trading would make it easier for the firm to steal land away form the local landowners for much less than it is worth. In other words, we should ban insider trader so that firms can more easily insider trade in the property market. But I also question the relevance of the example. How often would a rising stock price of a corporation tip off a specfic landowner in a specific location that their land is valuable. Suppose you owned a ranch in Texas and you suddenly saw Exxon of Chevron stock rising. What implications would you draw?
And I also question his assertion that insider trading rarely has a significant effect on stock prices. The sort of insider trading that is legally questionable, like buying on knowledge of a future takeover attempt, often moves stock prices quite sharply before the announcement.
Let the market determine what sort of insider trading is efficient. If employees sign an agreement not to trade on knowledge of a takeover attempt, then let the courts enforce those contracts. The burden of proof is on those who want an insider trading ban. As you say these laws are very difficult to craft, as the term “insider trading” is often in the eye of the beholder. So far they haven’t provided evidence to overcome the presumption that no regulation is best unless proven otherwise.
Regarding Shiller and Roubini, there is no mystery as to how much volatility we have had in the past, the mystery is how much we will have in the future. I don’t think it is at all obvious that we will see 1970s-style inflation again. Indeed I think it quite unlikely that double-digit inflation will return. (Or double-digit deflation as in the 1930s.)
I think you overrate Shiller and Roubini due to survivor bias. Those academics who predict recent events well will be better known that those academics who don’t predict recent event well. But they might have just gotten lucky.
You said;
“How do they select mutual funds they like? I think that EMH is correct in that most people are incapable of selecting a goood mutual fund. On the other hand, I can imagine myself investing in some low cost mutual fund that exploits Shiller effect and some other inefficiencies, and I would recomend such funds for some people.”
That’s fine, but it doesn’t address the issue of why people pay others to manage their money. The money managers are not putting them into funds that exploit the Shiller theories, they are going into high-cost funds. And these are college professors in a business school.
But then I also frequently hear professors ‘brag’ about how big their tax refund was. So maybe I should discard the EMH.
🙂
123#2, Almost economists, including Krugman, were clueless during the crisis. The crisis did not occur in 2005, nor did Krugman predict crisis that actually did occur. (Roubini did however.) Krugman predicted a housing bubble, which is very different from a banking crisis. It should also be said that as recently as 2007 or 2008 he predicted Fannie and Freddie were in good shape. How did that prediction turn out? For more than 15 years I have been calling Fannie and Freddie a time bomb waiting to explode.
When it comes to the policy implications of the crisis, Krugman completely missed the problem of the federal government regulators trying to get as many low income people into housing as they could. He even mocked those who were worried about that issue. Instead, he argued we needed even more such “regulation.”
31. October 2009 at 14:01
I thought that only Krugman would say “steal land away”, and I thought it is better to say “purchase land at prices that do not reflect information produced by the oil company”. Bainbridge uses Texas example to prove a narrow point that rules that permit corporations to delay disclosure are OK, and shorter version of his post could have ommited this example. I also have an impression that Bainbridge might agree with you that a rising stock price of a corporation would not help a specfic landowner in a specific location.
Bainbridge does not deny that insider trading might strongly change prices in some circumstances, he just correctly argues that the effect is not strong enough to reach the social optimum and efficient pricing.
I agree that market should help to determine what kind of insider trading is efficient. It would really be nice if there were no federal insider trading laws and state regulations according to the place of incorporation would apply.
Fortunately neither Shiller nor Roubini forecast double digit inflation or deflation. Maybe Nassim Taleb does, but he is too nervous, just look how his hands are shaking when he is interviewed on TV.
Your reaction to Q3 GDP report is no more positive than Roubini’s.
I don’t overrate Shiller and Roubini. I think that Roubini is top quartile economic forecaster, sometimes right, sometimes wrong but with positive alpha. Shiller is not an active forecaster, but his model is important and there is strong statistical evidence that his model isn’t just a result of survivorship bias.
Your story about business school professors investing in high cost funds makes me think that it is easier to find John Galt than to find the marginal investor who makes market prices efficient.
Krugman has got lots of details wrong, but his 2005 big picture assessment that housing prices were not efficient really helps his propaganda efforts.
Last September Krugman wasn’t clueless, he was one of the best sources for policy analysis:
http://krugman.blogs.nytimes.com/2008/09/30/people-i-agree-with-part-one/
Too bad that he later advocated wrong kind of stimulus, fortunately after a very long search I was able to locate a person who advocates right kind of stimulus in a blog called Money Illusion.
I agree with Krugman that structured finance and not low income housing regulations were the main source of the housing bubble. I agree with you that such regulations are harmful.
31. October 2009 at 15:45
Here is a timeline of Krugman losing belief in the credibility of Fed:
http://krugman.blogs.nytimes.com/2008/09/12/did-i-say-in-i-meant-de/
http://krugman.blogs.nytimes.com/2008/09/17/worse-than-ever/
http://krugman.blogs.nytimes.com/2008/09/22/the-humbling-of-the-fed-wonkish/
I think prescient is a better label than clueless…
2. November 2009 at 06:56
123, I agree with much of what you say, but it has no influence on my views that insider trading bans are foolish. Rather it makes me have a higher opinion of Bainbridge if he recognizes that insider trading can be helpful in some situations. I would just add that the government has absolutely no way to distinguish between insider trading that is helpful, and insider trading that isn’t.
Roubini made a great call on the financial crisis, but I certainly don’t see him as a forecaster. If I had followed his investment advice last spring I would now be about 40% poorer. That was just a horrible call he made when he warned people against investing in stocks.
I agree that Krugman often has excellent analysis. Recall I had a recent post (12 Krugmans) that praised Krugman for exactly the calls you mentioned in the links you provided. But I would add that there are some very serious errors in the last link you provide.
1. He claims the BOJ tried an easy money policy, and it failed. This is absurd, as I showed in endless posts. They targeted 0% CPI inflation, and got 0% CPI inflation. Arguably no other major central banks was so successful in terms of its own target.
2. He claims the point of easy money is to lower long term bond yields. That is wrong. A policy that is expected to be expansionary may lower short term yields, but it will usually raise long term yields.
So even the analysis you link to is hit or miss. As far as his prescience, he did not predict the economic crisis. Yes he predicted the housing bubble, And yes, he analyzed the economic crisis once it had occurred. But (unlike Roubini) he did not predict the crisis. Indeed he even admitted that he didn’t predict it when interviewed by NPR after winning the Nobel.
2. November 2009 at 13:43
Bainbridge recongizes that insider trading reveals information to the market, albeit to much less extent than implied by the EMH. It is amazing how a lawyer can have much better understanding of information economics than average economist.
Bainbridge is one of the loudest voices of the blogosphere saying that many recent SEC insider trading cases are not just.
General idea of insider trading laws is sound, as it is much better to compensate executives directly, instead of letting them make unlimited profits by using insider info. Insider laws that assign the ownership of material information to all shareholders have reduced the cost of capital in America.
I’ll answer about Roubini and Krugman later.
3. November 2009 at 06:44
123, If it is better to compensate CEOs directly, then why don’t companies have policies against insider trading? Indeed I was told that they often do have just such policies, which would seem to make the laws unnecessary in those cases.
3. November 2009 at 13:59
Enforcement of company policies is not strong enough.
4. November 2009 at 06:44
123, Wasn’t there a William’s Act, or something, that made companies go public after they had bought at least 5% of another company’s stock? If so, that seems to be a recipe for insider trading. Why not repael it, so companies could internalize the value added from the synergies that result frm mergers?
4. November 2009 at 15:25
5% law affects only the distribution of synergy gains between shareholders of acquiror and shareholders of the target company.
Your points re Krugman:
1. Now Krugman has changed his mind and thinks that BoJ has managed everything pretty well:
http://krugman.blogs.nytimes.com/2009/03/09/japan-reconsidered/
2. Krugman ignores expectations channel here, but in other places he uses models that include that. Model without expectations was a reasonable choice for his post, because Fed does not target expectations.
5. November 2009 at 18:07
123, I’m surprised by your response. I had thought there was fairly general agreement that the 5% rule was a huge inducement to do illegally (insider trading) what used to be legal (quietly buying up much more than 5% of the stock.) Is that wrong?
Is he saying Japan did well? I thought he said we are making the same mistakes. He doesn’t know much about Japan if he doesn’t think there was a lot of misery created. The unemployment rate there is very misleading, there is massive underemployment.
I would say the Fed does target expectations to some extent, just not very effectively. Krugman still thinks QE failed in Japan, even though the BOJ got precisely the 0% CPI inflation they targeted between 2001 and 2008. He seems to think thay failed to get inflation up as high as they wanted it, which is just not true.
11. November 2009 at 12:53
The real issue is not the perceived moral damage of insider trading, but the question of the optimal contract between shareholders and management, and the separate question of the optimal contract among shareholders. 5% rule has nothing to do with relationship between shareholders and management, and is regulatin only the relationship among the shareholders.
Maybe 5% rule is not the optimal rule, but it is close enough. In Germany lack of the enforcement of similar rules led to the absurdity of Volkswagen having the largest market capitalisation in the world for one day as a result of Porsche/Volkswagen affair.
Recently I saw an argument that growth of real GDP per person of working age was quite normal during two lost decades.
12. November 2009 at 07:20
123, If we are trying to protect shareholders from management, we need to change the laws on coprporate control. Insider trading laws cannot protect shareholders from management.
I am not familiar with the VW issue. And why did high share prices only last for one day. Presumably you mean the buyers were ripped off, as prices fell sharply. Who were the buyers, and why did they suddenly pay such high prices?
I don’t follow the last statement–are you referring to Japan? Certainly unemployment and underemployment are much bigger problems than 20 years ago. Read any contemporary description of Japan and they talk about how the economic problems have severely affected the country’s psyche.
12. November 2009 at 13:38
Protect is the wrong word, or else I might say that NGDP targeting won’t protect workers who are now unemployed from companies.
Instead, if we focus on efficiency of institutions, NGDP convertibility and insider trading prohibitions will maximise the pie.
Prices were too high for more than a month, but world record lasted only a day. Porche secretly bought 100% or more of the available shares, so anybody who was short Volkswagen had to pay enormous money to close the short when knowledge of 100% ownership became public. One formerly rich businessman commited suicide after this.
Oops sorry I was indeed referring to Japan. Current unemployment must be structural.
15. November 2009 at 09:10
123, I still don’t understand the VW stock example. Were the high prices expected to be permanent? What did the futures markets show? No short seller should have been forced to buy stocks at an unsustainable price, if there is a futures market for that stock.
15. November 2009 at 09:58
Expected future prices were not determined, as Porsche and few other shareholders owned more than 100% of VW.
15. November 2009 at 11:02
If insider trading were legal, think of how the job of stock analysts would change. They would no longer accept “I can’t tell you that” as an answer. They would work like lobbyists and salespeople, wining and dining managers and sales execs and accountants to find out if a big contract just got signed or if the exploration well succeeded. This would make markets less efficient, not more, because information would be less symmetrical; big firms with big resources would consistently know crucial information ahead of smaller rivals; financial services providers would actually serve a service, unfortunately.
OK, companies can have policies about confidential information, etc., but a company policy is not as strong a deterrent as jail-time. Jeez, who gives a shit about company policy anyway? I’m not doing my job if I can’t get someone to tell me something they’re not supposed to. (But they wouldn’t go to jail if they got caught.)
This is tangential, but having worked in Latin America a lot I have a theory about corruption. In Mexico, for instance, workers for Pemex can go to jail if they merely make a mistake on a commercial contract and fail to follow procedures to the very letter. Why? Because there is much corruption and there is the perception of much corruption. The perception becomes key. But it also becomes self-fulfilling: if you can go to jail for something that isn’t really corruption, why not take a bribe? You aren’t much more at risk whether you do or don’t.
Somehow I think that is within the ongoing theme of expectations creating reality.
17. November 2009 at 10:04
123, I’ve probably reached the limit of what I can say about VW, I just don’t know a lot about the case.
rob, Corruption is usually caused by people getting around government regulations. So I think the ban on insider traders just encourages people to evade the ban, it actually creates crime. And the activity you describe makes markets more efficient, by moving the price to its appropriate level.
17. November 2009 at 20:14
Scott, you said:
“And the activity you describe makes markets more efficient, by moving the price to its appropriate level.”
But what GOOD does this super efficiency do anyone? Does it prevent bubbles? Even the Enron guys believed their own BS until the last minute.
At the extreme, the perception of massive injustice creates problems. E.g., Socialism. Should Ben Bernanke be allowed to trade bonds? It would make prices slightly more accurate, but there might be other problems…
With insider trading, at some point one’s duties come into conflict with one’s interests. Imagine a CEO who shorts his company and then drives it into the ground. It’s a lot easier to have control over surprising on the down side. Why does the public have such a problem with the idea that Pete Rose bet on baseball? I have no problem with it, because I know in my heart Pete never bet on the Reds to lose. But someone eventually with no pride will be motivated to bet on their team to lose…
I think the costs of insider trading outweigh whatever subtle benefits.
17. November 2009 at 20:23
You also said:
“rob, Corruption is usually caused by people getting around government regulations.”
I think we are getting close to a tautology. Isn’t crime caused by people breaking the law? When I was much younger and worked for minimum wage I used to steal money out of the cash register at the end of the month to pay the rent. In my opinion then and now the cost to the company of my stealing was merely the same as my wages should have been in the first place. But I was breaking the law, because the social construct said otherwise…
But do I believe some laws exist for a good reason? Yes.
20. November 2009 at 14:48
VW is just a recent case of cornering the market. You might be more familiar with Hunt brothers or other cases from here:
http://en.wikipedia.org/wiki/Cornering_the_market
25. November 2009 at 13:17
Here is a good article by Russ Roberts that supports the views of Arnold Kling on banking that we discussed in this thread above:
http://www.invisibleheart.com/How%20Little%20We%20Know.pdf
26. November 2009 at 07:20
rob, I agree that there should be corporate policies against insider trader. But take something like the Martha Stewart case–what possible harm could be done by her insider trading?
Furthermore, I have a real problem with criminalizing activities that can’t be defined.
rob#2, I meant inappropriate government regulations.
123, As I recall the Hunt brothers lost money on their activities. That’s why cornering the market occurs so rarely–it is foolish. When you try to sell your holdings, the price collapses.
123#2, Yes, that is a very good article.