By 2001, as the demand for safe assets began to rise above what the U.S. corporate world and safe mortgage‐borrowers naturally could provide, financial institutions began to search for mechanisms to generate triple-A assets from previously untapped and riskier sources.
…even if correctly rated as triple-A, the correlation between these complex assets distress and systemic distress is much higher than for simpler single‐name bonds of equivalent rating.
His story is that there was an “excess demand” for safe assets, due to increased savings in emerging-market countries. He is correct that the financial sector satisfied this demand with assets that were in fact not safe.
His ideas bear some resemblance to my view (see lecture 9) that the nonfinancial sector wants to issue risky long-term liabilities and to hold safe short-term assets. But Caballero takes it as given that either the financial sector or government has to find a way to satisfy “excess demand” for safe assets. Instead, I think that the market needs to find a natural equilibrium, in which savers have to accept either low return or some risk and borrowers have to offer a high return or undertake low-risk projects (such as making a large down payment on a house).
Also, note that the excess demand for financial intermediation may have come from the borrowing side as well as the lending side. See real time economics.
Finally, Mark Thoma points to a post by Economics of Contempt that lists various economists and pundits who denied the existence of a housing bubble. I myself could have made that list if he had used my June 2004 article which said,
I would argue (and other economists would agree) that the bubble is in the bond market, not in the housing market. That is, if interest rates remain close to where they are today, then there is no reason for house prices to collapse.
However, in my defense, a fair amount of the run-up in house prices took place after I wrote the article. If prices had stayed where they were in June of 2004, we would not be talking about a housing bubble today. Indeed, the accusatory blog post contains too many quotes dated from the first half of 2005 or earlier (one quote even goes back to 2003). Given the data that was available on house prices as of early 2005, I think it is a stretch to call people fools for not calling a housing bubble. The really dramatic run-up in house prices came in the final two years of the bubble.
To put it another way, if you had asked these folks at the time they made their offending statements, “suppose house prices rise another 30 percent in the next two years. Would that cause you to worry?” My guess is that at least some of us would have said, “yes.”
READER COMMENTS
Mike Welker
Jan 12 2010 at 11:09am
Interesting to note, too, that John Taylor runs an estimate of the housing price “bubble” under the use of the Taylor Rule. His point is that no bubble would have come about, of course. Under the TR the Fed would have tightened credit at a pretty constant rate. Just an interesting counter factual.
E. Barandiaran
Jan 12 2010 at 11:51am
Arnold, I’d like to expand on the comments I made on Tyler’s post of Ricardo Caballero’s paper. Ricardo has put together several ideas on which he has been working for a long time.
Three comments. First, I agree with Ricardo that the original imbalance turned out to be a bluff, but I believe it was so because there was no global imbalance–the increase in the flow of global savings could have been easily intermediated by the financial system (it was not that easy in the 1970s when the system had to recycle the savings of oil-producing countries) and the fears that the increase could be reversed were largely based on political arguments. We know that there has been an increase in the flow of global savings but we don’t have a good measure of it and more importantly we don’t know its composition in terms of an increased demand for financial and nonfinancial assets. Anyway, it’s my impression that we are talking about increases that were smaller in relative terms than the increase in the flow of global savings and its related increase in the demand for financial assets during the 1970s. More important, in the 1970s there were too many restrictions on the financial system that made the recycling of savings a huge challenge.
Second, although the increase in the flow of savings has implied an increase in the demand for “safe” and “risky” assets, in the past 30 years the supply of new types of financial assets has been largely a response to the incentives from changes in information systems and regulatory regimes. Indeed it may be difficult to identify the causes of several innovations over a long period of time, but my point is that they started in the 1980s, well before a large increase in the flow of savings. In addition, we should not minimize the extent to which in the past 10 years the global financial system has been able to accommodate an increasing demand just by increasing the supply of old assets.
Third, regardless of how “safe” assets can be produced in normal times (a difficult question indeed because in these times most assets may seem safe), at a times of crisis in which “everyone” is expecting some sort of government intervention, only the government can supply “safe” assets, a point that was critical to solving the Chilean financial crisis of 1982/83 (that is, when Ricardo was a student at Universidad Católica de Chile). Government interventions are effective only if the government can issue “safe” assets, that is, if the fiscal position is sound enough to persuade people and financial institutions that the newly issued bonds will be serviced on time. A comparison between what happened in the second half of 1982 in Argentina and in the first half of 1983 in Chile illustrates this point.
Bill Shoe
Jan 12 2010 at 12:27pm
Arnold said later (2005?, 2006?) that based on house prices in his area of Silver Springs, Maryland he decided there was a housing bubble. I couldn’t find this particular quote, it may have been in an offsite essay or something. However…
I did find this more recent quote where he underestimates the maximum possible house price decline–
“I’m sorry, but unless by “some areas” you mean areas the size of a 9-digit zip code, we’re not going to see 40 percent declines in house prices.”
— Time to Panic? (Aug 2007)
However, Arnold has continuous blog postings going back to 2003 and they’re all accessible. Few econ bloggers have that kind of track record to conveniently pick through. And I’m essentially anonymous, so no track record for me.
wm13
Jan 12 2010 at 1:57pm
The theory that there was excess demand for “safe” assets seems problematic. During the period from 2002 to 2007, credit spreads were tightening, not widening, so there was, if anything, growing demand for risky assets.
Lord
Jan 12 2010 at 5:56pm
Well if people can only tell there is a bubble after the decline has started then the beginning of 2006 would have been the peak, but by all measures, http://www.calculatedriskblog.com/2009/02/house-prices-real-prices-price-to-rent.html
prices had exceeded historical levels by 2001 and while interest rates would justify higher levels, by mid 2003 they were already 30% higher. Certainly by 2005 anyone that doubted there was a bubble was delusional and should be remembered as such. The key was there was no reason to expect accelerating values as the Fed had already tightened considerably and was continuing to do so. Fools they were.
SydB
Jan 12 2010 at 8:28pm
On 3 April 2006 Mr Kling said “1. Relative to historical norms, prices are very high…2. Relative to rents, prices are reasonable” and that “For housing, I would rate the probability of a bubble at about 20 percent.”
I’d say this was at about the top of the bubble.
Not pointing fingers or anything like that. By 2003 I thought the whole thing was nuts.
Mark
Jan 13 2010 at 10:24am
Arnold,
IMO, your June 2004 bond-bubble argument is still the more important argument to make rather than its effect on housing.
Your argument ends up giving the rationale for the housing crash as well. When the risk premiums on MBSs widen and the entire shadow system begins its collapse, your one condition (assumption of interest rates not changing) goes away…
We might not be able to conceive it naturally, because history didn’t unfold that way, but this bond bubble very well could have shown up in the exaggerated appreciation of some asset other than homes…
Bond Bubble = Cause
Housing Bubble = Effect
Comments are closed.