Eggheads vs. “real world nitty-gritty”

I am getting burned out dealing with endless complaints about the way I think about markets.  Thus I thought it might be useful to compare my views to what I hear from those down in the trenches.  Those who actually know how auction-style markets work.  Bob supplied a typical complaint:

The financial instability caused a huge demand for commodities as investment – just look at the USO holdings of futures contracts last spring. Airlines don’t buy ETFs to get their oil.

I’m increasingly coming to the conclusion that academic economists have pretty close to 0 comprehension of how commodity futures markets actually function (this is really not even directed at you Scott). They don’t even understand the theory of futures pricing, let alone the real world nitty-gritty.

OK, so us academic economists are idiots.  That’s pretty much what I been saying about my fellow economists during this crisis, so I won’t get defensive here.  But where do I go for enlightenment?  Perhaps I should listen to those in the trenches.  So I turn on one of the Wall Street shows, and this is what I hear:

“There’s lot’s of cash on the side lines, waiting to go into the market”

I have a question; when tourists visit the NYSE do they get to see the room where they keep all the cash that investors put “into” the stock market?

Next I hear explanations for why stocks fell 1.8% on a given day:

“A selling wave hit the market in late afternoon, driving stocks sharply lower.”

Well thanks for clearing that up.

Before people accuse me of sarcasm, let me admit I make the same mistakes on occasion.  It’s human nature.  Sometimes humor makes a point more forcefully.  (BTW, Bob had a much longer comment, and may well have been correct.)  Of course I know what people “really mean” by the remarks I made up, and I certainly don’t think either economists or real world traders are idiots.  Often we simply use different language.  So the following is in my language.  Feel free to agree or disagree.

Let’s start with the selling wave on Wall Street.  Presumably what the trader meant is that at the previous price, more investors wanted to sell than buy.  Of course actual sales equal actual purchases, so on days when a selling wave hits Wall Street, a buying wave also hits Wall Street.  One might refer to types of buyers and sellers (specialists, market-makers, or whatever they call them) but even that is not really an explanation.  Stocks can fall sharply on massive volume with (at the end of the day) no change in the net holdings of any particular group.  What matters is price.  Indeed in the days before 24 hour trading the stock market would often crash without a single share being sold.  It would crash when the market was closed, and reopen the next day with the very first share being sold at much lower prices.  So my first claim is this:

1.  The fact that speculators “poured money” into a particular market tells us absolutely nothing about price movements.  Correct me if I am wrong, but the decision of investors to pour $100s of trillions into oil futures, would not make oil futures prices go up one cent.  This is because as fast as the money poured into the markets it would pour right out again.  (Every transaction is a purchase and sale.)

2.  Of course speculation can change the price of spot oil.  This can occur in several ways:

a.  It may lead producers to think that future production will be more profitable than current production.  Thus higher futures prices might cause suppliers to produce less.

b.  Consumers of oil may decide to burn less oil and store more oil, again in anticipation that future consumption will be more beneficial than current consumption.  There could be many reasons for this, but the most obvious would be expectations of higher future oil prices.

So I do understand that expectations of higher future oil prices can increase current oil prices.  I’m just not convinced that that was the key factor behind the 2007-08 oil price run-up.  I don’t see any reason why oil prices could not have stayed high if the world economy had not crashed in the second half of 2008.  Markets deal with enormous amounts of uncertainty.  Obviously when major unexpected shifts in the business cycle come along, prices will change dramatically.  And the pre-crash prices will look foolish in retrospect.

The data suggests world oil production rose slightly between 2007 and 2008.  I also believe that inventories were slightly lower in 2007 and didn’t change much in 2008.  Thus I don’t see evidence that speculators caused high prices.  If anything speculators seem to have held prices down in 2007, as oil stocks were shrinking, and been roughly neutral in 2008.  Here is the data I looked at:

World oil production, consumption and inventories

I freely admit that I may have misinterpreted the data.  It is possible that speculators played a major role in the oil price run-up.  Maybe oil production would have risen even more without speculation.  But I don’t see any evidence for that hypothesis.  And I don’t consider the fact that people bought and sold lots of oil futures to be evidence of anything.

There is a very deep human instinct to try to explain things that seem mysterious, but on closer examination don’t need any explanation.  In politics that shows up as conspiracy theories:

1.  The trilateral commission.

2.  Who really shot JFK?

3.  Who was behind 9/11?

4.  Where are those Iraqi WMD?

5.  The NBA draft lottery.

etc.

In economics it tends to show up in theories that mischievous speculators are causing undesirable price changes.  Speculators get a bad rap.  They are simply trying to make money—doing the best they can.  If they buy oil, it’s because they think oil will be higher in the future, and they want to conserve some today for when it will be more valuable to society.  (Of course that’s not what they are thinking, rather it’s what Adam Smith’s invisible hand is encouraging them to do.)  Sometimes they are right and sometimes they are wrong.  That’s all one can really say.

I know this seems to contradict my monetarist “excess cash balance” transmission mechanism.  But that is different.  In that case the money market itself is in disequilibrium, but it doesn’t cause any particular price bubble.  In a flexible price world, a 10% increase in the money supply should, ceteris paribus, raise all prices by 10%.  If prices are sticky it may have real effects, and cyclically sensitive prices may rise more.  But in the first half of 2008 the U.S was in a mild recession.  So I don’t see how monetary policy could have raised real oil prices through the normal monetary mechanisms (QTM or cyclical effects.)  And thus I don’t see why oil would go up because of easy money.  Why not equities?  If you say because the oil market was “hot,” my response is it was hot because of real factors–like rising demand in LDCs.  Note OECD demand was falling but LDC demand was rising, and not just in China.

OK. Fire away.  Tell me how I don’t know anything about markets.  I’m not thin-skinned.

PS.  A Chinese language version of this blog was just set up by Netease (which I believe is roughly the Yahoo of China.)  Someone told me that my picture was also on Netease’s finance and economics page.  Given the Chinese interest in stocks, that means probably about 100 times more people have now seen my face than in the previous 53 years of my existence.  Interestingly, I am going to China later this year.  Here is the site–notice all the pictures.  I plan a post on the Chinese yuan this weekend, to welcome the new visitors from China.

Disclaimer:  I have no control over this new site, and am not responsible for translations or choice of pictures, etc.


Tags:

 
 
 

61 Responses to “Eggheads vs. “real world nitty-gritty””

  1. Gravatar of Jimmy Daroukas Jimmy Daroukas
    23. May 2009 at 00:05

    I agree with your overall point about speculators, but a couple of points:

    “If they buy oil, it’s because they think oil will be higher in the future, and they want to conserve some today for when it will be more valuable to society.”

    I’m probaby nitpicking, but speculators don’t usually conserve any oil. They will mainly buy or sell futures and do not take or make physical delivery. Some entities are able to store oil and sell it later (during period of contango), but if I’m not mistaken, this is mostly done by the big oil companies, not speculators.

    “as fast as the money poured into the markets it would pour right out again. (Every transaction is a purchase and sale.)”

    While it’s true that every transaction is a purchase and sale, its not true that money that has poured in automatically pours out. This can be illustrated by open interest in futures markets. If Trader A goes long 10 contracts, with Trader B taking the shorts side, open interest has now increased by 10 contracts. This is new money in the market. They money will leave the market when both close out their positions. (Maybe someone can explain this better than me).

  2. Gravatar of RebelEconomist RebelEconomist
    23. May 2009 at 01:47

    Understanding asset price changes is rarely straightforward, but I do believe that there was an element of inflation expectations to the commodity price rises last year. Given the depressing effect of rising commodity prices – especially oil – on US economic activity, it seemed likely that (a) the Fed would be reluctant to resist higher inflation driven by rising oil prices and (b) that real earnings of US firms would be weak, meaning that stocks would be a relatively poor inflation hedge.

    A related, but wider question that I would be interested in your views on Scott is, in the case of deposit money, why should the quantity be expected to bid up anything, given that the deposits are generally matched by debt on the other side of the banking system balance sheet? If you have written about this already, feel free to refer me to a previous discussion.

  3. Gravatar of 123 123
    23. May 2009 at 02:30

    “Speculators get a bad rap. They are simply trying to make money””doing the best they can. If they buy oil, it’s because they think oil will be higher in the future, and they want to conserve some today for when it will be more valuable to society. ”

    Some speculators conserve oil today for when it will be more valuable to society. Other speculators conserve oil today for when it can be sold to greater fools.

  4. Gravatar of ssumner ssumner
    23. May 2009 at 06:55

    Jimmy, If there is no physical storage of oil, the spot price does not rise. Futures transactions only affect the spot price to the extent that they affect current supply and demand. Otherwise it’s just like if I make a bet on oil prices with my brother.

    Regarding expansion of the size of the futures market, I suppose that is possible. But of course that’s not what TV talkingheads mean when they refer to money on the sidelines waiting to jump into the market. I don’t know a lot about futures markets. Do they actually store a lot of base money (cash and bank reserves?) If not, money is not going into the market, it is going through the market. Remember that “the public,” broadly defined to included institutions, does not control the total stock of base money. So the public cannot decide they want to hold less base money and more of some other type of asset (except of course in real terms, which I discussed in the post.)

    rebeleconomist,

    On the first question, I would say that based on the TIPS markets, inflation expectations were no where near high enough to explain the commodity bubble. I think the longer term inflation spread peaked at about 2.5% In addition, isn’t gold a better inflation hedge than oil? Oil is costly to store. And if we go back to the previous commenter, remember than the inflation hedge role (in net terms) is only fulfilled in the spot market. Futures contracts net out to zero.

    123, If someone tells me they are buying an asset that they think the fundamentals will turn against, in the hope that there will an even greater fool out there to take it off my hands before it crashes, what do you think my advice would be to that person? My hunch is that people who think that way don’t have enough money to seriously impact markets, and certainly aren’t the marginal trader. But of course that is what the bubble debate is all about—how many foolish people are there, and how much money do they have to invest? I hope I’m wrong, I’d love to dive in and separate a lot of money from foolish owners. I’m ready to go as soon as someone gives me a plausible blueprint of how to do it. But I get discourage when Shiller and Greenspan tell me to sell on the eave of one of the greatest 4 year bull markets in American history. I should have been buying, so I could sell to the fools in April 2000. So I still don’t see a blueprint out there.

  5. Gravatar of Bill Stepp Bill Stepp
    23. May 2009 at 07:47

    Scott,

    A general comment about Krugman’s “Minsky moment” comment.
    A theory of the business cycle can’t pinpoint such a moment,
    because two things have to happen before the margin calls come in, namely (1) a bearish seachange in investors’ expectations and (2) a ratcheting up in the market’s discount rate used to value expected cash flows following on the changed expectations. Krugman and DeLong both overlook this, as far as I can tell.
    But even if we adjust for that, there will be lots of noise in the markets and no model can predict when it will happen.

    Have you read the Austrians on the business cycle? Hayek, Mises, Roepke, et al.? A more recent book by a non-Austrian is an essential supplement to the Austrians. Alfred Rappaport, “Creating Shareholder Value,” 2nd ed., discusses several issues and how to think about investors’ changing expectations and their impact on cash flows and stock values. You can then bring in the Austrians and fill in a couple of blank pages they left unwritten.

    I am unimpressed by DeLong and Krugman’s stuff on macro and the recent market gyrations.

    Also (this might be better left for another time), when are you going to consider the benefits of abolishing the Fed and going with free banking?

    Or maybe we do need centrally planned money. If the Fed has a monopoly of money, why not have a shoe reserve and give the Federal Shoe Reserve head the authority to determine the supply of shoes?
    And I’m not even going to bring up Uncle Ben’s getting a real job like pitching rice or piloting a real chopper!

  6. Gravatar of happyjuggler0 happyjuggler0
    23. May 2009 at 07:47

    Scott,

    I really hope you don’t stop posting and replying due to some type of “burnout” frustration. You are a gem, and even though I stumble sometimes trying to understand what you are saying, I nevertheless deeply appreciate your help.

    Let’s start with the selling wave on Wall Street. Presumably what the trader meant is that at the previous price, more investors wanted to sell than buy. Of course actual sales equal actual purchases, so on days when a selling wave hits Wall Street, a buying wave also hits Wall Street

    It is a bit disturbing how many people don’t quite grasp this, or forget it. People who should know better.

    Last spring/summer I was watching CNBC, and one of their anchors kept arguing that oil was in a bubble, because “there was no shortage”. She patiently explained that it was impossible for demand to be higher than supply, for every buyer there has to be a seller. Hence, in her view, if there was no shortage, there had to be a bubble.

    I wanted to cry, not only because she was so close, but yet so far, but also because she had a degree in economics from Wellesley, and she’s a libertarian to boot.

    Anyway, for those who can’t figure out where she went wrong, the clue comes from Scott’s quote above, namely: at the previous price….

    While the market price was at $140 per barrel, there was indeed a shortage of oil. The was a shortage at $139, an even larger shortage at $135, and a mega-shortage at $40 per barrel. The key of course is that with free markets determining prices, if there are more buyers than sellers at the current price, then the price rises enough until some buyers drop out and some new sellers come to market.

    So the price rises. In a commodity with highly inelastic supply and highly inelastic demand, you get a boom/bust game of musical chairs, where the chairs are auctioned off instead of first come, first served like in the children’s game. 11 seats, 10 players, and the seats are dirt cheap. 9 seats, 10 players, and all of a sudden the price of the seats skyrockets, at least if everyone still wants to play the game. And who doesn’t want to keep playing the game of driving?

    P.S. regarding “money on the sidelines”, it could well be that the recent few months of a US equity boom is the market discounting the day that that “sideline money” enters the market. One presumes they are waiting “until it is safe”. But what is safe? When one can see in the rearview mirror that the economy is expanding? If they wait to get in at that point, it is highly likely that that move would already be discounted, and “the discounters” might very well sell their positions to the “previous sideline money”. In this scenario, when the “sideline money” enters the market, we may get high volume with small price movement in any direction.

    Just my speculation, so to speak. Anyway, this is one post that I completely agree with you Scott.

  7. Gravatar of Thruth Thruth
    23. May 2009 at 08:39

    Scott, by and large I agree with the basic thrust of this post.

    >Of course I know what people “really mean” by the remarks I >made up, and I certainly don’t think either economists or >real world traders are idiots.

    I’ve always interpreted the “money on the sidelines” as a sort of “popular mechanics” way of describing asset class/sectoral shifts. In the wrong hands, it leads to stunning levels of misinterpretation.

    Onto the bits I disagree with:

    >So I do understand that expectations of higher future oil
    >prices can increase current oil prices. I’m just not
    >convinced that that was the key factor behind the 2007-08
    >oil price run-up.

    That sounds circular to me. The best theory of the price of long lived assets says they follow a martingale. If price is high today, it’s expected to be high tomorrow (go back to the intertemporal optimization for the why). All the usual caveats about frictions etc apply of course, and maybe you’re talking about those.

    > I don’t see any reason why oil prices
    >could not have stayed high if the world economy had not >crashed in the second half of 2008.

    That’s consistent with rising expectations.

    > Markets deal with enormous amounts of uncertainty.
    >Obviously when major unexpected shifts in the business
    >cycle come along, prices will change dramatically. And the
    >pre-crash prices will look foolish in retrospect.

    As I suggest in the other thread, to the extent market participants believed that Feds commitment to sustaining AD (Bernanke’s “we learned a lot from the great depression”) the price of commodities should be rising strongly as commodity based sectors are increasingly utilized to meet AD that can’t be met by the crushed financial and US housing sectors.

    >The data suggests world oil production rose slightly
    >between 2007 and 2008. I also believe that inventories
    >were slightly lower in 2007 and didn’t change much in
    >2008.

    I think we agree the one period change in inventories are not a sufficient statistic for changing expectations.

    > Thus I don’t see evidence that speculators caused
    >high prices.

    Assuming by this you mean, prices don’t change unless (expected) production and consumption plans change then I agree.

    Pulling in a couple of your comments from the other thread:

    >if it’s an argument that resources in banking shifted to
    >more energy intensive industries, that’s OK.

    yes, that’s the argument.

    >Traditionally both oil shocks and housing crises come
    >near the end of long upswings, don’t they?

    coincidence? Markets clearly recognize the problems well before the recession is called. They “shift” into defensive assets and sectors every time.

    >BTW, even if I am totally wrong about oil, it doesn’t
    >weaken my money argument. That is because oil shocks reduce
    >RGDP, not NGDP.

    I think my story about oil makes your money argument stronger because it gets rid of the distraction about oil prices causing the recession. Oil price rises were tied to the financial shock and the resulting monetary policy commitment (interest rate cuts of 4% over twelve months plus other measures). The subsequent decline in mid/late 2008 was the result of the failure of the money commitment.

    >I also don’t think “safe play” is the right term for the
    >commodities. When oil is $140, isn’t that a pretty risky
    >investment? Almost everyone realizes what goes up may come
    >down. I think you need some sort of expectations of
    >continued real growth, to supplement your other arguments.

    More “popular mechanics” wording :). I’m using safe in two senses: (1) relative to other asset classes; and (2) in the face of the Fed signaling it’s commitment to maintain AD.

    On real (or relative) growth: the consensus for financial and US housing seemed to suggest the potential years of problems from very early on in the crisis. So if authorities are credibly signaling the intention to maintain AD, that suggests outsized growth in other sectors.

    But I do think risk matters. No one knew how bad things could get — clearly RGDP would be higher if the banking/US housing problems were smaller.

  8. Gravatar of Devin Finbarr Devin Finbarr
    23. May 2009 at 09:05

    The data suggests world oil production rose slightly between 2007 and 2008. I also believe that inventories were slightly lower in 2007 and didn’t change much in 2008.

    Looking at inventory build up for oil is only a piece of the picture. Oil can simply be stored in the ground by the owner countries. The two articles that convinced me that “speculation” drove up the prices were There Is No Gas Shortage and Commodity currencies, the new derivatives.

    The fact that speculators “poured money” into a particular market tells us absolutely nothing about price movements.

    What? “Pouring money” means an increase in demand. If demand goes up, prices go up. I do agree that the language the Wall St journal uses (“putting money into the market” , “taking money out of the market”) is awful. No one puts money in the market, they simply give their dollars to someone currently holding the share in return for that share. But if the amount of dollars investors wish to spend on shares or futures contracts increases, then the price will increase.

    But where do I go for enlightenment? Perhaps I should listen to those in the trenches. So I turn on one of the Wall Street shows, and this is what I hear:

    TV is certainly not the place to learn about stocks or trading!

    I actually find that the best information from the trenches I get comes from the comments sections on sites like New Mogul, Hacker News, Brad Setser’s Follow the Money, or Econlog. I also have read a number of good posts the following blogs: Macro Man, Cassandra Does Tokyo, CT Oysters, Unenumerated, and Unqualified Reservations.

  9. Gravatar of RebelEconomist RebelEconomist
    23. May 2009 at 09:09

    Scott,

    I would question whether inflation expectations derived from TIPS are properly representative. First, I believe that many funds cannot buy them because they cannot cope with their unusual properties (eg unknown future cashflows, unlike conventional treasuries). Second, rightly or wrongly, the US authorities have introduced changes to the CPI methodology that have had the effect of reducing measured inflation, so there may be some doubt about how good an inflation hedge TIPS might actually be. Bernanke himself used to favour the UoM inflation survey. Nevertheless, changes in inflation expectations ought to be meaningful.

    I would be grateful for your thoughts on my second question (about whether the Q effect on expanded broad money is offset by the matching expansion of debt).

  10. Gravatar of Eric Morey Eric Morey
    23. May 2009 at 09:36

    You seem to be overlooking the fact that the “talking heads” on the “Wall Street shows” are not “traders”, otherwise they’d generally be too busy trading to be talking on TV. I’d be willing to bet that the majority of the “talking heads” never have and never will be “traders”. Listening to them is somewhat like a professional sports player player watching a sports TV show for coaching advise. I’d listen to Nassim Taleb before any “Wall Street show talking head”.

  11. Gravatar of Devin Finbarr Devin Finbarr
    23. May 2009 at 10:35

    Rebel-

    A related, but wider question that I would be interested in your views on Scott is, in the case of deposit money, why should the quantity be expected to bid up anything, given that the deposits are generally matched by debt on the other side of the banking system balance sheet?

    Because of a) the maturity mismatch and b) FDIC insurance.

  12. Gravatar of ssumner ssumner
    23. May 2009 at 11:38

    happyjuggler0, That’s a nice CNBC story. People often think rising demand implies shortage. There are so many fallacies one hardly knows were to begin.

    The blog will continue, but I need to slow down for a few weeks to revise my Depression manuscript, which I just got back from the publisher. So not many posts for a while, I’m afraid. But don’t worry about burnout, the show will go on.

    I’m glad you liked the post.

    Thruth, There is no circularity. I agree that current and future prices are linked. But they can be linked in many ways.
    1. A higher expected future price can cause lots of current hoarding of oil, pushing up the spot price.

    2. Or high current demand may be expected to persist, pushing up futures prices as well.

    The quantity consumed and produced data points toward option 2. That was my point.

    I see how your point could support my interpretation. I think where I have trouble is that I think (like most people) the banking/housing crunch pushed us into a mild recession in December 2007. I just have trouble seeing how that sort of shock could push up the price of a highly cyclical good like oil. LDC demand seems more plausible. But you may be partly right, and I’ll keep your view in mind as I think more about these issues. (As you know, I don’t think a view is wrong just because it’s counterintuitive.)

    Devin, I did address the owner country hoarding problem in the post. I said world oil output rose between 2007 and mid 2008. So I don’t see how supplier hoarding could have been a huge factor. But I’ll keep an open mind on the issue as more evidence comes in. Perhaps Qs would have risen even faster, without supplier hoarding.

    Transactions don’t cause market prices to change, indeed they are more a symptom. Prices often change sharply with no transactions. And the day stock prices collapsed 22% in 1987, investors poured record amounts of money into the stock market, in terms of actual transactions. Transactions reflect changes in which specific people prefer to own shares, they have nothing to do with why share prices change. Prices can rise on heavy volume or fall on heavy volume. When share prices jump 10% on news, if everyone sees the news in the same way there will be no transactions at all, no “money going into the market,” to use everyday speech.

    Also, demand curves are schedules of how much of an asset people want at various prices. In 2008 the world demand for oil soared, and yet “demand” as reported in the newspapers (quantity demanded) barely budged.

    Rebeleconomist, You might be right about TIPS, but I don’t recall the consensus inflation forecast of economists being very high either, at least long term. Is my memory wrong?

    Sorry, I completely forgot your other question. This is a common misconception about OMOs. You are right that swapping cash for bonds doesn’t change total wealth. So their is no direct AD increase from that channel. But during normal times cash and bonds are not close substitutes, and people only want to hold so much of a non-interest-bearing asset like cash. If you put more cash (or reserves) into circulation than people want to hold, they will try to get rid of their excess cash balances by buying more goods, services, and assets, pushing up AD.

    Regarding deposit money, people make the same argument. If you create more deposit money that the public wants, it tries to reduce its balances and drives up AD. But personally, I don’t define money to include deposits. I regard deposits as debt, not money,

    Eric, They interview people in the trenches all the time on the Wall Street show I watch (the all day one.) The comments are just as silly as non-traders, at least on many occasions. It’s all just a bunch of meaningless cliches. To use your sports analogy, it’s “we can’t look past this game” “we just got to take them one game at a time” and similar drivel.

    Devin #2, I like my answer better, but maturity differences and FDIC do influence the demand for deposits, so they do play a role.

  13. Gravatar of Devin Finbarr Devin Finbarr
    23. May 2009 at 11:57

    Transactions don’t cause market prices to change, indeed they are more a symptom.

    Right. Stock prices of index are simply the average price of all sales. Or in other words, total dollar amount of the transactions divided by shares sold. When people say “an influx of money caused share prices to rise”, they are talking about the special case when people decide to allocate more money to buying stocks, without a corresponding increase in the desire of sellers to sell shares. By definition, the average price of sale must rise in this case.

  14. Gravatar of Devin Finbarr Devin Finbarr
    23. May 2009 at 12:11

    Scott-

    But personally, I don’t define money to include deposits. I regard deposits as debt, not money,

    I think this raises a fundamental issue which is at the heart of many of our disagreements.

    Every complex human society selects a good to be the primary *intermediary good*. This good has no direct utility, It’s a good of indirect exchange. You trade your ox for the intermediary good, and then a few days, months, or years later, you choose to trade the intermediary good for a wagon. Thus all demand for the intermediary good is demand to be hoarded. The entire purpose of the good is to serve as a store of value in the duration between when you produce a good and when you want to consume a good. Most societies will denominate prices and contracts in terms of the intermediary good. However, not all societies will do so. For instance, in many third world countries prices will be denominated in the local currency, but people will use American dollars as a store of value. As soon as a company earns income in pesos, it converts it to dollars for saving. Or for another example, an arcade may denominate prices in “credits”.

    In the modern American economy, what fulfills the role of “the intermediary good”. Monetary base? M2? M3? MZM? M3+Treasuries – desposits backed by treasuries?

  15. Gravatar of RebelEconomist RebelEconomist
    23. May 2009 at 13:04

    Scott,

    Yes, last year the UoM one year ahead inflation expectations survey ( http://research.stlouisfed.org/fred2/series/MICH ) reached over 5%, a level not seen since the early 1980s. Of course Bernanke then stopped mentioning it! The five year ahead survey does not seem to be so readily available, but if I recall correctly, it followed a similar pattern but with a much more muted rise.

    You seem to have missed my point about whether deposit (inside) money can bid up prices, but from what you say I guess you emphasise base money anyway. That is perhaps most reasonable, but some would say that all money is ultimately inside money, because base money is essentially public debt.

  16. Gravatar of Carl Futia Carl Futia
    23. May 2009 at 15:37

    Dear Scott:

    I am a professional speculator – you can check out my blog for evidence: http://www.carlfutia.blogspot.com. As a speculator I spend my days deducing expectations from trading activity in the futures markets. In a previous life I was an economic theorist with publications in Econometrica, QJE, JET. etc. I have seen and worked on both sides of the fence – egg head and nitty-gritty practioner.

    In my view you are spot-on in your beliefs about how expectations are reflected in the activity and movements of speculative markets. Markets themselves reveal changes in expectations far more quickly and more accurately than any other time series. I also think that economists, especially macro-economists, don’t generally understand that the efficient markets hypothesis dooms their attempts to forecast turning points in economic time series.

    I might add that it is a fundamental theoretical observation and empirical fact that the participants in any market as a group know more than any individual can know, AND more than is revealed to anyone by price movements in the market itself. Any theoretical model which proports to explain movements in aggregate economic variables must reflect this fact. In particular, economic models cannot allow economists to forecast the economy better than do the hypothetical economic agents in these models.

  17. Gravatar of 123 123
    24. May 2009 at 03:18

    “If someone tells me they are buying an asset that they think the fundamentals will turn against, in the hope that there will an even greater fool out there to take it off my hands before it crashes, what do you think my advice would be to that person? My hunch is that people who think that way don’t have enough money to seriously impact markets, and certainly aren’t the marginal trader.”
    Some of the largest asset managers are using this kind of strategy.

    “But of course that is what the bubble debate is all about””how many foolish people are there, and how much money do they have to invest? I hope I’m wrong, I’d love to dive in and separate a lot of money from foolish owners. I’m ready to go as soon as someone gives me a plausible blueprint of how to do it. But I get discourage when Shiller and Greenspan tell me to sell on the eave of one of the greatest 4 year bull markets in American history. I should have been buying, so I could sell to the fools in April 2000. So I still don’t see a blueprint out there.”
    It seems you want to get a theory that allows you to outperform the market by 10X. Shiller’s theory allows you to add 1-4% points per year over a very long term (say 20years) with huge deviations from benchmark.

  18. Gravatar of Thruth Thruth
    24. May 2009 at 04:43

    Scott, “I just have trouble seeing how that sort of shock could push up the price of a highly cyclical good like oil. LDC demand seems more plausible. But you may be partly right”

    LDC demand isn’t formed in a bubble. And it wasn’t just oil that displayed this pattern. The 4% rate cut and all the other measures taken between Jul07 and Jul08 had to show up somewhere.

    Somewhat related, I’d like to hear your thoughts on Fed policy during the oil price shocks. (Perhaps a blog post when you are less busy). Found an 2005 FRBSF letter on oil price shocks, which included this snippet.

    “Other explanations for the 1970s

    Considerations like these have led a number of economists to suggest that the recessions of the 1970s reflected other kinds of shocks. For instance, Barsky and Killian (2001) argue that the great stagflation of the 1970s was the result of monetary policy alternating between periods of stimulation and restraint and not oil price shocks. Similarly, Burbidge and Harrison (1984) examine developments in five major industrial economies including the U.S. and conclude that even though the oil shocks in the early 1970s did have a significant effect, recessions were already on the way even before the jump in oil prices. They also find that the 1979-1980 oil shocks had a minimal effect on all these countries except Japan.

    Others have argued that the recessions may have been caused by the Fed’s reaction to the oil shocks. Bernanke, Gertler, and Watson (1997) show that postwar recessions have been preceded not only by rising oil prices but also by a tightening of monetary policy, which makes it difficult to distinguish between the effects of the two. According to them, the confusion between oil shocks and the response of monetary policy explains why oil shocks appear to have an effect that far exceeds what is expected based on a comparison of energy costs to total production costs. Their own analysis leads them to conclude that oil shocks have not played a major role in recessions and that endogenous monetary policy can account for a major portion (and sometimes all) of the effects attributed to oil shocks.”

    http://www.frbsf.org/publications/economics/letter/2005/el2005-31.html

  19. Gravatar of ssumner ssumner
    24. May 2009 at 06:46

    Devin, When lots of people who don’t own stocks rush in to buy stocks, there is no flow of money into the stock market. The money is given to those selling, and is still on the “sidelines” every bit as much before the sales took place. The concept of money on the sidelines waiting to go into a market makes no sense. People should talk about changing expectations leading to price changes, period.

    Devin#2, Base money–check out the other post where I respond to you and Nick.

    Rebeleconomist, Michigan one year surveys means little. We were in the midst of rapid (oil) inflation at the time. Most people don’t even know what the term ‘inflation’ means, they think a higher cost of living means higher inflation. I’d look at 5 years TIPS first, then 5 year consensus forecasts of economists.

    Carl, Thanks, I have been trying to make many of those points in my blog. Macro is still in the stone age in many respects. Macroeconomists act as if they know more than financial markets, and just dismiss TIPS spreads when it goes against their theories.

    123, I don’t know how many money managers use those theories, but I still doubt it represents the marginal trader. Otherwise trading stocks would be like taking candy from a baby. And I’ve never found it easy to make money in the market.

    I don’t know what Shiller study you refer to, but all market efficiency studies that I have seen use historical data. Which isn’t at all meaningful. I could come up with 100 models that beat the averages, merely by looking at what happened to key indicators in the past, and fitting my model to those stylized facts. I.e., I could check out how the market performed at various P/E ratios over time, and come up with a formula. But what would that prove? I’m not saying it’s impossible that Shiller’s formula could beat the market by 1-4% over the next 20 years, my point is how can I know that? Or even have strong reason to believe it?

    Thruth, I don’t agree with the implication that the Fed rate cuts were an easy money policy. I have had posts where I provided evidence that they were actually contractionary.

    I agree that the stagflation of the 1970s was monetary policy, not oil, but probably couldn’t add much to those other studies. If it was oil, we should have had 10% inflation in 2007, instead we had around 3%.

  20. Gravatar of happyjuggler0 happyjuggler0
    24. May 2009 at 06:52

    Thruth @ 24. May 2009 at 04:43,

    I’ve long believed that the bulk of the recessionary pain on the 70’s was due to bad monetary policy based on trying to exploit “inconsistent expectations”, not “oil shocks”. However I don’t think that the “oil shocks” were completely exogenous with regards to monetary policy.

    It is worth pointing out that the world was tied to the US dollar until the early 70’s, and that in theory that any country could swap dollars for gold if they wanted to. In practice of course not enough could.

    However, looking through that lens, many people, and countries, looked at the world through a gold lens. When the US killed Bretton Woods, closed the gold window, and devalued the dollar in the early 70’s, many countries (and individuals) with large dollar “incomes” felt they were getting robbed, including Petroleum Exporting Countries.

    It is with that thought in mind, along with seemingly getting screwed by the Texas Railroad Commission (the effective precursor to OPEC) over the years, that OPEC started “feeling its oats” and raised prices in accordance with the price of gold.

    So, part of the reason that oil went up in price was the US going off gold and inflating the dollar. It is also worth pointing out that the world goes through commodity cycles, where large amounts of commodities go through boom/bust cycles at roughly the same time. (Not all of them, and not strictly in tandem though). So for oil and gold to have demand outstrip supply at the previous price isn’t likely to be entirely a coincidence. And with the dollar inflating relative to gold, and oil and other commodities being loosely correlated with each other, it isn’t really entirely correct for anyone to say that the “oil shocks” of the 70’s were exogenous.

  21. Gravatar of Carl Futia Carl Futia
    24. May 2009 at 07:03

    Scott and all:

    Here is an explanation of how two sided auction markets “work”. I call it the parable of the musical chairs.

    Let’s begin by imagining two groups of traders – I call them the low frequency traders and the high frequency traders. The low frequency traders base their buying and selling on informational events that generally are not directly observable and that happen less frequently than the events high frequency traders key on. To keep things simple let’s suppose that both low and high frequency information events are independent of one another and across time.

    Let’s see what happens when a new, low frequency piece of information is discovered by some low frequency traders. They come into the market and buy ( i.e. the information implied higher future returns for the asset being traded). Of course the market is anonymous so initially the high frequency traders meet the low frequency demand at prices that do not fully reflect the significance of the new, low frequency information.

    Now begins a game of musical chairs played by the high frequency traders. The low frequency group has removed some chairs (the asset) from the circle of chairs accessible to the high frequency traders. As a group the high frequency traders now own less of the asset than they desire to hold. Moreover, this condition will persist at least until the next low frequency piece of information is seen by low frequency traders. Why? High frequency traders have accommodated an excess demand from low frequency traders. But high frequency information arrives much more often than low frequency information. So there will be a considerable interval of time containing many high frequency events but no low frequency events. During this time average, desired, high frequency holdings will remain unchanged. This is because these several high frequency events are uncorrelated.

    So in the interval between low frequency information events the high frequency traders play a game of musical chairs. They keep trading the asset among themselves in a collectively futile attempt to bring their actual asset holding up to the desired level. In the process individuals bid up the price on each other in an attempt to restore individual portfolios to desired levels.

    Eventually the price gets high enough and/or new, low frequency information arrives to encourage some other low frequency traders to reduce their holding of the asset. This accommodates the excess demand from high frequency traders and the bidding war stops at a price higher than the one at which it began.

    If you think about it you will see that this parable has the low frequency traders making money at the expense of the high frequency traders. But you have to remember that there is information at many frequencies. Traders a given informational frequency make money from higher frequency traders but lose it to lower frequency traders. The net result, after commissions and bid-ask spreads is that traders as a group lose money on their purely speculative activities.

    Every two sided auction market has many of these games of musical chairs going on simultaneously. Needless to say, price movements are driven by changes in expectations of returns across a portfolio of assets, not by these mysterious “money flows” in or out of a given asset. There are no such flows.

  22. Gravatar of Thruth Thruth
    24. May 2009 at 08:45

    Scott: “I don’t agree with the implication that the Fed rate cuts were an easy money policy. I have had posts where I provided evidence that they were actually contractionary.”

    I didn’t say money was “easy”, but surely fed policy was “easier” than if they didn’t cut rates at all? (Isn’t this the what you are saying in your “Would you have blamed the Fed for this policy” post) I guess I’m arguing that the Fed did enough to keep demand from imploding through the early stages of the crisis. As you’ve shown us, by mid 2008 there were signs that the Fed wasn’t doing enough to keep up, and by October it was obvious.

  23. Gravatar of RebelEconomist RebelEconomist
    24. May 2009 at 10:10

    Scott, the CPI is a cost of living index, so the implied inflation expectations from TIPS is a market forecast of the cost of living. A cost of living index generally gives a lower measure of inflation than a fixed basket approach.

  24. Gravatar of 123 123
    24. May 2009 at 12:13

    “I don’t know how many money managers use those theories, but I still doubt it represents the marginal trader. Otherwise trading stocks would be like taking candy from a baby.”
    Soros and GMO are marginal traders. It is not easy to take candy from Soros and GMO.

    “And I’ve never found it easy to make money in the market.”
    It is not easy to outperform market using EMH.

    “I don’t know what Shiller study you refer to, but all market efficiency studies that I have seen use historical data. Which isn’t at all meaningful. I could come up with 100 models that beat the averages, merely by looking at what happened to key indicators in the past, and fitting my model to those stylized facts. I.e., I could check out how the market performed at various P/E ratios over time, and come up with a formula. But what would that prove? I’m not saying it’s impossible that Shiller’s formula could beat the market by 1-4% over the next 20 years, my point is how can I know that? Or even have strong reason to believe it?”
    Yes there is a risk that market might become efficient in the future and then you will have only 50% risk adjusted probability of outperforming the market by using Shiller’s formula.

  25. Gravatar of Bob Murphy Bob Murphy
    24. May 2009 at 17:29

    Forgive the self-promotion, but I’ve actually done a lot of writing on the speculator issue, and I basically agree with Scott, but I fleshed out the details. E.g. here and here.

    Having agreed with you on that point, Scott, I must protest your knock on conspiracy theories. You’re saying you actually believe the Warren Commission? Have you read anything by Mark Lane? I’m not saying I know who killed JFK, but I don’t think it was a lone gunman named Oswald.

  26. Gravatar of Thruth Thruth
    25. May 2009 at 15:15

    happyjuggler: that all sounds plausible. I wonder whether we might to see the same mistakes repeated in the (near) future.

  27. Gravatar of ssumner ssumner
    25. May 2009 at 16:17

    happyjuggler0, I agree. But there were some shocks in 1973 and 1979 that were partly exogenous, as real oil prices rose very dramatically. But I do agree that a good part was endogenous–even part of the relative price increase. Monetary policy was the main problem in the 1970s.

    Carl, Yes, it is not money flows as you say. I think the necessary price changes happen pretty quickly–so the process you describe doesn’t take long.

    Thruth, Yes, I agree.

    Rebeleconomist, I think the CPI is a fixed basket index.

    123, I didn’t say I’d be taking money from Soros. But if it is possible to identify when prices are too high relative to fundamentals, then the right strategy is to sell, and buy when prices are below fundamentals. It is not clear to me that prices are too high relative to fundamentals in the early stage of the bubble. For instance, house prices in the recent bubble where I live went up 100% and then down 10-15% at most. So I’d say in the early stages of the run-up the prices were not too high relative to fundamentals. I very much doubt Soros could have gotten rich by jumping into market when they were already viewed as frothy, or bubble-like. But regardless of what Soros did, it is clear that if it is possible to recognize when markets are way above or way below fundamentals, it would be easy to earn a very high return.

    Regarding Shiller: I am saying something more than that his formula might stop working, I am questioning if it ever worked. It means nothing to say such and such a formula correlates with past movements in asset prices–that proves nothing. People used to use a Super Bowl index to predict the stock market, and it worked something like 14 out of 15 years. But it never really “worked” it was just a fluke correlation. What I am asking is how do I know that any of these other “formula” are any better than the Super Bowl formula?

    Bob, There are three kinds of people. Conspiracy buffs. The group in the middle who only believe the well-established conspiracy theories—like the CIA overthrowing anti-American governments in Chile, Iran, etc. And then there’s me, the anti-conspiracy buff. I don’t even believe in the sort of conspiracies most others do believe in–the ones well-established in history books.

    I haven’t read Lane. Have you read “Case Closed?”

  28. Gravatar of Alex Golubev Alex Golubev
    26. May 2009 at 11:04

    i don’t think economists, wall street, or joe shmoe DO or CAN know much about the prediction business, which is closely followed by the hindsight explanation business in its complexity. I’m all three and i can argue AGAINST most theories too, but i can’t argue FOR practically anything. One of the best ways to win an argument is to argue that the other person is wrong and pretend like implicitly your theoriy thus must be right. Of course that ignores an infinite number of other possible theories. I’m not sure what the best way of finding truth is, however. :/ Then again, most things only matter to a relative extent, so being more right than someone else is equivalent to being right. Any sort of over-confidence is a tell though. This doesn’t have much direct reponse to your post, but yes you’re mroe right than the person that you quote and most predictions and explanations are moronic, unfortunately THAT doesnt’ help anyone be more right.

  29. Gravatar of 123 123
    26. May 2009 at 12:45

    “But if it is possible to identify when prices are too high relative to fundamentals, then the right strategy is to sell, and buy when prices are below fundamentals.”
    Many speculators say that if prices are too high relative to fundamentals and there are more fools rushing in to the market, the right strategy is to wait until the final moment when the bubble bursts – this is what Soros often does.

    “But regardless of what Soros did, it is clear that if it is possible to recognize when markets are way above or way below fundamentals, it would be easy to earn a very high return.” Yes, it is possible to recognize when markets are way above or way below fundamentals, but high return (extra 1-4%) is realized only in the long run (earlier than when we are dead but still) with large upside and downside deviations from market benchmark over the medium term.

    “Regarding Shiller: I am saying something more than that his formula might stop working, I am questioning if it ever worked. It means nothing to say such and such a formula correlates with past movements in asset prices-that proves nothing. People used to use a Super Bowl index to predict the stock market, and it worked something like 14 out of 15 years. But it never really “worked” it was just a fluke correlation. What I am asking is how do I know that any of these other “formula” are any better than the Super Bowl formula?”
    You have reminded me of an interesting letter to The Times (London) in 1977 from Cambridge professor, where he argued that he does not believe that money supply causes inflation because he found out that correlation between Scottish Dysyntery and UK inflation was even higher.

  30. Gravatar of ssumner ssumner
    27. May 2009 at 06:12

    Alex, Those “moronic” quotes are the way most people talk, including many commenters on this blog, and even me once and a while. I’m trying to get people out of that habit. (BTW, I don’t actually mean people are moronic.)

    123, Regarding the QTM, that is exactly right–no one should believe that money causes inflation just because of the relationship between M & P. Almost all nominal variables are strongly correlated, and many are more strongly correlated with P than the money supply is.

    My point about Soros is this: Either he can tell whether prices are higher than fundamentals or he cannot. If he can tell, then the most profiticable strategy is not to buy when they are above, but rather when they are below, and sell when above. Now you might make a bit of money buying when above and selling when even further above. But Soros would be crazy not to buy when assets are cheap–that’s where the biggest capital appreciation can occur. Of course all this assumes Soros knows the “fundamental price” which I highly doubt, but I am provisionally accepting your assumption for the sake of argument.

  31. Gravatar of Alex Golubev Alex Golubev
    27. May 2009 at 09:05

    with Cramer, CNBC, daily news, most professors (in all disciplines. less so in math.) and Yahoo Finance headlines, politicians, you’re fighting a losing battle, my friend. We’re not rational, we’re rationalizing… even if the rationalization itself is flawed. It’s not a conspiracy theory, it’s abuse of human flaws on a large scale. You can definitely blow off some steam once in a while, but don’t waste yourself on the lowest denominator. only the 2nd lowest, so i can understand 🙂

  32. Gravatar of 123 123
    27. May 2009 at 09:32

    But as you don’t disprove QTM by comparing money supply to Scottish Dysyntery, you don’t refute Shiller by comparing inflation adjusted 10 year average P/E ratios to baseball results.

    Re Soros – let’s assume he is not just lucky but has a real edge in estimating fundamental values. He has two choices – buy undervalued assets and receive profit nobody knows when, or he can buy overvalued assets that he believes will get more overvalued as more fools are rushing in and receive profits right now. He says he does both.

  33. Gravatar of ssumner ssumner
    27. May 2009 at 16:47

    Alex, Since I have so many commenters talking about “money pouring into markets” I thought I should say something about it, so I don’t have to keep repeating myself.

    123, I am not trying to disprove Shiller, I am trying to show that he has no evidence to prove his model. It’s his job to prove his model. With the QTM there is an economic theory to support the correlations, not with the anti-EMH people–it’s nothing but correlations.

  34. Gravatar of 123 123
    28. May 2009 at 10:04

    Shiller has the economic theory to support the correlations – it is the same Keynes’ “The market can stay irrational longer than you can stay solvent”. See also his beauty contest metaphor from Ch12 of General Theory. And this is exactly what Soros says he does.

    p.s. Sorry for mentioning Keynes. It is like mentioning Allah in catolic church.

  35. Gravatar of ssumner ssumner
    28. May 2009 at 17:03

    123, I consider that a psychological theory, not an economic one. Economics puts limits on one’s imagination. Without those limits, literally anything is possible. Human beings have limitless ability to create stories. I stick with theories consistent with rational behavior, until someone shows me an alternative framework that explains the world in a more coherent way.
    You are free to mention Keynes’s ideas, as long as they are from “The Tract on Monetary Reform” (1924).

    I’ll give you the last word, as I am going on vacation.

  36. Gravatar of 123 123
    29. May 2009 at 09:49

    I also stick to theories that are consistent with individual rationality together with:
    agency costs,
    information costs,
    limits to leverage,
    shorting limitations,
    margin calls,
    incomplete markets.
    EMH is a special case where abovementioned costs are zero or very small. Shiller’s correlations help us determine if these costs are important or not in practice.
    I find the worship of EMH as strange as worship of Miller-Modigliani theorem would be.

    I’ll refrain from citing Keynes here in respect for victims of cult of Keynes (Nixon – Ford era savers and others).

    Have a nice vacation.

  37. Gravatar of ND ND
    2. June 2009 at 20:33

    I don’t understand this post’s comments about the following two statements:

    “A selling wave hit the market in late afternoon, driving stocks sharply lower.”

    “investors to pour $100s of trillions into oil futures, causing prices to rise”

    It seems pretty clear what someone means when they utter one of the above. They mean that late in the afternoon a large number of motivated sellers became active in the market (without a corresponding number of motivated buyers). You could also call these people price takers as opposed to price makers.

    Suddenly there are a lot of people in the market who want to sell and are very concerned with execution speed but not with execution price. Indeed for every sale there is a buy, but those buys may mostly be done by liquidity providers (market makers, perhaps hedge funds) who are only willing to buy the shares in exchange for price concessions, so the price falls.

    Of course the same story can be told about oil futures. Suddenly there are motivated buyers of oil futures in the market (funds that have just raised capital from individual investors say). There are no natural sellers to fill this demand. Of course people will come in and provide this liquidity in exchange for price concessions, so futures prices go up. As futures prices go up and trades get done the liquidity providers find themselves with more and more short oil exposure. They do not like this. They go into the spot market to buy oil for storage in order to hedge this exposure. The price of spot oil goes up along with the futures as people hedge their futures positions (or just arbitrage the expanding spread between futures and spot).

    I do not understand what is silly about these explanations, and it seems obvious to me that this is what people mean when they talk about money pouring in or waves of sellers, no?

  38. Gravatar of ssumner ssumner
    6. June 2009 at 13:01

    123, I’d be interested in Shiller’s pronouncements on “irrational exuberance” if I could see any connection between his forecasts and the list of market imperfections that you list. How do things like agency costs and margin calls cause stocks to become 100% overvalued? I don’t see the connection. Do agency costs and margin requirements change dramatically during bubbles?

    Given that I have said many times that the EMH is not literally true, but rather is useful, I am not sure you could say I “worship” the EMH. I don’t recall many religious folks that regard “God” as a useful fiction. (Maybe I’m a Unitarian EMH supporter.)

    ND, Your explanation is either empty, or it is wrong. For instance, you use terms like “natural sellers.” What does that mean? All sellers and buyers are equally natural.

    When people suggest that a selling wave drove stocks lower, they aren’t really adding anything to the statement that “prices fell.” If they think they are adding something explanatory, then they misinterpret the forces that change prices. Suppose some bearish news at 2:00 am drives the equilibrium DOW 15% lower, but no shares change hands. Now suppose some bearish news drives shares 15% lower at 2:00 pm, and billions of shares are sold. The reason why prices fell in each case is exactly the same. The only reason why shares change hands in the latter case is that people will differ on the implications of the bearish news. If everyone interpreted the bearish news in the same way, then no shares need change hands when prices change.
    Both sides of each transaction are equally active, and equally passive. Anyone who sells shares does so because the shares are worth less to him than their market price. Vice versa for anyone who buys shares. If you try to justify the silly explanation that a selling wave drove shares prices lower by arguing that you meant there were more people who wanted to sell than buy at the pre-existing market price, then your explanation is entirely circular. Thus when people say “prices fell due to a wave of selling” all they are really saying is that “prices fell because the equilibrium price fell”. If they think their “explanation” has any more content than that, then they are confused about how markets work.

  39. Gravatar of 123 123
    7. June 2009 at 01:35

    “How do things like agency costs and margin calls cause stocks to become 100% overvalued?”

    If you short a worthless stock and it doubles before going to zero you face margin calls that can wipe you out and you also may be fired as an agent.

  40. Gravatar of ssumner ssumner
    8. June 2009 at 03:20

    123, I understand how that could restrict the extent to which a few individual investors of modest means would be able to engage in short selling. But it certainly wouldn’t stop big institutions that have tens of billions in assets from selling short. And it wouldn’t explain why the average investor had priced stocks at twice their true value.

  41. Gravatar of 123 123
    8. June 2009 at 13:48

    Institutions that have tens of billions in assets are constrained by agency problems. Each individual institution is also constrained by margin and leverage requirements.

    There is a good paper by Brunnermeier and Nagel:

    Hedge Funds and the Technology Bubble

    Abstract:
    The efficient markets hypothesis is based on the presumption that rational speculators would find it optimal to attack price bubbles and thus exert a correcting force on prices. We examine stock holdings of hedge funds during the time of the Technology Bubble on NASDAQ and find that the portfolios of these sophisticated investors were heavily tilted towards (overpriced) technology stocks. This does not seem to be the result of unawareness of the bubble: At an individual stock level, hedge funds reduced their exposure before prices collapsed, and their technology stock holdings outperformed characteristics-matched benchmarks. Our findings do not conform to the efficient markets view of rational speculation, but they are consistent with models in which rational investors can find it optimal to ride bubbles because of predictable investor sentiment and limits to arbitrage. Moreover, frictions such as short-sales constraints do not appear to be sufficient to explain why the presence of sophisticated investors failed to contain the bubble.

  42. Gravatar of ssumner ssumner
    9. June 2009 at 05:44

    123, The abstract you present supports my argument, not yours. You claim that constraints prevented the big smart players from shorting tech stocks during the bubble. So they weren’t able to prevent tech prices from getting out of hand. But the abstract you present makes the opposite argument–they claim the big players were’t trying to prevent tech stock from rising, rather they were part of the problem. The big institutions actually were holding tech stocks, thus constraints on short sales certainly could not explain any failure of the EMH. Here is what your abstract says:

    “We examine stock holdings of hedge funds during the time of the Technology Bubble on NASDAQ and find that the portfolios of these sophisticated investors were heavily tilted towards (overpriced) technology stocks.”

    That is exactly my argument. During the tech bubble even smart investors thought these were good investments. I’ll conclude with another quote from the abstract you present:

    “Moreover, frictions such as short-sales constraints do not appear to be sufficient to explain why the presence of sophisticated investors failed to contain the bubble.”

  43. Gravatar of 123 123
    10. June 2009 at 00:44

    Yes, during the tech bubble smartest investors thought these were good investments. They were overpriced, but with excellent price trend, and hedge funds were the first to stop this greater fool game.

  44. Gravatar of ssumner ssumner
    11. June 2009 at 05:15

    123, Well I don’t know the data on hedge funds. Are you saying hedge funds weren’t holding tech stock in April 2000? If so, then more power to them. But I suspect they were holding tech stocks. Is data publicly available?

  45. Gravatar of 123 123
    23. June 2009 at 11:35

    Here is another summary of what Brunnermeier and Nagel are saying about hedge fund tech holdings:

    “This article documents that hedge funds did not exert a correcting force on stock prices during the
    technology bubble. Instead, they were heavily invested in technology stocks. This does not seem
    to be the result of unawareness of the bubble: Hedge funds captured the upturn, but, by reducing
    their positions in stocks that were about to decline, avoided much of the downturn”

    The article uses Form 13F filings that are publicly available. The technology
    exposure of hedge funds peaked in September 1999, about six months before the peak of the
    bubble.

  46. Gravatar of ssumner ssumner
    3. July 2009 at 12:30

    123, Sorry I missed this reply. I’d want to see complete data. It could be that they held 30% tech in 1999 and 20% in 2000. I doubt it was all or nothing. And did they go back in too soon, say when the NASDAQ fell in half? If data is publicly available, and hedge funds outperform index funds, then why don’t managed mutual funds just mimic hedge fund holdings?

  47. Gravatar of 123 123
    3. July 2009 at 13:26

    1. “I’d want to see complete data.” Raw quarterly data is available at SEC’s EDGAR database.

    2.”And did they go back in too soon, say when the NASDAQ fell in half? ” They did not go back too soon – performance of hedge funds during the dotcom crash caused the exposive growth of the hedge fund industry.

    3. “If data is publicly available, and hedge funds outperform index funds, then why don’t managed mutual funds just mimic hedge fund holdings?”
    Hedge funds outperformed index funds during the dotcom bubble and crash. I am not saying that hedge funds always outperform index funds. My guess is that in future index funds will outperform most hedge funds (net of fees). After Lehman equity hedge funds underperformed the index even before the fees (reduced leverage and fund redemptions caused forced sales of securities). Many mutual funds analyze 13F filings of other hedge and mutual funds.

    4. Hedge funds are not the only market participants that use strategies that are described in Brunnermeier and Nagel paper.

  48. Gravatar of ssumner ssumner
    4. July 2009 at 07:50

    123, I lost the original premise of our discussion. Let me just say that I would not want to dispute that a certain segment of the investment industry might outperform index funds during specified periods. So you may be right about hedge funds in the tech boom. My views on the EMH are that such an out-performance is likely to be random and unpredictable. Otherwise everyone would copy them. So we may not be far apart.

  49. Gravatar of 123 123
    7. July 2009 at 10:53

    We both agree that EMH is an excellent theory for analyzing the merits of mutual fund and hedge fund industries. However I believe that the stockmarket was extremely inefficient during the dotcom bubble. Brunnermeier and Nagel show that the most sophisticated investors bought overpriced securities during the bubble and sold them at a profit. Shleifer and Summers wrote about limits to arbitrage in their paper ‘Noise trader approach to finance’, and tech bubble is an example that proves their thesis.

  50. Gravatar of ssumner ssumner
    8. July 2009 at 06:08

    123, One can always find, ex post, people who beat markets. In fact in any given case, 1/2 the public will be right and 1/2 will be wrong. But it is not possible to know, ex ante, which strategy works best. And your examples don’t show that. What about all the highly sophisticated investment banks that lost a fortune last year? (Bear Stearns, Lehman, etc.) This example seems to argue the exact opposite point–those “sophisticated investors” did worse than simpletons like me. But I would never say it proves the EMH. The simple truth is that there is almost no correlation between how well investors do one year, and how well they do the next. If the market could be beat by sophisticated investors, then there should be a positive correlation.

  51. Gravatar of 123 123
    9. July 2009 at 09:36

    While it is true that different investors did well in 1999 and 2000, it does not mean that market was efficient in these two years. In January 1999 it was not possible to determine when when exactly the bubble will burst, but Shillers model was indicating very poor long term returns.

  52. Gravatar of ssumner ssumner
    10. July 2009 at 15:39

    123, I agree there are models that show the market is too volatile, and I also find these to be among the more persuasive anti-EMH arguments. But that’s also true on the downside. Did Shiller tell people to buy stocks in March when the market seemed to overshoot on the downside? I’m also curious more generally whether bubble-types but out buy calls as often as sell calls.

    (I plan to do another another EMH post in a few days. So if I forget to answer these old posts, there will be more opportunities to continue this interesting discussion.)

  53. Gravatar of 123 123
    11. July 2009 at 12:02

    78 billion USD asset manager GMO that uses Shiller type models issued a buy call a couple of days before the March equity market bottom.

    Hussman who also uses Shiller type model has indicated that equity market valuation is attractive, but their short term market timing models suggested caution in March.

    Both GMO and Hussman say that March undervaluation did not reach extremely attractive 1974 and 1982 levels.

    I don’t know what Shiller said in March but right now he is saying that the stock market is now fairly valued on a long-term cyclically adjusted P/E basis, but he also notes that there are big downside risks due to precarious state of the economy.

  54. Gravatar of 123 123
    11. July 2009 at 12:13

    there is an interesting video interview with Shiller about current equity market valuation here:
    http://finance.yahoo.com/tech-ticker/article/278305/%22Precarious-State%22%3A-Stocks-Fairly-Valued-But-%22Could-Go-Down-a-Lot%2C%22-Shiller-Says

  55. Gravatar of Scott Sumner Scott Sumner
    11. July 2009 at 12:23

    123, Thanks for the video. The hosts asked him a straightforward question: “What were you saying back in March.” Shiller didn’t answer the question which I find “interesting.” I am aware that the Shiller model does fairly well when you look backward, and apply it to historical data in the US. But again, it’s not hard to develop models that work well looking backward. I’m heavily invested in stocks (although mostly out of the US) so I hope he is right about 7% a year–that sure beats the alternatives.

  56. Gravatar of 123 123
    12. July 2009 at 01:56

    gmo.com (free registration required) publishes monthly calculations of expected asset class returns according to their Shiller type models, and it might be useful to you if you have investments. The bad news is that markets are efficient enough and Shillers model is too simplistic, and 7% a year forecast has a wide variance attached to it. Shiller says that during huge macro disturbances 7% expected return market often goes down a lot and becomes even more attractive market.
    Shillers model does pretty well looking backward, but it is so simple that it is hard to argue that it is a result of overfitting or data mining.

  57. Gravatar of ssumner ssumner
    12. July 2009 at 08:17

    123, As you probably know the standard argument is that the US itself is an outlier, something like the anthropic principle is physics. Thus in the early 1930s people had no idea that the US would “bounce back.” If you look at economies that didn’t bounce back (Russia, Argentina) or stock markets that were destroyed by war or revolution (German and Japan), you might get different results.

    I do understand your point about simplicity making data mining less of a problem. But I still think most people underestimate the dangers of looking backward. The basic insight is simple, but the calibration of the model was based on historical data. That is a bigger issue that you might think.

    All the smart hedge funds on Wall Street were fully aware of Shiller’s model, but most ignored it last year. People used to cite the Harvard endowment as evidence against the EMH, but they also ignored Shiller, and lost a ton of money. I still think there is a major puzzle here as to why it is so hard to beat the market. Shiller’s answer is that it isn’t hard to beat it when it is at an extreme P/E. But then why do so many smart investors who know about his model still lose money?

    I used his model a little bit in late 2007 and early 2008, and made a little bit (or should I say cut my losses.) But I am kicking myself for not using it more completely. So I do understand the model’s appeal.

  58. Gravatar of 123 123
    13. July 2009 at 10:32

    Shiller’s model is crude and simplistic and it leaves out many important factors including tail risks, and it is revealing that such a simple proxy for a fundamental value is sufficient to prove that sometimes markets are very inefficient.

    Because it is hard to calibrate Shiller’s model, because of the uncertainty and because Shiller’s model takes only very limited information into account, people who are using such models are giving wide ranges of projected future returns. For example GMO are giving a central estimate and a range of plus/minus seven percent.

    Fear of Lenin/Hitler/Chavez type scenarios is one of the best explanations for equity risk premium puzzle, and I agree that survivorship bias distorts Shillers model. A case can be made that threat of such scenarios was greater during the 30s market bottoms, but I don’t think that in 1982 threat of such scenarios was larger than usual.

    Shiller’s model is useless for predicting one year returns, and this may be the reason why this model is not more popular
    in Wall Street.

    Harvard’s idea is an extension of EMH – they thought they have to diversify to other asset classes in order to outperform the stockmarket. Initially this idea worked very well, but in 2007 all risk asset classes became overvalued to a similiar degree.

    Harvard’s model can also be criticized from the perspective of EMH – some of the asset classes that they are using have enormous fees and according to EMH should underperform.

    You should be kicking Bernanke and not yourself for not using Shillers model more completely.

  59. Gravatar of ssumner ssumner
    14. July 2009 at 07:00

    123, Yes, I do kick Bernanke a lot, especially in this blog. 🙂

    I think what you say is reasonable. The early 1980s seemed worse at the time than they do in retrospect, but they were certainly not like the 1930s, as you correctly note.

    Stocks had done very poorly from 1966 to 1982, which is quite a long time. Of course the smart investors were those taking the long view. It is hard for people to do that when history has been moving the other way for so long.

    Does the Shiller model say there is some threshold where you should be 100% in or out of stocks? I.e. if P/E’s are more than 30% or 40% out of line?

  60. Gravatar of 123 123
    14. July 2009 at 10:33

    Model does not say that, because individual circumstances are also relevant (investment time horizon, risk tolerance, correlation of job income and economic conditions etc.). Expected returns from alternative investments (bonds, TIPS, etc.) are also very important, and GMO’s asset class return forecasts can help you with that.

  61. Gravatar of ssumner ssumner
    18. July 2009 at 05:15

    123, In a way that makes sense, but also remember that the more parameters you have the easier it is to fit it to past data.

Leave a Reply