Who are the famous bubble deniers?

More and more I think that the entire bubble/anti-EMH approach to economics is founded on nothing more than superstition.  Superstitions are caused by cognitive illusions; we think we notice more patterns than are actually there.  You dream your son got in a traffic accident, and the next day it happens.  You forget the other 10,000 dreams that didn’t predict the future. 

In economics people notice bubbles bursting, but fail to pay much attention to bubbles not bursting.  But I admit I might be wrong, so I’ll give my opponents one more chance.  If it’s not really a cognitive illusion, then bubble deniers who are right ought to be just as famous as bubble predictors who are right.  Indeed as we will see they should be even more famous.

People who actually understand finance know that if the term “bubble” is to mean anything useful, it must contain an implied prediction of the future course of asset prices.  Not a precise prediction (everyone knows that would be impossible) but at least a better than a 50/50 prediction.  If someone said in 2005 that housing prices were a bubble, but still was unable to offer more than a 50/50 odds on whether real housing prices would rise or fall over the next 5 or 10 or 20 years, then what would their assertion actually mean?  Asset prices are very volatile; we know that at some point all markets will go down.  When I read predictions from people like Paul Krugman, I infer that there is an implied prediction that real prices will fall over some reasonable period of time—say 5 years.

And of course Krugman was right in predicting that real US housing prices would fall in the 5 years after 2005, as was Dean Baker, Nouriel Roubini, and some others who became well known and lauded for their predictions.

At the same time we also know that bubble-like patterns don’t always yield reliable predictions of future trends.  The Australian housing market looked just as bubbly as the US market in 2005, but since then has soared much higher.  Some day it will fall, but the 2005 prices no longer look excessive.

If people like Robert Shiller (another person who became famous from bubble predictions) are right about asset prices being too volatile, then it should be true that US-type cases are more common than the Australian case.  I don’t think that’s true— in most developed countries real housing prices have risen since 2005, despite real upswings before 2005.  But let’s say I’m wrong and Shiller’s right.  Then the easy prediction to make is that prices will fall after a big upswing.  The much harder prediction is that prices will keep rising, even from inflated levels.  Those cases would be much rarer, and those who correctly call them when they occur (as in the Australian housing case) should be lauded as great heros of the investment world.

So who are they?

If they don’t exist, I have a theory why.  Most people are convinced bubbles exist, regardless of the data.  Hence if the prediction doesn’t pan out, then the market was in some sense “wrong.”  Traders haven’t yet woken up to the stupidity of their behavior.  When they do, prices will crash and the bubble proponents will be proven right in the long run.  So most people would implicitly think; “Why praise someone for being right for the wrong reason?”  Of course this makes bubble theory into a near tautology, irrefutable in volatile asset markets that will almost always eventually show price decreases.

Perhaps I am wrong and there are lots of famous bubble deniers out there.  But if not, that would in my mind be the last nail in the anti-EMH coffin, pretty much confirming that people are seeing what they expect to see, indeed given the satisfaction we get from seeing the high and mighty brought low, perhaps what they want to see.

One final point; I also have noticed that lots of people are given credit for bubble predictions that were wrong.  John Kenneth Galbraith saying stocks were a bubble in January 1987.  Robert Shiller saying stocks were a bubble in 1996.  Dean Baker saying US housing was a bubble in 2002.  The Economist magazine touting its successful housing bubble predictions of 2003 in an ad, despite the predictions being incorrect for most of the countries listed.  That’s how strongly we want to believe in bubble predictions—we even assume that people who were wrong, were actually right.

PS.  For those interested in global housing prices, The Economist has a great interactive graph.  It helps to show the pattern from say 2000:1 to 2005:1, and then from 2005:1 to the present.  In the earlier period almost all countries showed gains in housing prices, even in real terms.  The two notable exceptions were Germany and Japan, where prices fell sharply.  If I was to use a Shiller-style model that predicts asset prices will self-correct after excessive swings, I would have predicted most housing markets to slump after 2005:1, but Germany and Japan to rise.  Instead almost the opposite happened.  Germany and Japan continued to do very poorly, while almost all other markets rose in nominal terms, and most rose even in real terms.

The two nominal decliners (in addition to Germany and Japan) were the US and Ireland—which is why people assume they had had a bubble.  But why didn’t all the other bubbles collapse?  Perhaps because asset prices are not as easily forecasted as most people naively assume.

BTW, if you can’t get a 2000:1 starting date, then white out all the country boxes and start over.  That worked for me.


Tags:

 
 
 

113 Responses to “Who are the famous bubble deniers?”

  1. Gravatar of Morgan Warstler Morgan Warstler
    8. December 2010 at 06:21

    Its so weird you would count a bunch of guys without money in the market. Who cares what guys without skin in the game think?

    The bubble deniers are the guys who stay in and are proven right.

    The bubble predictors are the guys who get out before the pop. Mark Cuban was the Internet bubble predictor, everyone else was just yammering.

    Right now there is much discussion about Tech Bubble 2.0, because the ONLY place people are finding returns is my sector, so everyone is suddenly an angel investor again.

    —-

    BTW, I was super excited you were going to explain when during the past booms Scott Sumner’s NGDP targeting would have popped them an how.

    Shouldn’t you be able to model that?

  2. Gravatar of Mr. Winston Mr. Winston
    8. December 2010 at 06:32

    Jeremy Grantham would claim he is a bubble predicting superstar…and he certainly fits in the anti-EMH camp. You could take a look GMO’s Global Tactical Asset Allocation product to see if he really has added value over time.

  3. Gravatar of 123 123
    8. December 2010 at 06:38

    “Perhaps I am wrong and there are lots of famous bubble deniers out there.”

    In Australia bubble denial is a way of life, and the reason you know no famous Australian bubble denier is that there are too many of them. Do you know any famous Vegemite eater? It is very likely that he is a bubble denier too.

    I’m back home and will catch up with other comments soon…

  4. Gravatar of Mattias Mattias
    8. December 2010 at 06:52

    I think one possible reason for this is that, in general, pessimists are considered much more “serious” than optimists. For some reason people are attracted to doomsday visions and market crash forecasts, which is strange considering the armageddonists’ track record.

  5. Gravatar of Dan Carroll Dan Carroll
    8. December 2010 at 07:03

    It’s not hard to find those that recognized the Internet bubble, the housing bubble, and related credit bubble. Just not in the newspapers, who make money off of the sensational. Indeed, I think there may now be a bubble in predicting the next bubble. What many (but not all) missed was just how deep the credit bubble went and how destructive the burst would be.

    When comparing prices to fundamentals (p/e ratios, price/rent, etc.), against any reasonable forecast horizon, and against any reasonable estimation of risk, most bubbles are easy to spot. “Price-to-eyeballs” doesn’t qualify.

    I know nothing about Australia, other than the fact that it is commodity rich, so it is riding commodity prices and China. Thus, rising commodity prices and exports to China are likely to raise incomes, which impact real estate. Thus, I am not surprised to see rising prices in Australia. Whether there is a bubble, I can’t comment.

    Unfortunately, when bubble spotters are right, they tend to be ridiculed and even fired.

  6. Gravatar of Ram Ram
    8. December 2010 at 07:04

    Consider a view that is neither fully in favor of nor fully opposed to the EMH. On this view, what constitutes a bubble is a (long) period of time during which noise traders enjoy significant market power, such that other traders stand to make more money in the short run by predicting what the noise traders will do next than by predicting the long-run value of the security being traded. The view is that such episodes frequently occur, that continuous rising and falling trends make such episodes approximately identifiable in retrospect, but that they’re impossible to consistently forecast beforehand. In other words, the EMH is true for every practical purpose (inasmuch as it claims that you cannot consistently outperform the market over the long run), but we may in retrospect identify episodes during which markets failed to efficiently aggregate all (or nearly all) publicly available information. This view would have no policy implications different from the EMH, but it seems to me to be a more realistic view.

  7. Gravatar of Bill Gee Bill Gee
    8. December 2010 at 07:30

    First of all, I just want to say that I just found your blog and find it facinating. I’ve been teaching economics for two years and you’ve got to be one of the only economists I’ve found that doesn’t get “hooked on the hype” like so many of the talking heads and even many Economics textbooks seem to do.

    Regarding the ability to predict bubbles, I think you’re right on the money here. Predicting a bubble is a lot like predicting a hurricane three years before it happens. We all know that eventually something will happen, but damned if we’ll ever get a handle of what exactly will happen. But unlike a hurricane, whose effects can be predictable when and if it does hit a coastline, the effects of bursting bubbles can only be understood in terms of a postmortem.

  8. Gravatar of Lee Kelly Lee Kelly
    8. December 2010 at 08:05

    A bubble is just a naturally occurring Ponzi scheme. There is no central conspirator deceiving investors, but people behave much as though there is. I grant that an actual bubble is difficult to identify, but they are surely a theoretical possibility. The probability of bubbles and their size seems to increase when other factors “conspire” together to change expectations or distort the price system, e.g. excessively low interest rates created by central banks or government programs that quickly change the demands for particular assets.

  9. Gravatar of jsalvatier jsalvatier
    8. December 2010 at 08:12

    Now, bubbles are obviously hard to identify, but they do exist. The experimental evidence is basically indisputable: http://www.theatlantic.com/magazine/archive/2008/12/pop-psychology/7135/

  10. Gravatar of Bill Woolsey Bill Woolsey
    8. December 2010 at 08:14

    Scott,

    Will a monetary base of 1.6 trillion lead to nominal GDP greater or less than 15,000 billion over the next 5 years?

    Clearly, monetary theory is just an illusion.

    You tell me that there exists some quantity of base money sufficiently high that will generate 17000b in NGDP in two years? What does that mean exactly?

    Let’s turn it around.

    Clearing, increasing the quantity of money will not increase real output, income, and employment, except maybe in a slight and temporary fashion if it is unexpected. How do I know this? Because right now, real GDP is equal to the productive capacity of the economy. Real expenditures is just sufficient to produce the total amount of goods and services firms are able to produce. How do I know that? Well, if it weren’t true, then there would be surpluses of goods and services and resources like labor at current prices and wages. And if there were such surpluses, prices and wages would be lower, and real expenditure would be higher, so that no such surpluses would exist. The current level of prices and wages must be the one where real expenditures equals the productive capacity of the economy. QED.

    As you know, I believe that there is a quantity of base money that will result in NBDP being 17 trillion in two years. I don’t know what it is.

    I believe that real expenditure and output is a good bit lower than the productive capacity of the economy. Real output and real income have fallen enough to bring the demand for money to match the existing quantity.

    And, I believe that speculative bubbles are possible. And further, that there was one in housing.

    But I don’t pretend that I can predict the course of future asset prices. I sure don’t favor having a central bank manipulate short term interest rates or the supply of credit to try to pop bubbles before they get started.

    But logic tells us that bubbles are possible. And historical interpretation allows them to be identified. It is not much different than identifying situations where there is an excess demand for money.

  11. Gravatar of rob rob
    8. December 2010 at 08:36

    Isn’t Fama at least as famous as Shiller?

    Victor Niederhoffer is famous in the commodity trading world as a bubble-denier.

  12. Gravatar of Bill Woolsey Bill Woolsey
    8. December 2010 at 08:38

    Like usual, some awful typos.

    If there is a surplus of a good, there are profit opportunities from lowering prices. If there is a surplus of labor, there are profit opportuntities from lowering wages.

    But perhaps some people consider as providing an unfair benefit to buyers. If that sort of motivation results in some prices not falling, or falling too little, what happens? Sure, it is possible that rational and greedy sellers will undercut those trying to protect buyers, and rapidly force markets to clear. What if they don’t?

    If the price of an asset is higher than the net present value of the expected future profits, then then it is sensible to sell the asset, perhaps even borrow the asset and sell it.

    But if there are some investors who believe that assets change in the future like they have in the past, and so judge a move in an asset price from too low to just right as providing a reason that it will rise more (and in fact, go from just right to too high,) their purchases will tend to raise the price.

    Sure, rational investors might sell enough to prevent the price of the asset from rising about the fundamental value, buying it back when the foolish investors see that the price doesn’t rise after all. It rose, and then stopped. Then, they sell and the smart investors buy it all back. Could be. What if it doesn’t happen?

  13. Gravatar of Philo Philo
    8. December 2010 at 08:51

    Ram writes that “we may in retrospect identify episodes during which markets failed to efficiently aggregate all (or nearly all) publicly available information.” This sounds like hindsight illusion. *Ex post* we have a vital piece of information that was not available *ex ante*: we know what actually happened. That wrongly colors our assessment of the evidence that was available *ex ante*, and so we call something a “bubble” that was actually a prediction that, though wrong, was rational.

  14. Gravatar of james in london james in london
    8. December 2010 at 08:54

    If most participants in a bubble thought it was a bubble, it would burst. Unless they expected the Greenspan/Sumner Put to bail them out and mean no losses.

    If this is your way of somehow addressing the moral hazard problem with your macro policies I don’t get it. You still need to address the issue of market participants altering their micro behaviour in expectation of macro remedies to limit or eliminate the potential losses fromtheir micro risk and leverage preferring behaviour.

  15. Gravatar of Tom P Tom P
    8. December 2010 at 09:37

    Why not define a bubble as a deviation of prices, caused by false beliefs, above their theoretical levels in a perfectly competitive, perfect information equilibrium?

    On this definition, bubbles are plausible because people don’t have perfect information, they instead have “information” that makes them look good in their own eyes. We already know that they think this way from the fact that so much speculation is going on, when under perfect information, the no-trade theorem would tell us that we should see only diversifying/life-cycle trades.

    Scott, in the past you’ve denigrated studies on “anomalies” in the finance literature (exceptions to the EMH) as data mined, but by now most of them have had relatively long out-of-sample periods. My understanding is that many of these effects have continued to work. How about that?

  16. Gravatar of Tom P Tom P
    8. December 2010 at 09:38

    I should say, note that nothing in my definition of a bubble implies that you can predict that the price will revert to its equilibrium level “soon.” It might be a long time due to these informational/expectational differences in addition to the practical difficulty of, say, shorting the stock market for a decade.

  17. Gravatar of Benjamin Cole Benjamin Cole
    8. December 2010 at 09:53

    Fascinating topic and post by Scott Sumner.

    I covered Wall Street for many years; I would say on some levels Wall Street is in the business of “selling a story.”

    Over the decades, the story can be the Internet, REITs, high-tech something, the Pacific Rim, Japan, oil scarcity, boundless China, etc.

    At various junctures, a story can become compelling, and then it is sold hard and heavy–money is made on the transactions, the M&As, the underwriting.

    The housing story of late was in large part the MBS story–you could get higher and safe yelds on MBS. Without global selling of MBS, the housing bubble might have been muted. The Niagara of capital came into the US (to quote Anthony Downs, a good intellect btw).

    So, I think there are bubbles, in some regards they are manufactured. With the advent of the web, there are zillions of commentators now who may have hidden agendas. I would say we are prone to bubbles more than ever.

    Man is not only an economic animal, he is a social animal, and thus has animus, is given to fads and excesses. If you saw the bell bottom pants I was wore…

    Well, if I am making a point with this comment I am not sure, but it always a pleasure to hear the sound of one’s voice.

  18. Gravatar of Benjamin Cole Benjamin Cole
    8. December 2010 at 09:58

    Add on:

    At one point, in the 1980s, Wall Street was successfully selling REITs for more than net asset value. The justification: If you owned a single asset, such as office building, you were stuck with that asset, and faced risks if something went wrong.
    But with a porfolio of office buildings, you could continuously cull the weaker properties, and did not face single-property risks. Also, more-professional management could be brought in, and economies of scale obtained. Hints of favorable pending legislation were tossed about.

    Soon, real estate owners were grouping their properties and selling them as portfolios on Wall Street. Buy 10 office building for $100 million; sell on Wall Street for $125 million.

    They had a story to sell.

    Later, in the 1990s, REITs started selling for less than their net asset values, and I think still do today.

    The story really well done becomes a bubble.

  19. Gravatar of Mike Sandifer Mike Sandifer
    8. December 2010 at 10:00

    Scott,

    First, I don’t think bubble concepts have to yield predictions about bubble behavior to be valid. As Bob Murphy once pointed out, we’ve long known of the existence of earthquakes and other seismic activity, yet predictions were always nearly impossible and are still difficult now, even with antecedents. Vulcanoes can smoke for many years before erupting, for example.

    Second, since bubble deniers are always perhaps prima facie in the majority, it makes sense that successful predictions wouldn’t be as laudable. And anyway, haven’t you claimed that EMH merely says that market predictions aren’t beaten consistently on average?

    Third, I would think it’s relative price volatility that matters.

    Then, you’ve made statements(I’m pretty sure these were yours) in the past which seem to contradict your current comments. What about your claims that mortgage lending was too loose in the build up to the most recent bust? How about your claim that there is at least some evidence of momentum in housing prices? And it now occurs to me that if you admit some momentum in housing prices, might that mean that there can be momentum in related securities as well, including stocks?

    Also, you’ve claimed in the past that if stronger EMH theories were right, we should expect “bubbles” now and then as a matter of chance. But, chance behavior has mechanisms. They may simply be unknown.

  20. Gravatar of Greg Ransom Greg Ransom
    8. December 2010 at 10:01

    Here’s Scott refutiating the existence of Darwin’s concept of “species”:

    “if Darwin’s term “species” is to mean anything useful, it must contain an implied prediction of the future course of speciation.”

    Well, that’s bad science, not sound argument.

    Ditto in the case at hand.

    When you have a bogus understanding of “science” at the core of your “economics” it sends out a siren call, a false signal, leading away from good science and toward bad.

    This bogus picture of economic “science” of course, comes from Friedman, as much as anyone.

  21. Gravatar of Eric Morey Eric Morey
    8. December 2010 at 10:08

    Steve Keen has indicated that the recovery in the Australian housing market was due to their First Home Owners Scheme “releveraging” the private debt market. He states that the difference between “bailing out private borrowers” and “bailing out banks” has made all of the difference.
    http://www.debtdeflation.com/blogs/2010/11/15/why-credit-money-fails/

    See slides 68 to 79 on his presentation for some relevant graphs: http://www.debtdeflation.com/blogs/wp-content/uploads/talks/KeenHowCreditWorksWhyItFails.ppt

    Slides 29 and up on this other presentation also have informative graphs:
    http://www.debtdeflation.com/blogs/wp-content/uploads/lectures/bf/KeenBehaviouralFinanceLecture12GlobalEconomicCrisisPart1.pptx

  22. Gravatar of Doug Bates Doug Bates
    8. December 2010 at 10:16

    I think most of the confusion about bubbles comes from observers overidentifying bubbles without having strict criteria. But identifying bubbles using strict criteria still won’t help you to predict when they will pop. People don’t like conclusions that we can’t predict something … but that’s life. Not everything is predictable.

  23. Gravatar of Ted Ted
    8. December 2010 at 10:23

    One could argue when bubbles will pop it is unpredictable because they are driven by sentiment.

    A few points about housing bubbles and bubble predictors:

    i) Who cares if you did predict a bubble? I suspected a bubble and I thought the market would go down in 2006-2008 because I thought house prices in 2005 just seemed absurd. But does this make me a messiah – or did I just get lucky? That’s impossible to know and I wasn’t confident enough that I could arbitrage effectively to make a lot of money off my prediction (not that I even have the money to arbitrage, but that’s beside the point). But, again, how do you know it wasn’t luck? For most people, the reason they thought we didn’t have a housing bubble was driven just as much by animal spirits and sentiment as those that did believe we had a housing bubble. There was no technical analysis done to justify their points.

    ii) You absolutely need a reasonable time-length or we can’t know if you just got lucky. Shiller was talking about a housing bubble in 1998. If you wait long enough, just based on historical trends alone, you’ll probably see a housing bubble eventually.

    iii) Lots of people had to predict the bubble, by the way, so these guys aren’t special. A housing bubble is driven by expectations of future supply-demand of housing. Supply-restrictions weren’t plausible because we have plently of undeveloped land, so people had to be expecting a jump in housing demand. Such high prices had to imply people were expected larger increases in growth and productivity than panned out. This should have also implied an increase in future consumption. Which, in turn, raises the real interest rate – which should strongly tames the housing run-up. You need a large portion of the population to not actually buy the housing bubble for this to make sense.

    iv) How do you know if something was a bubble? A bubble in asset prices is typically driven by incorrect expectations of future supply-demand for the asset. But, let’s say instead that rather than having incorrect expectations what actually happened was investors received new information about the future supply-demand of the asset. And if it’s less favorable, asset prices will fall. But this wasn’t a bubble, this was a rational response to newly available information. How could you possibly know which was which?

    In short, bubble predictions may not be impossible, but it’s impossible to verify whether the person is correct so for all relevant purposes the predictions are irrelevant.

    Finally, I want to point out one of the most interesting papers I’ve read this year. It’s by Angeletos, Lorenzoni, and Pavan (“Beauty Contests and Irrational Exuberance: A Neoclassical Approach” – see link: http://econ-www.mit.edu/files/325 ). Essentially there model is one where dispersed, private information is aggregated and the financial and real sectors of the economy look at each other to determine the profitability of investment – which causes information spillovers that can generate noise. In a neoclassical model with perfectly rational-optimizing agents, they can generate something that looks exactly like a crazy irrational exhuberance economy. How valid the model is as a description of the real world remains to be seen, but it’s interesting you can generate such an economy using perfectly plausible assumptions about heterogenous information spillovers and perfect rationality to supposedly get “irrational exhuberance” and bubbles. I’m still trying to wrap my head around the model entirely, but it seems like a promising idea.

  24. Gravatar of Bill Woolsey Bill Woolsey
    8. December 2010 at 10:25

    James in London:

    You are mistaken to assume that nominal GDP growing on target is inconsistent with the prices of various assets falling (some stocks, all stocks, some houses, all houses.) Further, your are mistaken to assume that nominal GDP growing on traget is inconsistent with failures of finanical institutions (some investment banks, all investment banks, some commercial banks, all commercial banks.) It is true that if all commercial banks failed, closed, and remained closed, it might be tough to keep nominal GDP on target, but if bankruptcy allowed for rapid reorganizations, then all commercial banks could fail. Some former depositors would become the new stockholders, and the newly capitalized banks could create all the deposits and loans needed.

    The notion that we need to have nominal GDP drop and a recession to punish people for paying too much for assets is wrong. The problem is the relative prices of assets rise, adn they “punishment” can and should be a drop in the relative prices of those assets.

  25. Gravatar of ssumner ssumner
    8. December 2010 at 10:26

    Morgan, This is ancient history, but NGDP targeting would have prevented the Great Inflation, which started in the boom of the late 1960s.

    Mr. Winston, Yes, there are lots of bubble predicters, but who are the famous bubble deniers?

    123, The Australian housing market was not a bubble in 2005, so it looks like the Aussies are pretty smart. But in the US I saw far more people predicting bubbles, than denying bubbles.

    Mattias, Yes, pessimists are more intellectually serious, but in the long run optimists have been more correct (i.e the famous Erhlich/Simons bet.)

    Dan, I completely disagree. I think bubble spotters receive great acclaim, and I think bubbles are very hard to spot.

    Ram, You might be right, but how is it that we can spot a bubble in retrospect, but not at the time. What specific information allows us to say in retrospect something was a bubble, they wouldn’t allow us to notice the same thing in real time?

    Maybe it’s a question of semantics. I’ve always assumed the EMH allows people to make mistakes, as long as the mistakes are not obvious, ex ante.

    Bill Gee, Thanks. And that’s a good analogy.

    Lee Kelly, But then why aren’t bubble deniers famous?

    jsalvatier, I don’t see those studies as showing that bubbles exist. Many studies suggest that experimental economics outcomes don’t carry over to the real world where traders are much more experienced.

    Bill, You said;

    “But logic tells us that bubbles are possible.”

    Just the opposite, logic tells me the EMH is true. If everyone knew prices were too high, they would be lower. prices are where there are because the average market participant believes those prices are appropriate. Are some traders smarter than the market? Perhaps, but how would we identify those traders.

    I don’t believe the EMH applies to labor markets, because wages are sticky. So monetary shocks have real effects.

    But asset markets generally have pretty flexible prices, so I think they are pretty efficient. Although housing prices are a bit stickier than stocks, bonds, forex, and commodities.

    rob, Fama didn’t become famous by denying a particular bubble, and being right. I don’t know the other gentleman. Which specific commodity bubble did he deny where he was proved right in the long run? (You can’t use a current commodity bubble, as it’s too soon to say.)

    Bill, You second point is an argument I often see, but don’t buy.

    First of all, it doesn’t address the argument in my post. Indeed no one has offered the sort of evidence required to refute it.

    Second, isn’t it just the old “the market can stay irrational longer than you can stay solvent” argument? Which is incorrect. I can and will stay solvent my whole live. I’d love to know which investments will do better that index stock funds over the next 30 years. If the anti-EMH view means anything, it means you can beat the market over 30 years by following an anti-EMH recipe. If it can’t do that, then the anti-EMH view is completely empty, based on a cognitive illusion.

    Philo, That was my reaction too–see my response to Ram.

    James, You said;

    “If this is your way of somehow addressing the moral hazard problem with your macro policies I don’t get it.”

    My policy recommendations reduce moral hazard, they don’t increase it. And this post has nothing to do with moral hazard, I’m trying to show that bubbles don’t exist.

    TomP, No, I’ve read that anomalies usually stop working after they are published. If they kept working then the managed mutual funds would beat index funds. Just invest in “anomaly funds.”

    You said;

    “I should say, note that nothing in my definition of a bubble implies that you can predict that the price will revert to its equilibrium level “soon.” It might be a long time due to these informational/expectational differences in addition to the practical difficulty of, say, shorting the stock market for a decade.”

    You don’t need to short markets, just move into markets that are undervalued. There are dozens of stock markets around the world, and dozens of other commodity, bond, currency markets, etc. At any given time there is always something “undervalued” by the anti-EMH view—and if not then temporarily park your money in cash until the next opportunity opens up. At least 95% of the time at least one market will be undervalued, usually many markets.

    I claim this strategy won’t work, even though the anti-EMH view says it will. Anti-EMH types should have been screaming for people to buy stocks in the spring of 2009, but all I recall is Roubini saying don’t buy stocks.

    Benjamin. How do stories get manufactured? My hunch is that people actually believe these stories.

  26. Gravatar of ssumner ssumner
    8. December 2010 at 10:43

    Mike, The earthquake example is no good; if financial crises are like earthquakes (unpredictable), then the EMH is right.

    Second, I don’t think bubble deniers are in that much of a majority. In 2005 most people I talked to thought housing was a bubble. And in any case if bubbles do exist, then the bubble deniers are making the more difficult prediction, and deserve greater praise.

    My criticism of mortgage lending had to do with moral hazard, not the EMH.

    Housing has bigger transactions costs than other assets, which makes the housing market slightly less efficient. But not inefficient enough to produce such a big bubble.

    Greg, Your analogy is completely inappropriate. It is the bubble deniers who claim to be able to predict the future. When stocks are claimed to be a bubble, they suggest the expected return from stock investments (long term) is lower than when stocks are undervalued. They say “don’t buy.” Can you imagine a Shiller saying “stocks are a wildly inflated bubble right now, go out and buy.” He’d be laughed at. If markets are like evolution (unforcastable), then the EMH is true.

    Doug, I agree.

    Ted, You said;

    “ii) You absolutely need a reasonable time-length or we can’t know if you just got lucky. Shiller was talking about a housing bubble in 1998. If you wait long enough, just based on historical trends alone, you’ll probably see a housing bubble eventually.”

    Wow! Can someone find a link. If so, the most famous bubble advocate was wrong about the stock bubble in 1996, and also wrong about a housing bubble in 1998. That’s 0 for 2, and he’s considered a successful forecaster!

    In general your ideas seem very plausible. I’ll try to take a look at that paper.

  27. Gravatar of ssumner ssumner
    8. December 2010 at 10:47

    Eric, Thanks for the info on Australia.

  28. Gravatar of Benjamin Cole Benjamin Cole
    8. December 2010 at 10:47

    Benjamin. How do stories get manufactured? My hunch is that people actually believe these stories.

    Oh yes, people believe the story–there is a context whihc makes a story compelling.

    In the case of MBS, the underwriters had fancy models showing risks of default low, and the ratings agencies went along, rating MBS AAA, and also creating tranches into which the so-called riskier elements were dumped, so that an issue could have AAA tranches and down the line to BB, creating an illusion of safety or perceived risk.

    The “story: workjed, globa, investors hunting yield poured money into MBS. On the frontend you had Countrywide et al handing out the money.

    Homebuyers had watched housing prices rise for years. They could buy with little down. They could walk away if all went wrong–these are non-recourse loans.

    A bubble was born!

  29. Gravatar of gnikivar gnikivar
    8. December 2010 at 10:57

    You know, I’m not convinced there ever was much of a bubble and the real prices of houses did double. A housing bubble would have only been a collapse if you could have predicted the collapse from the beginning. For the first three quarters of the decade people expected NGDP growth around 5 percent. Then, NGDP collapsed, and the housing bubble burst. All the countries where NGDP has held up strong like Australia haven’t seen much NGDP decline. I think the housing collapse only happened because of NGDP mismanagement.

  30. Gravatar of Floccina Floccina
    8. December 2010 at 11:52

    I say if the price an asset goes high enough that the long term yield is likely to be significantly below 3% you are looking at a bubble.

  31. Gravatar of Mike Sandifer Mike Sandifer
    8. December 2010 at 11:55

    Scott,

    You replied: “Mike, The earthquake example is no good; if financial crises are like earthquakes (unpredictable), then the EMH is right.”

    But, you’re expressing skepticism about the very existence of bubbles as you define them, which is unwarranted.

    I don’t see why bubbles can’t exist, but be unpredictable, and hence EMH holding up. I also fail to see why bubbles can’t be predictable, but unexploitable, perhaps due to widespread cognitive biases, and thus EMH holds up. Am I correct in stating that EMH doesn’t require markets to be optimal, even if suboptimality is based on widespread information that’s misinterpreted? If I’m not, what difference does it really make?

    And earthquakes may be predictable in principle, with sufficient models and metrics. Maybe one day we’ll be able to predict bubbles and then prevent them with the mere beliefs in the relevant predictors. Or, maybe we’re hopelessly stuck with them until human nature changes in some fundamental way.

    “Second, I don’t think bubble deniers are in that much of a majority. In 2005 most people I talked to thought housing was a bubble. And in any case if bubbles do exist, then the bubble deniers are making the more difficult prediction, and deserve greater praise.”

    In the discussion we had over whether free market decisions that allowed smoking in restaurants or later referenda disallowing it better reflected market opinion, you argued for the former. In the case of bubbles, most of the money seems to be betting on sustainability on the way up in any one moment, discounting the casual or merely academic opinions you refer to which apparently are mere anecdotes anyway.

  32. Gravatar of D. Watson D. Watson
    8. December 2010 at 12:04

    Scott,

    There a few things I think are missing or over stated.

    I’ll take issue with using 50/50 as the cutoff for determining bubble-predicting success. It’s not just about “stocks going down.” In any given day, even a soaring stock goes down multiple times. For something to be a popped bubble, it has to lose a significant portion of its value while other stocks are not going down. That happens a lot more rarely. You don’t have to get more than half of the bubbles right, or even have more than half of your predictions be right. If bubbles happened to 1/100 stocks (boy, that’s a lot of stocks!) and you got it right twice out of 100 guesses, you’d be doing better than the market. The real trick is in knowing when to jump ship.

    You almost acknowledge this when you say that bubble deniers have “the more difficult prediction.”

    Another issue: you say “If everyone knew prices were too high, prices would be lower.” Two biggest words in that sentence: Everyone; Knew. It’s a probability, people have different subjective probabilities of where things are going. If enough people believe it is likely enough that prices are too high, prices would be lower … but how do we know they aren’t already?

    I know that stock zzyzx is overvalued and it will crash someday, but I know there are more suckers I can sell my stock to tomorrow. Those suckers have to fit into the EMH too.

    And that’s the final thing I see missing: Vernon Smith. The first time traders go into the experimental setting, they make stock bubbles EVERY time. If they come back another day for a second round, there may be a small bubble, but it rarely lasts and it’s seldom shirt-losing. If they come back for a third day, no bubbles.

    I hope those disjointed thoughts make sense.

  33. Gravatar of Mike Sandifer Mike Sandifer
    8. December 2010 at 12:05

    Scott,

    You replied to Ram: “Ram, You might be right, but how is it that we can spot a bubble in retrospect, but not at the time. What specific information allows us to say in retrospect something was a bubble, they wouldn’t allow us to notice the same thing in real time?”

    I’m reminded of cellular automata and similar phenomena. You maybe be pretty sure that certain rules will produce certain shapes, but can’t predict ahead of time when and where they’ll occur. Obviously, you can see them after the fact.

  34. Gravatar of Luis H Arroyo Luis H Arroyo
    8. December 2010 at 12:13

    Good, Scott.
    BTW, I send you a graph perhaps of your interest. I think it is conected to MP. It represent the national bonds yields of euro and non euro countries. Euro countries have to pay much more interest than the others, in spite of having similar fiscal problems.
    I´m Sorry very much, I know it is not today´s topic, but I thing it is elocuent

    http://2.bp.blogspot.com/_UlqNAo7QxaA/TP_frnUiBfI/AAAAAAAABVE/ZKHnVsp45y4/s1600/BY.jpg
    (Arrows are the media of each area)

  35. Gravatar of Winton Bates Winton Bates
    8. December 2010 at 12:20

    It seems to me that those who argue that central banks should be trying to bust bubbles may have a point if they are looking at asset prices as an indicator of market expectations about future growth of NGDP.

  36. Gravatar of Blackadder Blackadder
    8. December 2010 at 12:50

    Mike Sandifer,

    You say: “I don’t see why bubbles can’t exist, but be unpredictable, and hence EMH holding up.” If all that is required for there to be a bubble is that prices go down then obviously there are bubbles, as prices do go down from time to time. When people talk about bubbles they typically mean something more than this. There is an implied claim that the market is somehow acting irrationally, etc. If aliens showed up tomorrow and gave us cold fusion tech the price of oil would drop, but no one would say that oil was a bubble, because obviously there was no way to know that the aliens were going to show up.

    BTW, my impression is that people who said oil was a bubble in early 2008 haven’t been lionized to nearly the same degree as those who predicted a fall in the price of housing or tech stocks. This suggests that pessimism is really the driving force here. In the case of oil bubble deniers are the pessimists, so they get the glory, whereas with housing bubble deniers are considered optimists and so are scorned.

  37. Gravatar of Mike Sandifer Mike Sandifer
    8. December 2010 at 12:53

    Blackadder,

    I was implicitly using what I think is Scott’s definition of a bubble.

  38. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    8. December 2010 at 12:53

    Arnold Kling replies:
    http://econlog.econlib.org/archives/2010/12/what_is_a_bubbl.html

  39. Gravatar of Floccina Floccina
    8. December 2010 at 12:58

    BTW Warren Buffet and Jim Rogers played the bubbles correctly. They sold stocks in the late 1990s and Buffet, unlike Rubini, bought stocks in 2008 and 2009. Rogers bought commodities in 1998 and is still holding them.

  40. Gravatar of marcus nunes marcus nunes
    8. December 2010 at 12:58

    Bernanke is the “bubble”! Soon, one fun too many will “poke it”.
    http://blogs.wsj.com/economics/2010/12/08/the-daily-show-takes-on-bernanke-and-printing-money/?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+wsj%2Feconomics%2Ffeed+%28WSJ.com%3A+Real+Time+Economics+Blog%29

  41. Gravatar of David N David N
    8. December 2010 at 13:13

    Scott,

    What level of EMH are we discussing here? Is the present market value always the best estimate of value, or is it only almost always the best estimate of value? The latter, of course, opens the door for bubbles.

    Last time around I argued that a bubble requires a new story; a “paradigm shift” to explain why existing valuation methods no longer work. Pets.com is worth $1B on < $1M sales? Sure, it's a new era, etc, etc.

    So yes, I think the market can be wrong, even if it almost always isn't. Therefore, bubbles are possible. And if you can spot the new valuation story being touted and see the holes in it, you can recognize the bubble, even if you can't predict exactly when it will pop. I guess you will say that makes the bubble concept meaningless but I think you're holding it to an impossibly high standard.

  42. Gravatar of Ted Ted
    8. December 2010 at 13:18

    Robert Shiller was actually talking about a regional real estate bubbles in 2000/2001, claiming it started in 1998 – I misspoke and said he predicted it was happening in 1998. He didn’t get apocalyptic until 2004. I can’t find links on this since a lot of his 2000-2002 op-eds are pretty old on the matter and so I can’t find them on google, but I’ll try more tomorrow when I have time and post them.

    By the way, he also predicted a world bond-bubble in 2003 (“Will The Bond Bubble Burst?” http://www.project-syndicate.org/commentary/shiller2/English ).

    Also, I’m going to go further. Shiller was wrong about the housing bubble anyway. Housing prices, right now, are about as high as they were in 2003-2004. So, let me get this straight. We have a supposedly massive housing bubble. Then we see a huge run-up in energy prices. And then the Federal Reserve goes an induces a massive nominal shock causing a huge fall in NGDP and employment. And yet, housing prices are at their 2004 levels ?? To me, if we were really in such a massive bubble with those two factors housing prices should have fallen even further. This actually suggests, to me, housing prices are actually much more reasonable than people think they were!

  43. Gravatar of marcus nunes marcus nunes
    8. December 2010 at 13:24

    Scott
    This link is the most complete I could find. It correctly dates Shiller´s and J Campbell´s stock (in general not the Nasdaq in particular) “bubble” to their Fed presentation on Dec 2 1996, which was followed by Greenspan´s Dec 5 speech where he introduced “irrational exuberance”, but dates his mention of a house bubble as 2005. No mention of anything in 1998. But maybe there is something around 1998 because I clearly remember an american client of mine telling me in 1998 that his beach house in “The Carolinás” had trippled in value in a short time period!
    http://www.yalealumnimagazine.com/issues/2009_09/shiller032.html

  44. Gravatar of Tom P Tom P
    8. December 2010 at 13:28

    Scott, re: market anomalies, why not check out Ken French’s Data Library. He has returns for strategies like “sell stocks that went up in the last month” (“STREV”). This strategy has been out of sample for a while and seems to work well. We could take bets on its future performance if you like.

    Moreover, the fact that mutual funds have poor returns in equilibrium doesn’t mean that it is impossible to generate good returns. It could just be that producing low quality services is the market equilibrium when your customers (i.e. investors) seem willing to pay a lot of money for nothing.

    Even if fund managers did want to take advantage of anomalies, mutual funds may not be able to trade them because of low liquidity. There is a trade-off between size of the fund (which is the main source of profit) and its ability to take advantage of an anomaly. Fund managers maximize their personal compensation, not fund returns.

    Finally, some actively managed funds do beat the market. As Fama/French say on their blog:

    “[W]hen returns are measured before the fund expenses borne by investors, we find evidence that there are some true winners in the population of fund managers.”

    So some managers can predict the stock market. They simply charge too much for it. This point may not have huge economic magnitude, but it is still there.

    You also argued that, even if it were difficult to short bubbles, as an anti-EMH person you still could make money by buying undervalued stocks. This may be true, but my point was that bubbles can persist for a long time because no one is going to short them. You don’t seem to think so.

  45. Gravatar of Lorenzo from Oz Lorenzo from Oz
    8. December 2010 at 13:28

    Scott: This appears to be a link on Schiller and bubble predictions.

    123: Kling’s definition of a bubble:
    a bubble is characterized by unreasonably high expectations for continued appreciation.
    is really irritating. What is “unreasonably high”? And if he has a definition for that, why is it not in his definition?

    He sort of “cashes it out” later with:
    when people expect an asset price to rise for a long period of time at a rate that is much higher than the long-term nominal interest rate, we are going to see a bubble in that asset. There will be strong demand for that asset until expectations change. But not really.

    If you are going to try and define a bubble, this one looks better:
    An economic bubble exists whenever the price of an asset that may be freely exchanged in a well-established market first soars then plummets over a sustained period of time at rates that are decoupled from the rate of growth of the income that might be realized from owning or holding the asset. The question becomes, do people have expectations about capital gain which are not based in expectations about income from the asset?

    If all that is changing is expectations about income, that is just markets acting as markets in a fairly straightforward sense. It is where there is evidence of people having expectations about capital gain not based in expectations about income-from-the-asset that things become interesting. Why would they have such expectations? And would not such expectations be unstable in the sense of having a “tipping point” that, when such expectations collapse, would collapse prices back to those based on income-from-asset expectations? These seem to me to be the interesting questions.

    I particularly get unimpressed by folk who talk about “speculation”. Speculation is ubiquitous, it only becomes interesting when the speculation is overwhelmingly in a specific direction. In the case of suggested bubbles, when people have reason to discount downside risk. Such as, to take the Australian housing case, supply constraints (since we have adopted the British/Californian model of official land use controls) mean supply cannot fully respond to demand, which is then reflected in a continuing trend of higher prices. This has set up an expectation of continuing capital gains. It is all very well to say that such expectations can overshoot, but since it is not clear what before-the-fact information one can point to that will identify what will trigger the change in expectations, it does not have much predictive value. (Expectations about expectations: there’s an analytical reductio for you.)

  46. Gravatar of Greg Ransom Greg Ransom
    8. December 2010 at 14:01

    The business editor of the Orange County register also made this call, about the local housing market.

    _The Big Short_ is has the names of others who made the bet and earned millions, some of them billions.

    Scott wrote:

    “of course Krugman was right in predicting that real US housing prices would fall in the 5 years after 2005, as was Dean Baker, Nouriel Roubini, and some others who became well known and lauded for their predictions.”

  47. Gravatar of james in london james in london
    8. December 2010 at 14:07

    Bill/Scott
    I think you both still fail to understand the way the NGDP Put (or whatever the Greenspan Put morphs into) will be gamed. The banksters are TBTF and Too Big To Bail, they like it like that and plan for that. And authorities who don’t understand that or worse have beeen captured by the banksters, like it seems in the US, will be doomed to repeat the turmoil of the last few years. Some other countries are far less trusting of their big banks and seem to “get it”. In the US the banksters love the Fed and its magic wand to cure all bubbles and bursts. You still need to understand how micro actions can cause havoc depending on expectations of macro policy. Moral hazard remains a real challenge for your NGDP targeting.

  48. Gravatar of Johann Johann
    8. December 2010 at 14:57

    Kevin Murphy, Gary Becker, Casey Mulligan, Eugene Fama and I believe (but am not sure) John Cochrane at the University of Chicago.

    The first 3 give particularly convincing evidence.

  49. Gravatar of OneEyedMan OneEyedMan
    8. December 2010 at 14:59

    @Bill Woolsey

    Rational asset pricing bubbles is a paper by Santos and Woodford in Econometrica that actually claims that monetary general equalibria are bubbles.

    Money is actually something we might think would have bubbles because it has no intrinsic, it only has value to us as a store of value and medium of exchange. Therefore it only holds value when we believe others will take it. That’s one of those things that is true until it isn’t, usually intimately tied to the health of the state issuing it.

  50. Gravatar of Robert Waldmann Robert Waldmann
    8. December 2010 at 15:34

    Ah yes cognitive illusions explain the so called price earnings anomaly such that stocks with low price earnings ratios out perform stocks with high price earnings ratios.

    This is not a new claim. The dating of the claim depends on whether one refers to the raw price earnings anomaly or the beta adjusted price earnings anomaly. This is inevitable since the alleged price earnings anomaly is older than the CAPM and beta was just a greek letter when it was first asserted.

    So, as is typical of patterns which we see even though they aren’t there, the anomaly fits data before it was described and fails to yield superior forecasts. Except that it did yield superior forecasts. For decades after it was described it outperformed the CAPM (which isn’t saying much).

    I think you are completey ignorant about practically everything. As far as I can tell, your views on economics are a pure expression of prejudice uncontaminated by any knowledge of more than a few hundred data. I assume you are familiar with US GDP growth rates and inflation rates for a few years and you might even know unemployment rates for all I know.

    You are, however, obviously totally unqualified to express any opinion on the efficient markets hypothesis since your survey of what you think the debate to be shows you are totally clueless.

  51. Gravatar of OneEyedMan OneEyedMan
    8. December 2010 at 15:50

    @Robert Waldmann
    The superiority of the Fama French factors over the CAPM isn’t an indictment of the efficient market hypothesis so much as the simply showing that the CAPM wasn’t a very good model of stock market returns.

    We really think that some sort of C-CAPM should be holding where assets that have high returns when the marginal utility of income is high are the most valuable. The Fama French factors (of which the book to market factor is one) can hold with an efficient market simply by representing correlation with income not explained by the broad stock market.

    Plain vanilla CAPM typically is only holds when everyone can invest in the same assets. The presence of non-diversifiable assets like housing or human capital (is one of many reasons) it does not hold.

    Portfolios that outperform on average are not challenges to the efficient market hypothesis if that out-performance depends on taking risks that other investors are not. True arbitrage would be required, which involves some sort of first order stochastic dominance of the anomaly portfolio over the standard one.

  52. Gravatar of jason11 jason11
    8. December 2010 at 16:08

    Bubbles cannot exist because predictions of the future are not exact enough. Gosh, what a disastrous piece of logic’ that is.EMH like any good mythology was never based on facts or data or even reality……so those things pose no threat to the truly devout. The existence of consistent 5 year planning models that accurately show price movements 5 years out would be a good argument for EMH But somehow the lack of this also proves EMH. brilliant

  53. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    8. December 2010 at 16:13

    ‘I think you are completey ignorant about practically everything. As far as I can tell, your views on economics are a pure expression of prejudice uncontaminated by any knowledge of more than a few hundred data.’

    Maybe you can prevail on DeLong to figure out a way to surreptitiously delete Scott’s comments here too.

  54. Gravatar of Doc Merlin Doc Merlin
    8. December 2010 at 16:37

    “More and more I think that the entire bubble/anti-EMH approach to economics is founded on nothing more than superstition”

    Bah, I have said it so many times in this blog. Bubbles are NOT anti-EMH, you can have an EMH-consistent bubble theory.

  55. Gravatar of Greg Ransom Greg Ransom
    8. December 2010 at 16:53

    EMH & RE economics derived not from an understanding of real people in the real world — but from efforts to force economic phenomena into a statistical “scientific” program and into a fictional and dramatically artifical “model” of an “economy” that has very little in common with the real world.

    So in the first instance they aren’t recommended by the real world, they are recommended by “scientific” programs which require proving themselves on independent grounds, which for many of us begs the question.

    Darwinian biology can give us grounds for anticipating future events, but it doesn’t give us grounds for quantifiable predictions — the problem is an insuperable limit onmour knowledge of initial conditions, a knowledge problem directly comparable to the knowledge pronnlem in ecomomics.

  56. Gravatar of David K David K
    8. December 2010 at 16:55

    @OneEyedMan:
    “The Fama French factors (of which the book to market factor is one) can hold with an efficient market simply by representing correlation with income not explained by the broad stock market.”

    This is a contradiction. If there is correlation to income, not explained by the broad market, that is nonetheless completely predictable and broadly available, the market is definitionally not efficient according to the standard definition.

    It is true that the CAPM doesn’t correctly handle non-diversifiable assets but that’s a red herring. It doesn’t explain why BP (naive, simple BP, let alone better valuation metrics!) is good at picking which large-cap stocks with millions of shares traded a day to invest in within a given industry. Which it emphatically does.

  57. Gravatar of Tom P Tom P
    8. December 2010 at 17:22

    @OneEyedMan

    Let’s leave the Fama-French anomalies aside (as you point out, some of them may have explanations consistent with EMH, although I’m personally skeptical).

    How about the sustainable cash flows anomaly of Sloan (1996)? Companies that get more of their earnings from accounting juggling rather than hard cash tend to underperform. There is a plausible economic explanation: people just look at the earnings number without digging into how that number is derived. And the effect has persisted since then.

  58. Gravatar of Tom P Tom P
    8. December 2010 at 17:29

    @OneEyedMan (post at 15:50) — I totally agree that any so-called anomaly can be explained by the risk involved in investing in that anomaly.

    I just find it difficult to come up with plausible risks associated with certain anomalies. Tetlock found an anomaly that shows that stocks that have positive news written about them tend to go up in the short term. What risk is associated with that (beyond market risk, which CAPM controls for)? Don’t we have to give some weight to the much dumber idea that maybe people take at least a couple days to trade on new news and so the news is not priced in *immediately*?

    I’m not saying the market is totally inefficient or anything, in fact I think it is quite efficient. But I just want to introduce some evidence of predictability in stock prices.

  59. Gravatar of Noah Noah
    8. December 2010 at 18:52

    Dr. Sumner –

    I refer you to the experimental work of Smith, Suchanek, and Williams:
    http://www.jstor.org/pss/1911361

    Bubbles (prices far exceeding known fundamental values, accompanied by trading at high volumes) are an extremely robust feature in laboratory markets.

    So much for the EMH.

  60. Gravatar of OneEyedMan OneEyedMan
    8. December 2010 at 19:23

    @David K
    I’m not familiar with BP. Can you give a more complete citation?

    @Tom P
    My impression is that there is a huge publication bias in the anomaly literature. I believe that many of them are false positives, and the stranger the anomaly is the more likely I am to believe that it is just a false positive on a statistical significance test.

    I can appreciate that sounds like a cop-out. But since so many anomalies have vanished when longer series become available after identification, there just isn’t much to excite me in the anomaly literature.

    It is possible that some sort of ultra-weak EMH holds where it takes time for information to defuse. I could be convinced but in today’s markets for equities things are more voluminous, tighter spreaded, and faster than before so I’m skeptical.

  61. Gravatar of scott sumner scott sumner
    8. December 2010 at 19:47

    gnikivar, Wow! You are even more extreme than me. I’ve made that argument many times, but only blame a portion of the collapse on NGDP.

    Floccina, That would make cash a bubble. And who knows what the expected return on housing was?

    Mike Sandifer, I can’t tell if we disagree over anything substantive, or are just arguing over semantics. Unless you can predict which way asset prices are likely to go, I say there is no bubble by definition. Obviously the perfect foresight price would be different, even EMH proponents accept that.

    D. Watson, You said;

    “I hope those disjointed thoughts make sense.”

    I’m afraid not, but it’s probably my fault. It’s like when I argue with someone over free will. The argument never gets anywhere because we see the same picture, but put different labels on it. I see mistakes in the housing market and call it “efficient.” Others see mistakes in the housing market and call it “inefficient.” So what is the debate about? That’s where I get back to the original point I made long ago. This debate shouldn’t be about the EMH, it should be about the anti-EMH theory. The anti-EMH view performed horribly in the housing bubble. Investors and regulators were supposed to do policies that reflected the market irrationality. But they didn’t. Hence the anti-EMH is like the emperor with no clothes. We keep being told markets are irrational, but no one seems to be able to do anything useful with the anti-EMH view. Even Roubini failed to sell his stocks before the crash. If the anti-EMH view is useless for Roubini, how could it be useful for anyone else? He’s the guy who saw the crash coming more clearly than anyone else.

    Mike, I don’t know enough about cellular automata to comment.

    Luis, Thanks, the euro is a big problem for many countries right now.

    Winton, That’s exactly my view. But if that’s what we are doing, then we DESPERATELY need an NGDP futures market.

    Blackadder, Interesting point. The oil bubble seemed to last a shorter time, so there was less attention. And also there were all these conspiracy theories kicked around for oil, so the debate was about manipulation. Since houses are sold from one average guy to another, there are no manipulation theories for housing bubbles. I don’t know how much that explains, but I just throw the idea out there.

    123, Thanks, I’ll take a look.

    Floccina, You said;

    “BTW Warren Buffet and Jim Rogers played the bubbles correctly.”

    Buffett did very poorly, lost something like 25% in 2008. I probably did better than Buffett in 2008.

    Marcus, Thanks, That’s very funny.

    David, I’m not focusing on the EMH, but rather on the anti-EMH. I claim the anti-EMH is a useless theory, with no implications for policy.

    Ted. So Shiller was wrong about the bond bubble as well? Aren’t bond prices higher than 2003?

    Marcus, Thanks for the info on Shiller.

    Tom, Fama believes in the EMH, so I’d say his empirical evidence is at best ambiguous. I assume Fama has the same view I do, that the EMH is approximately true, and the anti-EMH does not have useful implications.

    I don’t buy the argument that mutual fund managers don’t have an incentive to do well. I recall that Fidelity made a lot of money when its Magellan fund had a good track record, and drew lots of investors.

    Lorenzo, I agree with your view on bubbles.

    More to come. . . .

  62. Gravatar of scott sumner scott sumner
    8. December 2010 at 20:27

    Greg, And how does the Orange county newspaper relate to this post?

    James, The Fed did not prevent NGDP from falling, so your argument makes no sense.

    Johann, Evidence on what?

    Robert Waldmann, You said;

    “I think you are completey ignorant about practically everything. As far as I can tell, your views on economics are a pure expression of prejudice uncontaminated by any knowledge of more than a few hundred data. I assume you are familiar with US GDP growth rates and inflation rates for a few years and you might even know unemployment rates for all I know.
    You are, however, obviously totally unqualified to express any opinion on the efficient markets hypothesis since your survey of what you think the debate to be shows you are totally clueless.”

    I may be clueless, but I do know something many anti-EMH types seem not to know. For ever 1 billion patterns of data, the EMH predicts 50 million statistically significant anomalies. So I am unimpressed by anomalies. Elsewhere I’ve argued that the search for anomalies is the wrong way to test the EMH.

    jason11, I can’t follow your comment.

    Doc Merlin, I don’t see how the two can be reconciled

    Greg, I don’t see the relevance of evolution. I’m a pragmatist. I find Ratex and the EMH to be very useful in my empirical work. The anti-EMH view? I find it of no value. How can I benefit from it? Will it tell me how to vote? How to invest? Will it tell the Fed how to run policy?

    Noah, I already dealt with that in the earlier comments. And the findings are not extremely robust. In some studies the effect wears off after just three rounds. In any case, experimental economics often produces results that don’t hold up in the real world.

    Everyone, This exercise has served to confirm my skepticism about the EMH. I expected specific examples of people who became famous from specific predictions made denying bubbles. None were produced. Some people seem to get unhinged debating what is basically a dry technical topic.

    I’m guessing if there really is someone smart enough to beat the markets, they aren’t spending time sending me hate mail, but are soaking up the good life on a yacht somewhere.

  63. Gravatar of Morgan Warstler Morgan Warstler
    8. December 2010 at 20:57

    Scott, I’d prefer if you told me when exactly between 2004-2008, NGDP level targeting would have departed from Greenspan’s decisions.

    You have something more recent than 1960 or it’s not a theory.

    —-

    I answered th question in the first post, and you respond with:

    “I’m guessing if there really is someone smart enough to beat the markets, they aren’t spending time sending me hate mail, but are soaking up the good life on a yacht somewhere.”

    Look, Cuban entertained countless offers to acquire Broadcast, each time was a unique trading opportunity where he traded correctly by not trading, until it was perfect.

    And then right when the right when the emperor had no clothes, he not only said it, he went borrowing cash to bet against his acquiring company.

    Now, you can complain that someone who beats the market, can’t beat it again, but what it takes to do that is a lot of inside knowledge baseball, that some column writing economist would only have if he wasn’t a column writing economist.

    Google Vivi Nevo. Or look up and number of serious hedge fund guys.

    The really smart guys don’t talk – they bet their own cash and keep their mouth shut… why would you not factor that into your equation? Just because you want there to be no one smarter than you?

  64. Gravatar of Greg Ransom Greg Ransom
    8. December 2010 at 21:26

    The So Cal housing market has gone through 2 very clear and very pronounced bubble and crash cycles in the last 25 years.

    Business reporters at the OC Register has very special local understanding of the characteristics of this phenomena — local understanding and experience is part of the stuff of rivalrous entrepreneurial judgment.

    All part the the substantive contrast between the real empirical world of economic phenomena — and the fake world of the macroeconomist’s “hypotheses” and “models”.

  65. Gravatar of Greg Ransom Greg Ransom
    8. December 2010 at 21:29

    By the lights of your “pragmatism” test, Darwin means nothing to you, evidently.

    I have no interest in helping you earn money.

    I am interested in explaining economic patterns, in the same what that Darwin was interest in explaining biological patterns.

    Perhaps we don’t share the same interest.

    Scott writes.

    “Greg, I don’t see the relevance of evolution. I’m a pragmatist. I find Ratex and the EMH to be very useful in my empirical work. The anti-EMH view? I find it of no value. How can I benefit from it? Will it tell me how to vote? How to invest? Will it tell the Fed how to run policy?”

  66. Gravatar of Greg Ransom Greg Ransom
    8. December 2010 at 21:34

    Explain:

    “I find Ratex and the EMH to be very useful in my empirical work.”

    Ratex and EMH usually are used as assumptions to make “models” work, or to explain the relation of investors using statistical market data to “predict” a future stream of statistical patterns.

  67. Gravatar of Tom P Tom P
    8. December 2010 at 21:37

    Scott, the reason no one gets famous for denying a bubble is that no one gives you credit for following the herd.

    Regarding anomalies, again you are claiming that they are data mined, yet my evidence (I think) shows they have held up out of sample. I would be happy to bet on the short-term reversal factor if you are interested. Why not bet?

  68. Gravatar of TGGP TGGP
    8. December 2010 at 21:41

    Falkenstein says Shiller did not actually cry “bubble” on housing prices in 2005.

  69. Gravatar of Edwin A Edwin A
    8. December 2010 at 21:44

    David Lereah, former chief economist for the National Association of Realtors, was a prominent bubble denier. This book of his was published in 2006. Poor guy

    http://www.amazon.com/Real-Estate-Boom-Will-Bust/dp/0385514352/ref=sr_1_9?ie=UTF8&s=books&qid=1291872411&sr=8-9

    He was extensively cited as an housing expert at the Washington Post before the bubble popped.

  70. Gravatar of Didier Sornette Didier Sornette
    9. December 2010 at 04:56

    To add to the debate, let me point out to our research papers that identified correctly not only the presence but the timing of the reversal of the UK [1] and of the US [1] real estate markets [note that the papers were available about a year before the turning point in each case, from the international arXive of scientific papers: http://arxiv.org/. See also [3] for a recent assessment of two other real time predictions on the chinese markets. The website http://www.er.ethz.ch/publications/finance/bubbles_empirical provides access to the documents on many other cases.

    [1] Wei-Xing Zhou and Didier Sornette,
    2000-2003 Real Estate Bubble in the UK but not in the USA,
    Physica A 329 (2003) 249-263
    (http://arXiv.org/abs/physics/0303028)

    [2] W.-X. Zhou and D. Sornette
    Is There a Real-Estate Bubble in the US?
    Physica A 361, 297-308 (2006)
    (http://arxiv.org/abs/physics/0506027)

    [3] Zhi-Qiang Jiang, Wei-Xing Zhou, Didier Sornette, Ryan Woodard, Ken Bastiaensen, Peter Cauwels,
    Bubble Diagnosis and Prediction of the 2005-2007 and 2008-2009 Chinese stock market bubbles,
    Journal of Economic Behavior and Organization 74, 149-162 (2010)
    (http://arxiv.org/abs/0909.1007)

  71. Gravatar of James in London James in London
    9. December 2010 at 06:59

    “James, The Fed did not prevent NGDP from falling, so your argument makes no sense.”
    My argument was that the banksters would be bailed out through an attempt to prevent NGDP falling. Whether NGDP was prevented from falling was not really an issue for the banksters, they don’t care much about non-banksters and their problems. Whether the Fed/Greenspan/Sumner Put encourages banksters and shadow banksters to overleverage is the question, that I am still waiting for you to address.

    I understand Bill Woolsey’s comment that the Put may end up bailing out the banksters, but this is a price worth paying. I think it is a price that get’s higher and higher each time, and that a better solution would be for the wage earners to “expect” that their stickiness will hurt them in the long run via inflation and real wage reductions and general economic inefficiency and clear favouritism for banksters. Better to have innovative market solutions to the perceived liquidity trap, like German short-time working, than macromancy solutions that usually throw up all sorts of unintended consequences.

    The link to the NY Fed piece on Shadow Banking is here:
    http://www.ny.frb.org/research/staff_reports/sr458.pdf

  72. Gravatar of Doc Merlin Doc Merlin
    9. December 2010 at 07:56

    @OneEyedMan

    This isn’t true because of tax structure. We have to pay our taxes in currency, which gives fiat currency a sort of built in demand.

    @Didier Sornette

    Its no use, I’ve linked to your papers here several times. Btw, the name “dragon kings” has got to be the best name for any statistical phenomena I have run into.

  73. Gravatar of Mike Sandifer Mike Sandifer
    9. December 2010 at 08:17

    Scott, you replied:

    “Mike Sandifer, I can’t tell if we disagree over anything substantive, or are just arguing over semantics. Unless you can predict which way asset prices are likely to go, I say there is no bubble by definition. Obviously the perfect foresight price would be different, even EMH proponents accept that.”

    I don’t think one needs to be be able to predict the direction of asset prices for bubbles to exist. I think I’m using the word the way you are. There can be good reason in general to believe a phenomenon can be detected, but unpredictable. With bubbles in particular, we just may not be able to understand how to even detect them reliably, and usually only see them in retrospect.

    In the case of cellular automata, we can even know very simple rules that decide how different examples will evolve, but be very incapable of predicting the shape of the outcome. Essentially, you only know after the fact. Here is a very brief intro:

    http://mathworld.wolfram.com/CellularAutomaton.html

    I also don’t think the existence of bubbles, whatever the cause, necessarily has anything to do with whether semi-strong EMH is true.

    You’ve shaped my opinions about bubbles and EMH more than anyone, and looking your at references to Fama and Cochrane helped strengthen them. Scientifically, based on my perhaps extremely limited knowledge on these topics, I think one has to be agnostic.

    But, then there’s my tendency to think semi-strong EMH is too strong and that bubbles do exist at times. Anecdotally, I was able to spot stocks I thought were undervalued for nearly 7 years and very rarely missed. I suffered big losses in late 2007 though, which I think reflects an almost total lack of understanding of macroeconomics. I recall you saying that the tech boom and bust was suspicious in your view.

    When it comes to your ultimate call for examples of people who’ve famously correctly denied the existence of specific bubbles, deniers legion and unremarkable, again due to people voting with their money.

    That being said, there are those who invest in bubble assets even when they think they’re bubbles, like Soros claims to, so maybe the existence of bubbles doesn’t quite so strongly indicate a lack of bubble belief by participants.

    So primarily, while I can’t say whether bubbles exist scientifically, I don’t except your framework for skepticism.

  74. Gravatar of Greg Ransom Greg Ransom
    9. December 2010 at 08:29

    The difference between those who talk of bubbles / malinvestment and those who talk of EMH is pretty basic.

    The EMH was invented to account for the problem of taking a simple stream of numbers in a very limited data set, and then using it to extrapolate — “predict” — a future stream of data points, most importantly including the timing of the unfolding of that pattern.

    This is a completely different from using a structural / causal understanding of the economy, your local knowledge, and your background understanding to explain malinvestment “bubbles”.

    Comparing the two is a scientific / explanatory mistake of the first order.

  75. Gravatar of Greg Ransom Greg Ransom
    9. December 2010 at 08:31

    “Unless you can predict which way speciation is likely to go, I say there is no species by definition.”

    That’s the logic of your argument, Scott. And its bad science.

  76. Gravatar of scott sumner scott sumner
    9. December 2010 at 08:31

    Morgan, Policy might have been a bit tighter in 2004-06, although it’s hard to say. Policy would have been much easier in 2008.

    You said;

    “The really smart guys don’t talk – they bet their own cash and keep their mouth shut… why would you not factor that into your equation?”

    I do, I had a whole post pointing out that if anyone had a magic formula for beating the market, they’d never talk, as the formula would become worthless.

    Greg, You fail to address any of my arguments. I pointed out that some bubbles crash and others don’t. You point to a bubble that crashed twice–and that’s supposed to somehow refute my argument?

    Darwin’s theory is very practical. It helps understand the breeding of animals. It helps medicine. (antibiotics)

    I use Ratex and the EMH to understand how markets reacted to policy shocks.

    Tom P, I don’t have to bet you, and you don’t have to bet me. We can simply bet the market, if we don’t believe in the EMH. In any case, a bet would tell us nothing about whether the EMH is true. The EMH allows for either side to have been right on a market call.

    If most people think bubbles exist, why would a prediction that housing was not a bubble have received even more acclaim? Everyone I talked to thought the bubble would burst. Predicting it would won’t burst would seem the more adventurous prediction.

    There have been studies that show anomalies do not tend to hold up out of sample. Of course some do, as you would expect if the results were random. The meta studies of all anomalies are more relevant to the EMH hypothesis.

    In my view the question is not whether anomalies exist, but rather whether they are useful. I see no evidence they are useful. If they were useful you’d have anomaly mutual funds set up and outperforming the markets. I’m told that studies show this is not the case.

    TGGP, Thanks for that info.

    So Stiglitz likes Chavez’s economic policies–why am I not surprised.

    Edwin, You misunderstood the point of the post. I am looking for bubble deniers who were right, not bubble deniers who were wrong.

    Didier, There are 7 billion humans. That fact that some thought housing prices would fall and some thought housing prices would rise, has no implications for the EMH debate. I can also show you studies from about 2005 suggesting it was not a bubble.

    James, You said;

    “My argument was that the banksters would be bailed out through an attempt to prevent NGDP falling.”

    Not true. Because the attempt failed, the banksters suffered large losses. Look at what happened to the prices of big bank stocks like Bank of America and Citibank. I wish the attempt had succeeded. Then more loans would have been repaid, banks would have done much better, and we wouldn’t have 10% unemployment.

  77. Gravatar of Greg Ransom Greg Ransom
    9. December 2010 at 08:32

    “Unless you can predict which way three body interactions are likely to go, I say there is no interacting bodies by definition.”

    QED. This is bad science.

  78. Gravatar of Greg Ransom Greg Ransom
    9. December 2010 at 08:33

    Let’s fix the grammar.

    “Unless you can predict which way three body interactions are likely to go, I say there are no interacting bodies by definition.”

  79. Gravatar of Greg Ransom Greg Ransom
    9. December 2010 at 08:36

    It’s become self-evident that you haven’t read “The Big Short” …

    Scott writes,

    “If most people think bubbles exist, why would a prediction that housing was not a bubble have received even more acclaim? Everyone I talked to thought the bubble would burst. Predicting it would won’t burst would seem the more adventurous prediction.”

  80. Gravatar of Greg Ransom Greg Ransom
    9. December 2010 at 08:39

    People who bought houses at the top of the market in Ladera Ranch, CA, and then almost immediately lost $400,000 to $500,000 were folks who predicted the housing bubble didn’t exit ..

    Read about “Zombieland” in MONEY magazine here:

    http://money.cnn.com/2010/12/07/real_estate/ladera_ranch_foreclosure.moneymag/

    Note well.

    I live in “Zombieland” and I know a lot of these people.

  81. Gravatar of Doc Merlin Doc Merlin
    9. December 2010 at 08:40

    @Greg
    “The EMH was invented to account for the problem of taking a simple stream of numbers in a very limited data set, and then using it to extrapolate — “predict” — a future stream of data points, most importantly including the timing of the unfolding of that pattern”

    @Scott:

    “Didier, There are 7 billion humans. That fact that some thought housing prices would fall and some thought housing prices would rise, has no implications for the EMH debate. ”

    He didn’t just think that Scott. He created a systematic framework for identifying bubbles and then repeatedly stuck his neck out by making large public predictions.

  82. Gravatar of Doc Merlin Doc Merlin
    9. December 2010 at 08:40

    Woops, didn’t include my reply to Greg. I was going to say that he makes a really good point.

  83. Gravatar of Greg Ransom Greg Ransom
    9. December 2010 at 08:48

    The long term “bubble” in the value of land on the Island of Manhattan would suggest that what some might like to call “bubbles” aren’t what people are talking about when they talk about bubbles, they’re permanent changes in relative prices.

    Some species continue for millions of years without changing in form and dividing in to new species — i.e. without speciating. This doesn’t disprove the phenomena of speciation.

    There are all sorts of underlying structural things involved in unsustainable malinvestment bubbles that are not involved in all asset / production goods price & quantity movements.

    There are feature of housing bubbles in So Cal that are characteristic of the “up” cycle of the bubble, house flippers, large percentages of homes bought as investments and not to live in, housing moving out into regions hours away from where people work, etc.

    It’s more than just a simple stream of prices — “data” — for one good across time.

    Scott wrote.

    “Greg, You fail to address any of my arguments. I pointed out that some bubbles crash and others don’t. You point to a bubble that crashed twice–and that’s supposed to somehow refute my argument?”

  84. Gravatar of Greg Ransom Greg Ransom
    9. December 2010 at 08:53

    Soctt, you’ve got this backwards. The breading of animals — artificial selection — helped Darwin come up with his mechanism for explaining past historical patterns, patterns which Darwin himself said we had no hope of predicting into the future.

    “Darwin’s theory is very practical. It helps understand the breeding of animals.”

    Note well, Darwinian explanation is not genetics.

  85. Gravatar of scott sumner scott sumner
    9. December 2010 at 09:06

    Mike, What I am really saying isn’t that bubbles don’t exist (an impossible question, since no one can agree on how to define bubbles) but rather that anti-EMH theories are useless.

    Greg, You said:

    “The EMH was invented to account for the problem of taking a simple stream of numbers in a very limited data set, and then using it to extrapolate — “predict” — a future stream of data points, most importantly including the timing of the unfolding of that pattern.”

    Just the opposite, the EMH says markets are impossible to predict. The EMH is very Austrian in spirit.

    Your analogies completely lose me–I see no relevance for the EMH. Maybe someone else can explain the analogy–I don’t see it.

    We can’t predict markets or the way species will evolve. So the EMH seems similar to evolution in that sense. There are fundamental factors driving both markets and species that humans cannot detect. So what’s the difference?

  86. Gravatar of scott sumner scott sumner
    9. December 2010 at 09:12

    Doc Merlin, You said:

    “He didn’t just think that Scott. He created a systematic framework for identifying bubbles and then repeatedly stuck his neck out by making large public predictions.”

    And so did people who stuck their neck out saying it was not a bubble, in academic articles. What’s your point? On any coin flip, some will be right and some will be wrong.

    Greg, The Manhattan example supports my point. Some people thought OC would be the next Manhattan, because of zoning restrictions. They were wrong. Big deal.

  87. Gravatar of Mark A. Sadowski Mark A. Sadowski
    9. December 2010 at 09:19

    Scott,
    It’s far more fun to point out famous bubble deniers who were wrong. The list inludes such people as Alan Reynolds, Kevin Hassett, David Malpass, Steve Forbes, Brian Wesbury, Larry Kudlow, Casey Mulligan etc. The common thread among most of these people is an unflappable devotion to ideology and a completely unnuanced view of reality.

    Were anyone to actually come up with a denier of bubbles who later proved to be right I suspect that they would exhibit very different personal characteristics.

  88. Gravatar of Ted Ted
    9. December 2010 at 09:30

    @Scott

    Just though I throw in a little biology. We actually do predict evolutionary behavior using statistical genetics – in the very short-run (or, at least we try). This is most often done in the case of viruses (particularly influenzas) because they evolve rapidly. This has also been done on plants and fungi. While the method is far and away from perfect, it does perform better than random chance. I don’t think we will ever be able to actually determine the path of evolution with any substantial level of certaintly, but we might one day be able to put reasonable probabilities on certain evolutionary paths.

    Also, speaking of evolution, I’m curious Scott, what do you think of Andrew Lo’s “adaptive market” hypothesis (if you are familiar with it)? I personally don’t think it’s ready to be taken as a serious alternative, but at least Lo’s made a real attempt to try to get somewhere on combining obvious behavioral biases and occasional market inefficiencies without throwing out the EMH entirely. Most people who try to find an EMH alternative, create one and then ignore all the evidence in favor of the EMH and then just show how their theory performs better on various narrow grounds (e.g. “anti-EMH predicts the behavior of glamor stocks better” – while ignoring it does a crappy job with everything else).

  89. Gravatar of Doc Merlin Doc Merlin
    9. December 2010 at 10:18

    @Ted:

    I am not Scott, but the adaptive market hypothesis puts a name to what I have been trying to express for a long time. That markets aren’t efficient, they are “efficiency seeking.”

  90. Gravatar of Harpreet Harpreet
    9. December 2010 at 10:29

    It’s simple to understand why bubble predictors are famous. It is because they are riding against the tide of the time. That by default makes everyone else a bubble denier, because they believe in it so much that they are willing to put their money into those bubbles themselves. How else would the bubble inflate?

    So that is the reason why you will not find many famous bubble deniers, because there are many and when they deny that there is a bubble, they mostly have a vested interest in it.

  91. Gravatar of Barry Ickes Barry Ickes
    9. December 2010 at 10:38

    I don’t get the point. Wasn’t Kudlow, http://old.nationalreview.com/kudlow/kudlow200506201040.asp, a bubble denier? He is always on TV, blowing hard. What does that prove? You can find many people who denied there was a bubble at the time in housing. So what? They turned out wrong.

    Peter Garber wrote an excellent book showing that Tulipmania may have been the result of rational behavior. Well known book, well known economist shows that tulipmania was not a bubble.

    I don’t understand what your point was in asking this question.

  92. Gravatar of Mariano Chouza Mariano Chouza
    9. December 2010 at 11:14

    @Greg

    The motion of three or more bodies under gravitational attraction aren’t very difficult to predict:

    http://en.wikipedia.org/wiki/N-body_simulation

    :-D

  93. Gravatar of james in london james in london
    9. December 2010 at 11:50

    A bubble denier who was right seems a contradiction in terms. A bubble is a bubble, a fact, almost all observed with hindsight, denying it would be denying the truth.

    By the time most people think something is a bubble then it’s usually burst, and we are all using hindsight. It’s hard to imagine a bubble still going on where almost everyone thinks it’s a bubble apart from some hero, that’s not how bubbles/manias work.

  94. Gravatar of james in london james in london
    9. December 2010 at 12:14

    Scott
    The banksters are almost all still in their jobs, only a few fell by the wayside. The industry is growing again. Relative rewards in banksterland (ie in the big 5 wholesale/investment banks) are still way higher than most other trades and professions. The leverage game and its ugly, but, necessary sister, capture/fool the regulators, is back on in earnest.

    The latest agreement to keep tax low and introduce new tax breaks and keep entitlement and other spending high, plays into their hands. Your economy and politics are in a mess, moral hazard keeps being preferred over tough love and real solutions.

    The fight to control these (wholesale/invstment bank) monsters in Europe is still ongoing, and there are signs of succcess in Switzerland and the UK, but it is tough. They remind me of old-style labour unions, but much, much more savvy. And just like those old-style labour barons they love economists who favour bail-outs (and Puts). Ask yourself, have you been captured by them?

  95. Gravatar of marcus nunes marcus nunes
    9. December 2010 at 12:40

    STOP! This comment section has “bubbled up”

  96. Gravatar of Miguel Miguel
    9. December 2010 at 13:31

    Definitely, people is biased, not towards rising asset prices, but nonsense predictions. Take Nouriel Roubini. I’m sure everybody remembers his prediction about the housing bubble, but has anyone checked the other predictions he has made since then? For instance, take a look at this one: http://www.bloomberg.com/apps/news?pid=newsarchive&sid=aS0yBnMR3USk&refer=home

    The point is that if you make predictions all the time, sometimes you are right and sometimes you are wrong. The likelihood depends on what those predictions are. If several people make predictions all the time, probably one or some of them will end up being right all the time. It’s just about statistics. So I guess Mr Roubini is just a statistical phenomenon.

    By the way, this is very much like the Fama/French research on active investors: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1356021

  97. Gravatar of Mariusz Mariusz
    9. December 2010 at 14:24

    Scott,

    What does that mean? “If someone said in 2005 that housing prices were a bubble, but still was unable to offer more than a 50/50 odds on whether real housing prices would rise or fall over the next 5 or 10 or 20 years, then what would their assertion actually mean?” How do you offer better odds? How do you judge that one offer is better than the other?

  98. Gravatar of Greg Ransom Greg Ransom
    9. December 2010 at 14:30

    I’m not confused here. The problem didn’t have a solution better than the market itself. Hence EMH as the “solution”.

    There is no “just the opposite” here.

    “The EMH was invented to account for the problem of taking a simple stream of numbers in a very limited data set, and then using it to extrapolate — “predict” — a future stream of data points, most importantly including the timing of the unfolding of that pattern.”

    Just the opposite, the EMH says markets are impossible to predict. The EMH is very Austrian in spirit.

  99. Gravatar of Greg Ransom Greg Ransom
    9. December 2010 at 14:34

    EMH is not Hayekian. It pretends local knowledge, judgment and macro structure don’t exist — and is the “solution”‘ to a “problem” Hayek see as a non-starter.

  100. Gravatar of Simon K Simon K
    9. December 2010 at 14:52

    Scott – Is it possible to believe both that the EMH is broadly correct and that there are bubbles? Or to put it another way, is there a definition of a bubble that’s compatible with the EMH? I think there might be – The EMH says that all information that can be incorporated into market prices already has been, so you can’t make money by acting on it. I believe this is essentially true. In fact I don’t really see how anyone could not believe it to be true.

    But there are facts that don’t qualify as information in the required sense, because although they will affect market prices when they are discovered, they have not been as yet. In precisely the same way, there are presumably physical facts that are true, but have no as yet been incorporated into the theories of physics, so when we make predictions we omit them.

    For example, in the dot-com bubble there was real uncertainty about what constituted a successful internet business. Clearly actual currently-realized revenues were not the right metric, but investors were extremely uncertain about what was. There were relevant facts, some of which were known to some people, but translating the effect of technical facts about logistics and computer networking into securities pricing is an extremely uncertain business, and at the time almost no-one had the faintest idea how to do it. We found out, as a consequence of the boom and crash, which businesses were actually viable and which were not. But had we known certain things in advance, we would have not had to have the bubble. For example, to pick a rather trivial example, we could have figured out that domain names were not a very important component of internet marketing, a fact most people did not appreciate at the time. But of course we didn’t know those things, or at we didn’t know them in a form where they could be reflected in market pricing. Does it make sense to say there was a bubble, then?

    I think it does, but not in the sense you’re looking for – that we could algorithmically take public information, or even private information that could become public, and use it to come up with better prices than the market. That would violate the EMH. But in the sense that someone could have figured out certain facts that were not generally publicly available or well understood and make money by exploiting them. It makes sense to me to say that, say, Jeff Bezos or John Paulson, made money by exploiting a better understanding of the situation and not just by luck, and at the same time to say that the EMH is true, because the knowledge they had was not generally available in the form of information the market could assimilate.

    Its harder to apply this reasoning to understanding whether there was a housing bubble – what fact was not publicly known that we could have known? I suspect that thanks to the distortions of the American mortgage market, its actually the net demand for housing equity that was the open question, and this might be part of the answer to why the US saw a crash and other countries with similar levels of house price appreciation did not. We don’t really know what the true willingness of Americans to pay for houses is, thanks to the distortions created by the FHA, Fannie and Freddie. Unfortunately I don’t think the crash has actually answered this question, although it has answered some related questions about the desirability of living in (say) San Francisco versus Tracy.

  101. Gravatar of David Pearson David Pearson
    9. December 2010 at 17:25

    Regarding EMH: information asymmetries do occur in markets. For instance, in the fall of 2006, New Century (a leading subprime originator) argued that the spike in delinquencies it was experiencing was just a return to “normalcy” after record low levels. However, it was obvious that the early delinquency experience of the 2005 and 2006 vintages was running way, way above, say, the “normal” 2000 vintage. This information was “lost” at higher levels of aggregation (i.e. averages across vintages).

    New Century went on to “kick off” the financial crisis when, together with HSBC, it finally warned over subprime delinquencies in January of 2007. By August, the run on shadow bank funding was in full swing.

    Why didn’t analysts realize in 2006 that subprime delinquencies were trending double or triple “normal” rates? This is something I asked myself repeatedly at the time. I think the short answer is that their own “stress tests” of subprime ABS portfolios assumed that such delinquency rates “couldn’t happen”. They relied on an analytical paradigm — delinquency models based on historical regressions — that in the end led them off a cliff.

    Does the above violate EMH? I’m not sure. If 99% of market participants use a certain model, then the market is “efficient” in pricing in that model’s result. If a new analytical model comes along, that is “new information”. The process of adopting a new model offers an arbitrage opportunity to first movers. This is not necessarily a violation of EMH, IMO — i.e. the arbitrage opportunity cannot be consistently exploited.

    The analytical support for what we call the “credit bubble” came from PHD’s from elite university math and physics departments (these were literally, in some cases, “rocket scientists”) running regressions on too-small data sets. They simply didn’t know how much they didn’t know. One lesson from the above is that when an analyst/economist asserts, with certainty, “according to my model, such-and-such a thing can’t happen”, a healthy skepticism might be in order.

  102. Gravatar of TheMoneyIllusion » More evidence that the BOJ is not trying to create inflation TheMoneyIllusion » More evidence that the BOJ is not trying to create inflation
    9. December 2010 at 18:13

    [...] few reactions to comments on my recent post on the EMH.  Some people get extremely angry when you defend the EMH.  The debate reminds me of arguments I [...]

  103. Gravatar of scott sumner scott sumner
    9. December 2010 at 20:06

    Mark, We know that bubble deniers will be right roughly 50% of the time, say 40% at worst, the fact that no one can come up with someone who was famous for those predictions, speaks volumes about cognitive illusions.

    Ted, I’m afraid I don’t know about the adaptive market model. What are its practical implications for policy?

    Harpreet, I don’t agree, see part four of my next post.

    Barry, Neither Kudlow of Garber became famous for making a prediction that later turned out accurate. So those two cases don’t have any bearing on my question.

    James, You said;

    “A bubble denier who was right seems a contradiction in terms.”

    I am refering to a situation where it was widely regarded as a bubble, but someone denied it. Say the gold market right now.

    Regarding banksters, yes some are still employed, but they made much less money than they would have made with a good monetary policy, keeping NGDP growing at 5%.

    Miguel, Thanks for the links. I frequently make comments on Roubini’s horrible stock market predictions.

    Mariusz, If you really believe it’s a bubble, doesn’t that mean you think prices are likely to fall over some reasonable period of time? Otherwise, how can one say it’s a bubble?

    Greg, I don’t follow your comment. I thought Austrians talked about how market prices aggregated local information.

    Simon, That’s probably why so many people don’t agree with me. But if the problem is uncertainty, then I don’t think we should call it a bubble. And I think Fama agrees with me. I think many bubble theorists do see it as evidence of inefficient markets. But I agree that different people have different definitions.

    David, That’s a good example, and it raises difficult philosophical issues. That’s why I’ve ended up with the pragmatic question: What use is the anti-EMH view? Obviously the market makes all kinds of mistakes. The real question is whether anti-EMH theories have any practical value. I just don’t see it. To say “try harder next time to be smart all you people” is not very useful, but it’s all I can see for the anti-bubble models.

  104. Gravatar of Simon K Simon K
    9. December 2010 at 21:28

    If we’re going to be pragmatic, then obviously the anti-EMH view is very useful to people who are trying to sell us their investment expertise :-)

    I agree maybe we shouldn’t call periods of uncertainty bubbles. For one thing, there are obviously periods of negative uncertainty (like now!) and we don’t call those bubbles. But I think its periods of positive uncertainty – where there are opportunities of great advantage, but none know what they are – that we end up identifying as bubbles.

  105. Gravatar of Mark A. Sadowski Mark A. Sadowski
    9. December 2010 at 22:00

    Scott wrote:
    “Mark, We know that bubble deniers will be right roughly 50% of the time, say 40% at worst, the fact that no one can come up with someone who was famous for those predictions, speaks volumes about cognitive illusions.”

    Isn’t that part of the title of your website? Man this is getting deep!

  106. Gravatar of Greg Ransom Greg Ransom
    9. December 2010 at 22:05

    Simon is channeling Hayek & Kirzner, Scott.

    No, you don’t “get” Heyek because Hayek is about rivalrous, adapting, and non-identical entrepreneurial judments as much as he is about prices as signals — the economic actor is a learner in Hayek, and some learners prove to have come to better understandings than others.

    The EMH is the answer to the failure of the dream of a robot / machine solution to ambition of beating the market, using a simple stream of numbers alone, stuff significant to a mere machine.

    Hayek rejects that picture of what an entrepreneur is doing — he isn’t limited to a tiny and bare stream of money numbers.

    And he gives irrefutable reasons for understanding the economy as having a deep production / price structure that can’t be fully known to one mind, and can’t be “modeled” with a simple data stream of statistically related prices.

    And we can have some sense of distortions in this deep structure — but no one can have a perfectly accurate picture of this structure, and all will see it from their own local position, each therefor will have a rivalrous stance toward working successfully within it.

    Those esp. with no sense of the deep structure will be most susceptible to being fooled by price relations — they won’t believe in “bubbles” because they can’t imagine the causal structures whcih sustain them.

    Their lack of imagination and understanding doesn’t make those stucture any less real — just as the inability of pre-moderns to image or understand does not make it that case that oxygen didn’t combine with carbon to produce fire in the pre-modern age. (I know this comment about pre-moderns and fire opens an avenue for you to get off topic — please resist).

  107. Gravatar of December 10, 2010 Crossing Wall Street December 10, 2010 Crossing Wall Street
    10. December 2010 at 04:40

    [...] Who Are The Famous Bubble Deniers? [...]

  108. Gravatar of Greg Ransom Greg Ransom
    10. December 2010 at 09:29

    This is absurd as a supposed defense of your argument:

    “Greg, The Manhattan example supports my point. Some people thought OC would be the next Manhattan, because of zoning restrictions. They were wrong. Big deal.”

    We now KNOW there were structural factors behind the bandwagon that created the ilusion for those not very alert or ignorant of good economics (I don’t count mainstream macro as particularly good economics).

    Ignorance was revealed by history — many of the most ignorant were identified and targetted even at the time.

    All sorts of people on the ground were in on the game, and they knew their “marks”.

    They didn’t knowmthe timing but they knew the game had an expiration date.

    You are forced to profess belief in far too many obviously false things for me to believe you hold your own position with any firm conviction grounded in facts or real wolrd phenomena — the “model” tells you what to believe, the nature of reality and genuine scientific practice be damned (after the fsct explanation of things with insuperable barriers to precise prediction aremdime a day in real science.)

  109. Gravatar of james in london james in london
    10. December 2010 at 10:16

    Scott
    You say: “Gold today is widely regarded as a bubble”. It’s good you have come up with an example. Is it true? Given it’s so easy to short gold your statement can’t be true. Please try again.

    The banksters would certainly have made even more money if nominal GDP had kept rising at 5% and you’d managed to rescue them 100% from their leverage gamble with a more fully successful version of the Fed/Sumner Put.

    Jimmy Cayne and Dick Fuld and Kerry Killinger and (Sir) Fred Goodwin and Adam Applegarth et al would all still be with us. And they’d all be crowing they’d survived a minor liquidity crisis – and by now they’d all be paying themselves more than the millions they’d earned in 2006. The top guns at your admired GS would not have had to leave behind those halcyon days of $70m a piece they earned for 2006.

    But would we be heading for an even bigger, leverage-driven, systemic, crash? Now there’s a counterfactual.

  110. Gravatar of Greg Ransom Greg Ransom
    11. December 2010 at 10:57

    Was the Bernie Maddoff fund a bubble?

    How does the EMF apply to the Maddow fund?

  111. Gravatar of scott sumner scott sumner
    11. December 2010 at 15:05

    Simon, I agree.

    Mark, My blog title has multiple meanings.

    Greg, You said;

    “And he gives irrefutable reasons for understanding the economy as having a deep production / price structure that can’t be fully known to one mind”

    Exactly, no one mind can know what’s going on.

    Greg, Sorry to disappoint you, but I do believe what I write.

    James, Just because you can short gold doesn’t mean it’s not a bubble.

    We’ll cross that “even bigger crisis” bridge when we come to it. And NGDP targeting will get us across even the biggest chasm.

    Greg, The Maddoff fund was not a bubble. And no, I don’t think the EMH has much of interest to say about Maddoff.

  112. Gravatar of 日銀がインフレを起こそうとしていない更なる証拠 by Scott Sumner – 道草 日銀がインフレを起こそうとしていない更なる証拠 by Scott Sumner – 道草
    22. December 2010 at 08:47

    [...] EMHについての私の最近の投稿へのコメントにいくつかお答え。君達がEMHを擁護しようとすると、ある種の人々は激怒する。このディベートは、私が自由意志に関してする議論を私に思い起こさせる。私は、出来事には原因があるか、または偶然であるかどちらかだと指摘する。どちらの場合にも、自由意志の入り込む余地はない。私の相手は、彼の目の前にある塩入れと胡椒入れのどちらを拾うか自由であると言い返す。このディベートは、本当に決して交わることなく、それ故に殆ど時間のムダだ。自由意思について反対したり、EMHを擁護したりする議論でこれ以上時間を浪費すべきではないと私は確信した。 [...]

  113. Gravatar of Rick Hull Rick Hull
    17. May 2011 at 17:25

    Could it be that EMH + ZIRP = HE PRIZM?

    No, seriously: that bubbles are effectively malinvestment, driven by relentless (and especially, acute) monetary expansion. All those dollars feel safe in numbers until it goes critical, and by then the nimble ones have chosen the new asset du jour.

Leave a Reply