What is a bubble?

Think quick.  You have a housing bubble.  After you recognize the problem should you respond by building more houses or fewer houses?  The US responded by building fewer houses after 2006.  The Chinese government is responding with policy initiatives to encourage more construction.  Who is right?  I say they both are.  I would also argue that this shows that the term ‘bubble’ is vague, indeed almost vacuous. Does it mean:

1.   Prices are rising fast.

2.  Prices are obviously too high.

3.  Prices are too high but the public doesn’t know it yet.

4.  Prices are too high but neither the public nor the experts know it yet.

5.  Prices are being pushed up by excessive demand.  (US housing in 2006)

6.  Prices are pushed up by insufficient supply.  (Chinese housing today)

7.  Prices are pushed up by a cartel (OPEC, 1970s)

And are there negative bubbles?  If so, what are they called?  And why don’t people talk about them?


Tags: ,

 
 
 

45 Responses to “What is a bubble?”

  1. Gravatar of Michael Michael
    2. January 2010 at 20:45

    In popular use (as opposed to academic use), it seems to imply some form of widespread irrationality and “excessively high” prices.

    I think the biggest problem with the term is that the discussion often doesn’t extend any further than that, as if the bubble itself is the problem rather than the underlying issues (such as excess demand or inefficient supply).

    Maybe thats because of when the discussion of bubbles usually takes place: just after the bubble has burst. At that point I suppose it is the bubble itself that is the problem, not whatever started the bubble.

  2. Gravatar of Bob Murphy Bob Murphy
    2. January 2010 at 21:12

    I think a bubble in this context means that demand is being fueled by anticipations of speculative gain (as opposed to people who want to buy the asset and hold it or use it, if it’s a durable consumer good). Because of this, the price is much higher than the “fundamentals” can support, where this is defined by the market price if there were no speculative demand. If and when prices even just taper off, the speculative demand disappears and then prices collapse (the bubble pops).

    My popularity in high school was a negative bubble. Everyone agreed that on paper everyone should want to hang out with me, and yet no one did. I propose defining a negative bubble as a Bobby.

  3. Gravatar of Tom Hickey Tom Hickey
    2. January 2010 at 21:50

    A search on bubble operational definition turns up the following. Some authors think that cash flow rather than price is the standard of bubble measurement.

    When is a Housing Bubble Not a Housing Bubble?-Margaret Hwang Smith & Gary Smith
    http://www.economics.pomona.edu/GarySmith/SoCalHousingBubble.pdf

    Bubble, Bubble, Where’s the Housing Bubble?
    http://www.brookings.edu/es/commentary/journals/bpea_macro/forum/bpea200603_smith.pdf

    Bubbles: some perspectives (and Loose Talk) from history
    Maureen O’Hara

    What Is an Asset Price Bubble? An Operational Definition, Jeremy J. Siegel
    http://www3.interscience.wiley.com/journal/118851313/abstract?CRETRY=1&SRETRY=0

    Bubbles
    http://bubbles.behaviouralfinance.net/

    Experienced traders try to identify market excess through predominance of momentum plays (MOMO) over value. Unsustainable upward volatility is momentum-driven. When the second derivative levels off, that it the get-out signal. By the time the first derivative levels, people are rushing for the door.

    Would it be more productive in analyzing the present financial crisis catalyzed by the putative housing bubble to examine it chiefly in terms of finance rather than price volatility?

    It seems that Minsky’s financial instability hypothesis and Irving Fisher’s debt deflation speak directly to present conditions. The underlying problem was Ponzi finance in Minsky’s sense, which was exacerbated by predatory lending practices, widespread fraud, and lax oversight, and compounded by securitization, enabled by rating agency complicity.

    It seems to me that demand dried up in the housing market chiefly because leverage collapsed, not because of increased selling pressure in the market. Prices got too high for incomes to support, so that there were too few qualified buyers at the higher prices (which many people would still have gladly paid).

    All traders know that there are negative bubbles and they usually follow positive bubbles, because markets tend to overcorrect and the greater the excess on the upside the more extreme the correction on the downside. And the bigger and quicker the bounce.

    As far as building more houses in response, building take a long time to get going and a long time to stop. When the bubble burst, a lot of development was already in the pipeline and a lot of contracts signed. So building continued for quite a some time “” at least six months. Further problems developed when developers were pressed by carrying charges (“alligators”) and had to unload at a discount or loss, or else default. That further depressed prices.

  4. Gravatar of Darren Hom Darren Hom
    3. January 2010 at 00:11

    Good question. Here’s my take on it:

    A bubble exists during the final phase of a long-term market cycle in which highly uneducated investors rush in (out) of a market. Once the supply of these investors has been exhausted, there is nothing left to continue holding prices high (low).

    By the time these uneducated investors are jumping in (out), prices are usually quite distant from where fundamentals indicate they should be. So fundamentals may be helpful in identifying a bubble but not as helpful in choosing entry and exit points.

    This above is quite unhelpful in the actual process of timing and taking advantage of bubbles. Once a heavily overbought or oversold market has been identified, I like the 200-day moving average.

  5. Gravatar of Darren Hom Darren Hom
    3. January 2010 at 00:15

    Example:

    This past month, two insurance company reps tried to sell me whole life policies that contain T-bonds. The reasoning was that T-bonds are not speculative and that stocks and baskets of currencies are. I had to explain that T-bonds are actually speculative, that they historically have been nearly as volatile as the S&P 500, and that they tend to fall in value when long-term interest rates go up.

    These brokers told me that they have had a lot of success selling these structured products because the 2008 crisis has caused people to want safe, risk-free investments.

    That month, I started watching the 200-day moving average.

  6. Gravatar of jsalvatier jsalvatier
    3. January 2010 at 01:29

    I agree with your implication that people talk about bubbles as if it is a well understood phenomena (or even claiming that they understand them well, which they should do because there are certainly people who do not understand bubbles well). There’s some interesting experimental economics work on bubbles which I would be interested to see people discuss more; I think one of the main results is that bubbles are pretty easy to create even in circumstances where they are obvious (future asset cashflows are fully known). Also, I think Tyler Cowen was throwing around the idea of a negative bubble some time back.

  7. Gravatar of 123 123
    3. January 2010 at 01:50

    Tyler Cowen during the oil bubble of 2008:
    “If Krugman’s cited data don’t do the trick, why do I agree with his conclusion that speculation is not the villain? The simplest alternative story, again blogged by Arnold, is that the earlier low price of oil was an anti-bubble of sorts and one which now has been corrected by market forces. It was a kind of collective blindness, akin to the view that real estate prices would continue rising in value. No, I can’t prove that is true but I find it the most plausible story, with p (truth) = 0.57.”

  8. Gravatar of 123 123
    3. January 2010 at 01:50

    source for Cowen quote:
    http://www.marginalrevolution.com/marginalrevolution/2008/06/exasperating-pa.html

  9. Gravatar of Doc Merlin Doc Merlin
    3. January 2010 at 04:18

    1. I uncharacteristically disagree with Bob here: A bubble is when gains or prices are significantly above the “long term” exponential trend. You see this by looking at the long term trend and seeing the bubble collapse. Speculation alone doesn’t produce a bubble, only incorrect speculation.

    Everyone likes to blame speculation for bubbles, but they are very hard to distinguish from shocks.

    2. Econophysicists, finance people, and experimental economists have studied bubbles, a lot in fact. They have several characteristics, one of them is faster than exponential growth and log periodic power scaling. Lin, Ren, and Sornette have a pretty good economophysics paper on the shape of bubbles, where they derive the shape commonly seen in bubbles once using EMH also with a completely different method.

    Here is the paper. Note: Not for the faint of heart. http://arxiv.org/abs/0905.0128v1

    3. People do talk about antibubbles, just not macro people. The LPPL model also works for the longer than “very short term” antibubbles. Anti-bubbles are a bit rarer as those with power prefer high prices to low prices, so they tend to make short positions harder/more_costly to maintain than long ones. Even so, we see anti-bubbles all the time, we just don’t usually call them that. Look at the crash in 2009, it has all the characteristics of an explosive ant-bubble.

    4. I disagree that trying to prop up bubbles is a good thing. It doesn’t allow proper price seeking and then shifts micro incentives so that individuals look to government for their revenue stream. This imo necessarily worsens the collapse later.

    5. http://bubblehunter.blogspot.com/
    …just one blog for bubbles and antibubbles

    6. Bubbles are scale invariant so small, medium, and large (anti) bubbles all share the same characteristics.

  10. Gravatar of Doc Merlin Doc Merlin
    3. January 2010 at 05:30

    7. You don’t need financial speculation for bubbles to occur, a high time-inconsistent discount time preference rate (like happens in fads) or high barriers to entry or high costs to change behavior are enough to create them because economic actors must work on expectations not on actual future knowledge.

    8. As to the question:
    “More houses or less houses.”
    Neither, it is a bad question. Bubbles can arise in because quantity supplied cannot ramp up fast enough to meet quantity demanded before the structure, preferences, or other factors shift because prices are too very high relative to what they could be with some changes in structure. By the time one notices a bubble, its way too late to fix it.

    9. Over long time periods, demand and supply curves are not independent, but over short time periods they are. This itself makes bubble-like behavior:
    Another way to say this is: Long term demand and supply often have much higher elasticity than short term demand and supply.

    Example:
    Oil prices rise heavily due to slowly decreasing supply in a few countries. This causes the price to skyrocket, because the demand for oil is highly inelastic. In the long term, the high price causes people to increase efforts to look for more fuel efficient cars (a time costly expense) which then drops the demand curve (not just the quantity demanded) downward.

    Another real world example (this actually happens in SF):
    In San Francisco housing prices are bubble very frequently. Housing supply in San Francisco is very inelastic, due to the small amount of space and the regulatory overhead required to build new houses. As housing prices rise the expected profits on house building/sell increases, and more people apply for permits and prepare for the building, and prepare to move. Building takes a while, and there are high costs to moving so during this time housing prices stay high. Finally, eventually, the price crashes again.

    I am not just confusing demand/supply with quantity demanded/supplied. Modern econ ignores the effect of price changing elasticity because it looks at things mostly without understanding the role time plays.

    10. Because of 7. and 8 the best thing a government can do to fight bubbles isn’t to try to deflate them or to try to prop them up, but to make a good legal system that reduces barriers to entry, and helps make transfer costs small.

    Oh, Its way too late/early, I am going to bed.

  11. Gravatar of David David
    3. January 2010 at 08:07

    The answer is 5, with some elaboration on the term “excessive.” Seller’s interests (e.g. National Associaton of Realtors) won’t admit a bubble exists. Instead they will come up with an explanation; a “new paradigm” for the anomalous price appreciation. In my opinion the birth of this explanation is sufficient evidence of a bubble. You saw it in real estate and in dot com stocks. The simplest explanation, which worked with tuilp bulbs and seems to work well even today, is the expected price appreciation itself as casue for high demand. Sometimes the explanation doesn’t even have to make sense (e.g. “They ain’t making any more land.”).

    Another characteristic of a bubble is a shift in resources and entrepreneurial activity towards exploiting it. When becoming a mortgage agent or a house flipper or a Beanie Baby collector is the hot thing to do, but historically it wasn’t, that is strong evidence of a bubble.

    If a bubble may be defined as new, urgent economic activity developed to take advantage of expected price appreciation, then perhaps a negative bubble is rushing to act before prices decline. For example, the Gold Rush. There’s a lot of gold out West, and if it all gets found prices will come down, so I better get there first.

  12. Gravatar of Greg Ransom Greg Ransom
    3. January 2010 at 08:55

    No. Let’s make sure to think like, well, economists.

    BUILDERS responded. BUILDERS built fewer houses.

    Scott wrote:

    “You have a housing bubble. After you recognize the problem should you respond by building more houses or fewer houses? The US responded by building fewer houses after 2006.”

    Compare:

    “The Horse (i.e a nameable Platonic entity) has a population spike. After The Horse recognizes the problem should The Horse respond by producing more offspring or fewer offspring? The Horse responded by producing fewer offspring after 2006.”

    But horses are populations made up of individuals — and the characteristics of the population are understandable in terms of the doings of the individuals within it. There is no Platonic kind and entity nameable as “The Horse” and acting as a causal agent.

    To talk about what The Horse did reifies and animizes something which isn’t even a natural kind — and the the process misleads one into believing this “Platonic” non-agent can bear the weight of ascribed explanatory causation.

    But in fact, this explanatory causal efficacy is an illusion — creating false “understanding”.

    Members of populations interact — there ain’t no Platonic entity or kind here acting or interacting with anything. It’s all a crude explanatory blunder caused by misusing language on the model of primitives ascribing reified and animistic causation to The Forest, or whatever.

  13. Gravatar of Greg Ransom Greg Ransom
    3. January 2010 at 08:57

    In other words, saying “the US responded” encourages one to think like a constantly blundering Keynesian macroeconomist, instead of like a real economist.

  14. Gravatar of Greg Ransom Greg Ransom
    3. January 2010 at 09:01

    It’s easiest to understand “Housing Bubble” is a simple indexical — go to Orange County, CA and look around. Talk to some people. Read the newspaper.

    Get out of the Ivory Tower.

  15. Gravatar of Greg Ransom Greg Ransom
    3. January 2010 at 09:14

    Scott — I charge you again with a failure to disaggregate, i.e. for failure to think like an economist.

    “4. Prices are too high but neither the public nor the experts know it yet.”

    We’re talking about heterogeneous populations here, not Plantonic entities. Different people understood things different — some “experts” had no clue, more than a handful of others did have a clue.

    It makes a difference that the population changed over time, as did the characteristics of the members of the population. Taking this into consideration is what it means to be an economists. Failing to do so make one, well, how do you like the term “Keynesian shaman”?

    Scott writes:

    “5. Prices are being pushed up by excessive demand. (US housing in 2006)”

    Haven’t we made enough fun of the blackboard non-contextualized concept of “demand” to know how ridiculous and unhelpful this sentence is? A mere 5th wheel, write out a “let’s make it easy to grade” textbook.

  16. Gravatar of Greg Ransom Greg Ransom
    3. January 2010 at 09:20

    Talking to people on the ground matters, i.e. getting a sense of what people knew and when they knew it re: the housing bubble.

    Talk to people in the mortgage origination business — many of them saw what was happening. All sorts of people on Wall Street saw bits of what was happening.

    A few home buyers clearly knew they were essentially renters, who were living beyond their means at what turned out to be zero or no rent (they never paid a penny on their liar loan, or took far more cash out of their house than they ever put in, etc. — read back issues of the Orange County Register, full of such stories).

    And I could go on and on.

    Knowing the on the ground facts rather than merely posited blackboard Ivory Tower “facts” and some downloaded statistics makes a difference in understanding what happened.

  17. Gravatar of scott sumner scott sumner
    3. January 2010 at 10:07

    Michael, That is my problem, the term ‘excessively high’ is hopelessly vague. In any market, bubble or not, the seller of an asset thinks it’s worth more than he paid, and the buyer thinks it is worth less.

    Bob, But how do you know if the expectations of speculative gain are rational? For instance in China nominal incomes have been growing at 12% to 15% a year. Obviously that suggests that Chinese house prices will be much higher in the future. So why is the anticipation of gain a sign of a bubble? Isn’t that how markets are supposed to work?

    Tom, Interesting points, but I still haven’t had anyone address the issue of whether bubbles imply we should build more houses or fewer. I am not a fan of many of the bubble theories you cite, as they violate the EMH. In several other posts I defend the EMH.

    Darren, But once again, my interest is in the implications for output. Should the Chinese build more houses or fewer, if they have a housing bubble? I agree that prices are unpredictable during bubbles.

    Darren#2, I am also a contrarian. I love to buy things when everyone I talk to thinks they are horrible investments.

    jslavetier, I agree that they can be created in the lab. But these experiments imply there are easy algorithms one could use to beat the market. But in the real world there aren’t. So it’s not clear the lab results apply to the real world, where much more money is at stake.

    123, Thanks, I agree the high oil prices were not irrational. But not knowing the definition of bubble, I can’t say whether they were a bubble.

    Doc Merlin, The problem is that the trend line may not be reliable. In the 1980s I seem to recall that NYC condos rose above trend, but there was no bubble, indeed the prices were probably still too low.

    You said;

    “4. I disagree that trying to prop up bubbles is a good thing. It doesn’t allow proper price seeking and then shifts micro incentives so that individuals look to government for their revenue stream. This imo necessarily worsens the collapse later.”

    Who advocates this policy?

    You said;

    “8. As to the question:
    “More houses or less houses.”
    Neither, it is a bad question. Bubbles can arise in because quantity supplied cannot ramp up fast enough to meet quantity demanded before the structure, preferences, or other factors shift because prices are too very high relative to what they could be with some changes in structure. By the time one notices a bubble, its way too late to fix it.”

    In this case the long run adjustment is not likely to be less demand, but rather more supply.

    David, OK, but as I mentioned earlier, it is not clear when expectations of further price gain are rational, and when they are irrational.

    greg, You said;

    “No. Let’s make sure to think like, well, economists.”

    You seem to think more like a philosopher than an economist. I just wanted to toss this grenade out there to see what sort of reactions I’d get. I am surprised there are so many comments. I have never regarded the bubble concept as being interesting.

    The most important characteristic of the sub-prime crisis wasn’t excess production of houses but rather excessive refinancing. That’s where the big losses were. The problem was people who already had houses, and had lots of equity in their houses, refinancing in such a way as to extract all that equity.

  18. Gravatar of David David
    3. January 2010 at 11:36

    The rationalization of decision-making that would previously have been considered irrational is exactly my point. In NYC during the bubble, many rental properties were selling with negative cash flow. Cap rates were less than the 10-year Treasury yield. That is an ahistorical, if not totally unprecedented view of rental property. Justifying it requires paragraphs like “things are permanently different now because of X.” I suppose X could be rational, but wrong. It’s the novelty of X, and sometimes the source of X, that interests me.

    I don’t think Chinese real estate values that keep pace with income growth is a sign of a bubble. In many parts of the world in the last decade values completely outpaced income growth, and the explanations offered for that were in my opinion not compelling.

  19. Gravatar of JimP JimP
    3. January 2010 at 13:03

    off topic – as usual

    Maybe this explains it:

    http://www.washingtonpost.com/wp-dyn/content/article/2009/12/31/AR2009123101532.html

    Robert Kagan is described in this Post article as thinking that “Obama has given up before even playing, simply managing what he sees as the country’s inevitable decline into a “post-American” world. Worse, by trying to be a convener of nations and a “friend to all,” Obama is casting aside old alliances and accommodating rivals such as China and Russia.”

    Maybe that is it. Obama is managing decline. As is Bernanke.

    I just have never understood why Bernanke is doing this. He knows better. He has said so directly. He knows that price level targeting is better. But perhaps both he and Obama just think we are sort of doomed. We are in decline – and are just getting what we deserve. Declinists are after all deflationists – and maybe this is what they want – or at least when they deeply expect. Decline.

    Why else would Obama simply ignore the simple fact that he does have two empty seats on the Fed board. He could fill those with optimists. But both he and Bernanke are just to pessimistic and sad to do so. They have just given up.

  20. Gravatar of Doc Merlin Doc Merlin
    3. January 2010 at 13:20

    “Doc Merlin, The problem is that the trend line may not be reliable. In the 1980s I seem to recall that NYC condos rose above trend, but there was no bubble, indeed the prices were probably still too low.”

    Its not just the trendline that shows you a bubble. The log periodic behavior also is there. As the bubble grows, the oscillations in the trend-line decrease in amplitude (or so argue Lin et al.).

    “who advocates this policy.”
    Most of congress it seems with increasing the tax incentives on buying a home.
    The federal reserve by buying up “toxic” assets at higher than face value.
    The federal government by subsidizing house buying through the GSEs.

    “In this case the long run adjustment is not likely to be less demand, but rather more supply.”

    For ones that aren’t based on financial speculation, that is very true.

  21. Gravatar of Greg Ransom Greg Ransom
    3. January 2010 at 13:25

    Scott, there were both in OC, CA and in other parts of SoCal. Read the local papers or the blogs (Calculated Risk, etc. House builders boomed and busted, construction employment ditto. You seem simply not to be informed about facts on the ground.

    “The most important characteristic of the sub-prime crisis wasn’t excess production of houses but rather excessive refinancing. That’s where the big losses were. The problem was people who already had houses, and had lots of equity in their houses, refinancing in such a way as to extract all that equity.”

  22. Gravatar of Greg Ransom Greg Ransom
    3. January 2010 at 13:28

    Scott, the highest foreclosure rate in OC, CA is in a new construction zip code.

    OC, CA is ground zero in dollar terms of the housing boom and bust.

  23. Gravatar of Bob Murphy Bob Murphy
    3. January 2010 at 17:50

    Scott,

    You asked something like, “How do we know if the speculative demand is rational or irrational?” My answer is that you can check to see if they were right about the “fundamentals” rising (assuming we can define those, which I think we can at least imperfectly).

    So e.g. if I think Iran is going to get bombed next month and I go long oil futures, I push up the price and people might think, “What the heck? Output is still strong, why is oil going so high and getting added to inventories? That’s unsustainable!”

    But then if I’m right about the war, those inventories will come back down, and oil prices will have to be $200 a barrel to match quantity supplied with quantity consumed in the new equilibrium. So at that point, it would be clear that I did the right thing in my speculation.

    Now I think maybe you can get really philosophical and ask, “Suppose you were wrong in your forecast, and after it was obvious you were wrong, you closed your positions and the price fell back down. Was there a bubble in the interim, or do we call that something else?”

    Strictly speaking, I would call it something else, like a disequilibrium price caused by faulty speculation. I think the true essence of a speculative bubble is when people buy now expecting the price to rise, not because they think the market will eventually catch on to a trend in the fundamentals that the speculator has recognized earlier than others, but simply because prices apparently can get caught in a self-fulfilling prophecy.

    So in practice, you’re right, we can’t tell if apparently-too-high prices are a bubble, or a sign that speculators know something we don’t. But I thought you wanted to know in principle how to define a bubble, not how to identify it in progress.

  24. Gravatar of ssumner ssumner
    4. January 2010 at 06:43

    Greg, How do the “facts on the ground” allow you to know whether the fall in the housing market is due to a bubble popping, or to NGDP falling?

    David, If you don’t think China is a bubble, then that supports my point. Most people do regard China’s housing market as a bubble. My point is that people do not have any consistent definition. It is no good to point to an extreme case as a definition of bubbles. Even if you are right about NYC, it is an example, not a definition.

    JimP, If Obama is trying to manage decline in the US economy, his time in office will last precisely as long as Jimmy Carter’s did.

    Doc Merlin, I misunderstood your point about those advocating propping up the bubble. I entirely agree.

    Greg, Check out this post:

    http://www.themoneyillusion.com/?p=3255

    Bob, I don’t see how we can identify good and bad reasons for price rises. If the market consensus believes prices are rising for good reasons, and the experts think the reasons are bad, then how can we judge who’s right? In my view the markets are right much more often than the experts.

  25. Gravatar of David David
    4. January 2010 at 07:10

    I didn’t think my point was that subtle. I said “I don’t think Chinese real estate values that keep pace with income growth is a sign of a bubble.” That doesn’t mean China is not in some kind of a bubble. I haven’t been paying attention to China much so I don’t have an opinion.

    Of course my NYC example–which I am right about–is an example. What I said was, a necessary condition for and strong evidence of a bubble is that a new story be crafted to explain why the old way of valuing particular assets are obsolete. I gave more than one example, so I think its unfair to say I focused on an “extreme case.” Are bubbles not an extreme case anyway?

    A bubble is a (mass) psychological phenomenon and probably has an similar vague style of definition as a psychological diagnosis, like the presense or absence of some subset of possible traits. I’m proposing one trait, and rejecting the idea that a bubble depends on who knows about it and when.

    Even something as well-defined as a recession is nothing more than consensus opinion until it meets the standard post-hoc. Why should “bubble” be held to a higher standard?

  26. Gravatar of JimP JimP
    4. January 2010 at 07:45

    Scott

    Exactly so. Obama does seem like Carter to me. Sad, like Carter was. And supine in front of Congress.

    We need the optimistic Roosevelt. Where is he? I have the impression that, now that he is President, Obama does not actually like the job all that much.

    Expectations really do matter.

    This is a quote from the Kagan article:

    begin quote
    Obama and his foreign policy team have apparently rejected two of the main pillars of this post-World War II strategy. Instead of attempting to perpetuate American primacy, they are seeking to manage what they regard as America’s unavoidable decline relative to other great powers. They see themselves as the architects of the “post-American” world. Although they will not say so publicly, in private they are fairly open about their policy of managed decline. In dealings with China, especially, administration officials believe they are playing from a hopelessly weak hand. Instead of trying to reverse the decline of American power, however, they are reorienting American foreign policy to adjust to it.
    end quote

    Note especially this sentence:

    “They see themselves as the architects of the “post-American” world. Although they will not say so publicly, in private they are fairly open about their policy of managed decline.”

    The whole Obama administration seems to be on board for managing decline. And if we do get Bernanke’s high unemployment goal this decline meme will really be out there.

    If they want decline they can sure deliver it.

  27. Gravatar of OGT OGT
    4. January 2010 at 08:57

    Sumner- I use two definitions of a bubble. One is the pundits/analyst definition, which is any market or commodity that the pundit thinks is due for a sharp fall for whatever reason. This is the most common usage.

    The other is more economic(ish), a bubble is any sustained, substantial deviation of asset prices from the ability of the asset to produce revenue streams. Generally one sees ‘ponzi’ financing in bubbles in which the investment is only validated by rising asset prices with no hope of making debt payments based on the income. (This goes for non-debt financed investments too as the opportunity cost of one’s own equity must be considered too).

    David- But are Chinese incomes keeping pace with real estate prices, not according to the Chinese government:

    Other analysts also see a bubble, at least in terms of affordability. “Even Chinese government statistics point to real affordability problems, with the income-to-price ratio in Beijing hovering at 1:22, when the IMF and the UN say the ideal figure is 1:3 or 1:4,” said Ashley Howlett, head of China construction practice for Jones Day. “The fact is that the average people cannot afford to buy apartments in Beijing or other major cities.”

    http://edition.cnn.com/2009/BUSINESS/12/30/china.property.bubble/

    On the other hand the Economist last week claimed the price to rent ratio naitionwide is near its long term average. So, perhaps it is just frothy, as the Maestro would say.

  28. Gravatar of David David
    4. January 2010 at 10:32

    OGT — I guess I misread Scott’s response. I thought he claimed values were increasing inline with income growth. I really have no clue about China. If there’s a new and vastly higher multiple on China real estate, the explanation why could indicate a bubble.

    bettor — Rent control and stabilization has been present in NYC since WW2. If it’s a distortion, it’s a nearly constant one. Rent control didn’t seem to prevent positive risk premiums for condos/co-ops before 2005. Anyway the percentage of rent stabilized apartments is shrinking, not growing and doesn’t apply much to the condo market where virtually every rent is over $2000 a month and therefore exempt. The existence of rent control doesn’t invalidate my example.

  29. Gravatar of caveat bettor caveat bettor
    4. January 2010 at 13:21

    great discussion, but let’s throw out the NYC condo data set, ok? Remember, 2/3s of the rental market in Manhattan is subsidized (via rent control and stabilization policies). There’s enough noise without adding more distortions.

  30. Gravatar of David David
    4. January 2010 at 15:49

    There seems to be a comment timestamp issue. I didn’t anticipate bettor’s comment 3 hours before he posted it.

  31. Gravatar of Simon K Simon K
    4. January 2010 at 18:12

    My personal definition of a bubble is when “The Economist” magazine finally concedes that consistently inflated asset prices may represent a real sectoral shift and not a bubble, that’s when you know its a bubble. This predictive method was correct for both the housing bubble and the dot-com bubble, so it has a better track record than any other I know of 🙂

    More seriously – bubbles are one of those things everyone knows when they see, but are incredibly hard to define. As Doc Merlin said above, its really very difficult to distinguish a bubble bursting from a shock. In both cases investors may be very heavily leveraged, in the expectation of future price shifts, and in both cases their expectations will be wrong and will result in ripples of insolvency through the financial system. This is why bubbles are usually only clearly identified in retrospect – if the expectations turned out to be correct and the crash never came it wasn’t a bubble.

    But that takes me back to the “know it when you see it” thing. What makes people suspect a bubble is the inability of investors to explain precisely why the asset in which they’re investing in is going to gain in value beyond pointing to historical trends and claiming that other people are doing the same thing. When this strategy comes to dominate the marginal prices in a particular market, that’s a bubble. But of course there’s no way to determine what’s going on in investors’ heads by just looking at the numbers so its not a very useful definition.

    I read a great deal of discussion during the housing bubble, growing in volume as it started to burst, between people who believed that you could tell it was a bubble just by looking at the numbers (The Fundamentals in bubble-caller-speak) and that on that basis prices had been inflated since some date usually ranging from 2004 to 2001, 1997, or 1776, depending on the (often questionable) sanity of the poster.

    The problem with this sort of thing is that The Fundamentals it appeals to are either not really that fundamental, or not knowable. For example, one common ruler stick is to compare the monthly payment on a 30 year fixed mortgage for average house price and compare it with the average monthly income (looking for the 30% ish ratio mortgage lenders look for) or the average monthly rent (looking for rent to be higher). But as metrics these have all the advantages of theft over honest toil – they don’t account for medium term changes in population, changes in housing stock, changes in median income, or the relative proportion of rental versus for sale properties, let alone the tax, inflation and interest rate tradeoffs between owning and renting.

    But those things are the details in which the devil of whether an increase in prices is a bubble or a sustained shift lie. For example – think of an old, low density suburb that’s near to rapidly growing industries but where the local region has very limited scope for expansion. You’d expect prices to rise at least until the area was reconfigured for higher density, which (given the nature of local government) may just never happen. These things are not captured if you look only at rent-to-mortgage or mortgage-to-income comparisons, even if you do them properly.

  32. Gravatar of OGT OGT
    4. January 2010 at 19:51

    Simon KBut those things are the details in which the devil of whether an increase in prices is a bubble or a sustained shift lie. For example – think of an old, low density suburb that’s near to rapidly growing industries but where the local region has very limited scope for expansion. You’d expect prices to rise at least until the area was reconfigured for higher density, which (given the nature of local government) may just never happen. These things are not captured if you look only at rent-to-mortgage or mortgage-to-income comparisons, even if you do them properly.

    That’s not true of price to rent ratios, the rents would be just as affected by housing shortages as prices. In fact, if rents do not rise that’s a dead give away.

  33. Gravatar of Doc Merlin Doc Merlin
    4. January 2010 at 21:42

    “That’s not true of price to rent ratios, the rents would be just as affected by housing shortages as prices. In fact, if rents do not rise that’s a dead give away.”

    That was what tipped off Peter Schiff actually, that it was a speculative bubble.

  34. Gravatar of Simon K Simon K
    4. January 2010 at 21:45

    OGT – Its true that price-to-rent ratios have some value, but they ignore two things that matter, especially in the highest cost areas. Rented and owner occupied properties aren’t homogeneous, and real price-to-rent ratios are personal. The first point is obvious – rental properties are mostly apartments and get (and are usually designed for) minimal maintenace. Owner-occupied homes are mostly SFH and higher maintenance. The second point is less obvious, but true housing cost for owner occupiers depends on tax bracket, downpayment amount, credit score and interest rate and inflation expectations. Any metric that doesn’t account for this is not a useful guide. In particular you can’t look at one price-to-rent number to assess whether an area is over/under valued.

  35. Gravatar of Greg Ransom Greg Ransom
    4. January 2010 at 22:21

    Scott,

    “Falling NGDP” didn’t force my neighbors out of their houses in 2006 and 2007, it didn’t put prostitutes on my very short street as part of a mortgage origination scam, it didn’t encourage more people than I can name to buy more house and more Home improvement” than they could handle. It didn’t cause mortgage origination firms and WaMu and Countrywide to give liar loans to mariachies “secured” by a
    photo of a man in a mariachi suit — the insane stories in OC, CA were the subprime industry was invented and had it home are endless, repeated on the OC Register, on CNBC and told by word of mouth by people you run into every day.

    The burden of proof is upon people who seem not to have any knowledge of any of the massive wealth of information that tells one story — and not another.

    The game of “shifting the burden of proof” is ridiculous when the burden flipper seemingly has only the slimmest of background knowledge or understanding of the phenomena at issue.

    Scott writes,

    “Greg, How do the “facts on the ground” allow you to know whether the fall in the housing market is due to a bubble popping, or to NGDP falling?”

  36. Gravatar of caveat bettor caveat bettor
    5. January 2010 at 05:11

    David, there may be some comment latency. I skimmed all the comments before posting mine, and did a word search on ‘condo’.

  37. Gravatar of caveat bettor caveat bettor
    5. January 2010 at 05:13

    Ah, I see, David, you replied to my comment, but you appear to have time traveled ahead me. Neat trick.

  38. Gravatar of Greg Ransom Greg Ransom
    5. January 2010 at 08:50

    Here’s the mariachi / liar loan story:

    http://www.nytimes.com/2008/12/28/business/28wamu.html?_r=1&hp=&pagewanted=all

    Part of a NY Times story on the lending standards of WaMu.

    The OC Register must have published a hundred stories like this in 2006, 2007, and 2008.

  39. Gravatar of Greg Ransom Greg Ransom
    5. January 2010 at 08:55

    Scott, everything that happened “on the ground” is at once part of the “trace data” that must be explained and part of the evidence supporting a classification of the overall causal event.
    (See Larry Wright’s work on the nature of explanation).

    The bare graphs and crude “macro” data and supply / demand inventions of the economists have almost no meaning absent all of the background knowledge and understanding of our on the ground learning and experience.

    And gain, let me emphasize that the burden on proof is on the macroeconomists to prove they have any idea what they are doing with their bare graphs, crude data sets and their constantly failed “models”.

  40. Gravatar of scott sumner scott sumner
    6. January 2010 at 12:10

    David; You said;

    “Even something as well-defined as a recession is nothing more than consensus opinion until it meets the standard post-hoc. Why should “bubble” be held to a higher standard?”

    There are no serious economists who question the existence of recessions. There are lots of serious economists (including me) who have doubts as to whether bubbles exist. Or perhaps a better way of expressing my view is that I don’t believe bubbles are a useful concept, but I do believe recessions are a useful concept.

    JimP, I agree that the lack of optimism is very worrisome. Of course you don’t want recklessness, but without any hope life is hardly worth living. They should have a plan to get unemployment below 7% by year end. But where is the plan?

    OGT, If that’s the definition, then people should say “overvalued” not bubble. The meaning would be clearer. You said;

    “Other analysts also see a bubble, at least in terms of affordability. “Even Chinese government statistics point to real affordability problems, with the income-to-price ratio in Beijing hovering at 1:22, when the IMF and the UN say the ideal figure is 1:3 or 1:4,” said Ashley Howlett, head of China construction practice for Jones Day. “The fact is that the average people cannot afford to buy apartments in Beijing or other major cities.”

    The average person hasn’t been able to afford a house in San Francisco for 30 years. The price to income ratio is absurdly high. Any guesses as to when San Francisco house prices fall to the level where the average person will be able to afford them? Here’s my guess:

    Never.

    Beijing is the capital of what will soon be the greatest country on earth (in terms of GDP) I’m guessing real estate will do very well there over the next few decades. Right now prices are dirt cheap by the standards of Tokyo, NYC, London, etc.

    Simon K, You said;

    “More seriously – bubbles are one of those things everyone knows when they see,”

    If so, why do so many people lose money in bubbles?

    Every sports fan knows a hot streak when they see it in sports. But statistical studies suggest that hot streaks don’t exist. And hot streaks even have the same psychological correlates (euphoria.)

    As far as your statistical metrics for housing bubbles, those would have all predicted that San Francisco prices would have to fall back to earth. But 30 years later they are still very high.

    greg, You said;

    “Falling NGDP” didn’t force my neighbors out of their houses in 2006 and 2007,”

    That’s because NGDP didn’t fall in 2006 or 2007. But when NGDP did fall in late 2008 and 2009, it did cause a double dip in real estate prices already depressed by the subprime crisis. The data is very clear. Even parts of the country that avoided the sub-prime fiasco saw price declines after August 2008. And when NGDP stopped declining, real estate prices stopped declining in most markets.

    Before you make fun of my story, you might want to check out what my story is. I must have had a hundred posts where I said that falling NGDP did not cause the housing bust of 2006, 2007, and the first half of 2008.

  41. Gravatar of David David
    6. January 2010 at 21:15

    Scott, I really feel like you’re not giving me a fair reading, repeatedly. Did I say recessions were not a useful concept? No. What I said, or meant to say, was that in the first two quarters of a recession there can be disagreement about whether the economy is in recession or not. Once the data is in it becomes “official.” Similarly, we can argue about bubbles until they meet some post hoc standard. Like any economist who thinks he can spot a recession coming, I think I can see certain bubbles before they’ve popped. It’s an opinion and therefore can be wrong. Maybe we can agree on a post hoc standard so we can eventually settle the argument.

    After the fact, were tulip bulbs, South Sea charters, and dot coms not bubbles? Does NGDP explain CMGI and pets.com? In 2000 when the world’s central banks said it was a new age and sold all their reserve gold at rock bottom prices, was that based on sound, longstanding economic principles?

  42. Gravatar of Simon K Simon K
    7. January 2010 at 15:26

    Scott – I agree with you that the statistical metrics for housing bubbles don’t work. If I thought they worked, I wouldn’t own a house in an expensive suburb of San Francisco, would I?

    But I have a “rational” reason for my “investment strategy” (not that I expect to actually make any real return from it beyond living in it). We have an under-supply of housing, no land to build on, and an over-abundance of newly rich people. So no bubble (much to the disappointment of some).

    What I was driving at before is that maybe you can identify bubbles from the strategies and beliefs of market participants. If the market is dominated by leverage and momentum plays, but there’s no associated expectation of an increase in real returns, that would seem to spell bubble. Except of course its hard to measure expectations regarding real returns, especially for something like housing thats not a purely financial good.

  43. Gravatar of ssumner ssumner
    7. January 2010 at 19:58

    David, I think I do understand you. I am not trying to blow smoke. I am saying that recessions are obvious ex post, bubbles are not. You said:

    “After the fact, were tulip bulbs, South Sea charters, and dot coms not bubbles?”

    There is a famous book by a distinguised economist arguing that the Tulipmania was not a bubble. And I seem to recall Fama arguing that the dotcom bubble was not a bubble. I’ve never heard anyone argue this downturn is not a recession.

    How can we know if “event X” is a bubble, when no one can seem to agree on what the term means? My commenters give me all sorts of definitions, yet I don’t think they realize how inadequate those definitions are. Things like “out of line with fundementals” or “prices are only justified by expectations of further gains” These are worthless definitions. No one knows the fundamentals, and almost every biotech start-up is supported by nothing more than expectations of capital gains.

    Simon. The problem is that there must be some people who think there are real factors that support higher prices, or the prices would not be bid up. A person cannot lift himself up by his bootstraps, and neither can a market.

    I admit there are things I can’t explain, like Vegas and Phoenix, but someone must have believed their was a reason, or else why would they have paid such high prices? Why not just rent two years until prices fall in half due to endless new construction at virtually constant cost?

  44. Gravatar of David David
    7. January 2010 at 21:47

    Scott, I appreciate you coming back here to page 2 and responding. I agree there’s not at present a universal definition of a bubble. Perhaps it’s like Potter Stewart’s definition of obscenity: “I know it when I see it.” I know that’s not rigorous enough for you, but it may have practical use. And like the too-frequent misuse of the term “Ponzi scheme,” maybe the term “bubble” is often misapplied.

    The problem with using things like price or fundamentals to define a bubble is there can always be your skeptical EMH rejoinder, “Someone must have thought it was worth it…” That’s why I think a bubble might be defined as a rapid and easily observable change in people’s motivations. The new paradigm justification, the creation of a new speculator class, The rise to dominance of a formerly sleepy economic sector into our consciousness and media, the abandonment of normal standards of fiscal prudence…

    Finally, I also don’t know anyone who says we’re not in a recession. I’m certainly not saying that. Recessions are obvious ex post because there is a quantitative standard. Perhaps the standard for a bubble is qualitative and one can’t simply point to a P/E ratio and yell bubble. That doesn’t mean bubbles don’t exist or that having an opinion about them isn’t useful.

  45. Gravatar of scott sumner scott sumner
    9. January 2010 at 10:25

    David, This conversation is now carried on in a new thread (the CRE thread). In that thread I mentioned stocks like Microsoft that once looked like a bubble, but proved to be the real thing. So I think high prices can sometimes be justified on expectations, or hype, or whatever you call it. Google might be another example, although perhaps too soon to say.

Leave a Reply