Do information asymmetries explain the housing bubble?

By Felix Salmon
September 2, 2010
new paper out which reckons it can explain the entire housing bubble by looking at the supply of private-label mortgage-backed securities in the market, and the information asymmetries embedded in them.


" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Adam Levitin and Susan Wachter have a new paper out which reckons it can explain the entire housing bubble by looking at the supply of private-label mortgage-backed securities in the market, and the information asymmetries embedded in them.

They do have a point: since the banks putting together these private-lable securities, or PLS, knew much better than the buyers (and, for that matter, the ratings agencies) what was going into them, there was an opportunity — grasped with both fists — to take advantage of those asymmetries:

PLS are unusually complex, heterogeneous products. Any particular securitization is supported by a unique pool of collateral and has its own set of credit enhancements and payment structure. Complexity and heterogeneity shrouded the risks inherent in PLS. As a result, investors failed to properly price for risk, as they did not perceive the full extent of the risk involved. The structure of PLS (including the underlying mortgages) allowed investors to underestimate the risks involved and therefore underprice the PLS by demanding insufficiently large yield spreads. The housing bubble was fueled by mispriced mortgage finance, and the mispricing occurred because of information failures. Thus, at the core of the housing bubble was an information failure. Investors lacked adequate information about the risks involved with PLS.

I like the way that Levitin and Wachter look at the commercial real estate market as well as the residential market. And I also like the way that they say that the bubble was actually pretty short-lived, starting in 2004 (or possibly 2003) and bursting in 2006. But if that’s true, then the housing bubble started right when the homeownership rate peaked, and the homeownership rate actually declined for the entire duration of the bubble. Which seems a bit weird.

What’s more, while the paper nods to the odd fact that at one point structured mortgage-backed securities started trading through triple-A-rated corporate bonds, it doesn’t really explain why that happened. And it’s far too quick to dismiss other theories about what caused the housing bubble by simply saying that those theories don’t explain the low spreads on mortgage-backed bonds. Which might be right, but surely a bubble can have more than one cause.

And while there are occasional references to housing bubbles elsewhere in the world, there’s no attempt to see whether they can be explained in similar terms. I suspect that doing so would be hard, in which case it’s not that silly to assume that the causes of the US housing bubble overlapped with the causes of the housing bubbles in countries like the UK, Ireland, Spain, Australia, and even South Africa.

In fact, you don’t even need to look abroad. Private-label mortgages were subprime and alt-A, yet we had a perfectly healthy bubble right here in Manhattan, where there was no subprime mortgage bubble at all.

Most importantly, the people with the upper hand, when it came to the information asymmetries, ended up being the people taking the biggest losses when the bubble burst. Information asymmetry might explain why small Norwegian municipalities ended up losing millions investing in structured notes, but they don’t explain the losses at Merrill Lynch and Citigroup, or, for that matter, at Fannie and Freddie.

So while this paper constitutes an important contribution to the literature, I don’t think the authors have managed to explain the housing bubble with a single concept. In fact, I very much doubt that such a thing would even be possible. Modern economics and finance is far too complicated for that.

8 comments

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/

Right Felix, I agree that monocausal explanations here are not the way to go, but after having read the first half of M. Lewis’ The Big Short, it does smell as if there was a huge information asymmetry on Wall Street itself in terms of what the respective players were really getting into.

Claus

Posted by clausvistesen | Report as abusive

There was no housing bubble, at least not nationally. A handful of areas (FL, AZ, Las Vegas, and parts of CA) had bubbles where way too much housing was built because speculators had easy access to cheap debt, and thought prices would only go up. In the rest of the nation, purchase prices went up only because interest rates went down and rules were eliminated. And while purchase prices went up, the cost of owning didn’t, because interest rates were lower. If rates were lowered today, and the rules reset to 2003 (where anybody who was breathing could get a mortgage), prices would rise again, because the monthly payments would decrease.

If you want to talk about bubbles, talk about the debt bubble. That was real, and everywhere.

Posted by OnTheTimes | Report as abusive

“I suspect that doing so would be hard, in which case it’s not that silly to assume that the causes of the US housing bubble overlapped with the causes of the housing bubbles in countries like the UK, Ireland, Spain, Australia, and even South Africa.”

The three largest mortgage securitisation markets in the world? The US, the UK and Australia. Also the only three countries with subprime mortgage securitisation markets of any significant size. Spain was also in the top 5 private label mortgage securitisation markets globally, and while Ireland’s market was small, so is its economy. South Africa’s RMBS market started when the boom started. Now I’d be the first to say securitisation wasn’t the only factor in the bubbles (there’s a reason why so many of the most affected countries are former British colonies and hence common law jurisdictions), but it was certainly a strong contributor.

Subprime is something of a red herring, except that it exposed the fundamental weaknesses of the securitisation model in such a stark way and ensured that when the bubble burst it would be particularly bad. The fundamental issue was massive, often hidden leverage with risks recirculated within the system rather than dispersed, abetted by securitisation in many forms.

Posted by GingerYellow | Report as abusive

Right on point.

While interesting, Levitan and Wachter maybe showing some unfortunate “focusing effect.”

The U.S. housing bubble dissipated too much net worth, spread way too far, and involved such an ensemble cast that it can’t be reduced to a game of Clue in which “the butler did it in the Conservatory with the candlestick.” The housing bubble had more similarities to say, “Murder on the Orient Express”–everyone did it. Well, maybe everyone’s cognitive biases did it.

Of course, Levitan and Wachter are right. There are information asymmetries at each level of the mortgage production process and in PLS creation and sales: Homeowners know more about their particular house than brokers do. Brokers know more about the borrower and his house than bankers do. Bankers know more about it than CDO packagers, packagers know more than PLS buyers, etc.

But the fascinating (and horrifying) aspect of the U.S. housing bubble is that it destroyed a significant amount of the U.S. private sector’s main nest egg–home equity–while preying on a bevy of known cognitive biases, e.g. interloper effects (rating agencies), status quo bias (regulatory agencies, CDO buyers), availability cascade (home buyers), and Dunning-Kruger effects (regulators, rating agencies, and almost everybody now with negative equity), just to name a few.

What is more interesting is how we avoid a similar debacle the next time. Levitan and Wachter’s prescription would be to fix the information asymmetries of PLS, which may be a necessary but is not a sufficient condition for a fix.

Regions in the U.S. that fared better had legal, regulatory, or transaction restraints on home mortgages (like state consumer laws or NYC co-op boards). Regions that had stricter rules on second homes and investment homes tended to fare better. States whose economies had a proportionately smaller reliance on real estate and construction fared better. All of this reduced irrational decision-making in those places.

And the differential effect is clearly evident in the FDIC’s most recent quarterly report (June 2010). Home loan delinquencies (as a % of total home loans) in the FDIC Atlanta region reached 12.19% versus 4.71% in the New York region.

On the other hand, the regulatory authorities, in particular the Fed, let everyone’s cognitive bias run wild. Or better said, they refused to believe that such irrationality could exist. Hence, low interest rates created lower carrying costs which certainly goosed the demand for homes. At the same time, the Fed decided not to enforce HOEPA (1994) which accelerated the mortgage origination “corruption multiplier” (i.e. “anyone breathing could get a mortgage”) in non-bank mortgage originators. Non-bank mortgage practices later spread to banks and to the GSEs in their attempt to compete with new non-bank, corrupt product.

We can only hope that the Fed’s new, broader mantle of oversight is assumed with all of these causal factors in mind.

Posted by AABender1 | Report as abusive

It seems to me that a lot of bubble behavior is psychology-driven, so let me offer this hypothesis. Buying a home is something that the vast majority of people do perhaps three or four times in their lives. They rely on outside source of information as a guide, and realtors and agents are highly biased sources of that information. So they hear that they have to buy sooner or later, before prices increase, while the house or neighborhood are still available, and so one.

Throughout the bursting of the bubble, we are still hearing the real estate people saying that it’s the perfect time to buy. The combination of expertise and bias, coupled by any information to the contrary, is what drives bubbles such as this.

Posted by Curmudgeon | Report as abusive

What AABender1 said.

To which I might add that this question of information asymmetry is hardly unexplored territory. There is a good survey of the literature here: http://www.bankofcanada.ca/en/review/aut umn09/paligorova.pdf. There is a very accessible blog post on the subject here: http://newyorksocietyofsecurityanalysts. typepad.com/nyssa-news/2010/04/the-seven -deadly-frictions-of-subprime-mortgage-c redit-securitization.html.

If you are going to focus on just one issue, I think that agency conflicts are more significant than information. Many agents have incentive to disregard their lack of information. This was recognized quite early on, e.g.: http://siteresources.worldbank.org/INTFR  /Resources/KaneCaprioDemirgucKunt-The20 07Meltdown.pdf.

Finally, first-loss was widely presumed to be the remedy to agency conflict both before and after the crisis. But Hartman-Glaser et al have an interesting argument that the timing of payments to the servicer is more important than the first loss position: https://wpweb2.tepper.cmu.edu/wfa/wfasec ure/upload/2010_3.42582E+08_PA_OSMH.pdf. Others have noted that first-loss is unlikely to be effective when the first-profit position is actually larger (sorry, can’t recall the authors or paper to mind.)

Posted by Greycap | Report as abusive

Good stuff Mr. bender. Was that mother Goose?

Posted by hsvkitty | Report as abusive

Surprise, surprise, Information Asymmetry strikes again. But how can we avoid this from adding to systemic risk – indeed, aren’t markets through their diversified channels and means of ongoing price discovery assist in overcoming this disconnections?

Maybe it was the reality that these PLS were stand alone securities that were not traceable and accountable throughout the market both at the end-point in the securitisation process but also in their initial supply.

A solution might be to ensure that some level of market coherence is set in – ie an open marketplace that will identify buyers and and issuers of the securities ensuring that reputation and branding becomes associated with the price discovery process. This should avoid the incentive to purposely ‘game’ the system with an ongoing engagement of the market ensuring a positive incentive to participate openly and efficiently.

I am not saying, for as much, that these securities need be exchange traded automatically – certainly we have many forms of financial arrangements where reputation is engaged without accessing through the strictures of a strict central clearing exchange, eg. credit facilities and credit records – but to skip the whole accountability process entirely is evidently too risky.

Hopefully self regulation will prime in this above externally enforced rules, certainly from a buyer/securitiser’s point of view, the benefits of adequate traceability and fungibility of these securities is paramount for their operations. Another side benefit of self-regulation would be to ensure that compliance costs are properly internalised to the primary beneficiaries rather than redistributed on to other market by-standers who might not be trying to gain exposure, financial or regulatory wise, to the housing market.

TS
http://twentyfoursomething.wordpress.com  /

Posted by tariqscherer | Report as abusive