Private Information or (and?) Local Thinking during the Housing Bubble

A few weeks ago I came across a research paper named “Wall Street and the Housing Bubble” by Cheng, Raina and Xiong (2012). I found it while I was reading the always-interesting (and diverse) blog “Nada es Gratis”.  In here, you can find (in spanish) a thorough synopsis of what the article is about. The paper analyzes whether the experts in securitization were aware of the housing bubbles and the possibility of the collapse of housing prices in the US. Namely, they test the “Inside Job” view, popularized by the film of the same name which held the view (among other things) that insiders were aware of the financial bubble during 2004-2006 and its inner workings. At first glance their results are at least thought-provoking.

They construct a database with three different groups of professionals. A “treatment” group composed of the experts in securitization in the housing market. And two “control” groups that consist of financial equity (non-housing) analysts and lawyers. The study is an experimental one in nature as the authors take care of these groups being very similar in terms of individual characteristics of each group. Their hypothesis, as stated in the paper, is the following:

Hypothesis 1 (Inside Job View): Securitization agents exhibited more awareness of the housing bubble relative to lawyers and non-housing equity analysts in two possible forms:

A. (market timing form) Securitization agents who were homeowners were more likely to divest homes and down-size homes in 2004-2006.

B. (cautious form) Securitization agents who were non-homeowners were less likely to acquire homes in 2004-2006.

Overall, securitization agents had better performance after controlling for their initial holdings of homes at the beginning of 2004. 

Their conclusion is that their data rejects their hypothesis in its two forms, and there appears to be no evidence that securitization agents (in average) were more aware of the housing bubble than other professionals.

On the one hand, I think this type of papers (and data) provide us of a much deeper understanding of both the financial crisis in 2008/2009 and the inner workings of the financial system. It could potentially open up new avenues of both empirical and theoretical research because the result challenges the median view on the “Inside Job” subject during the crisis. On the other hand, I believe that the authors make a somewhat rushed claim. Because the result is quite appealing, they rapidly suggest that their result “points towards the need to understand how beliefs were formed among people in finance during the bubble”. Thus, for Cheng, Raina and Xiong (2012), their paper gives support to Gennaioli, Shleifer and Vishny (2012) and Coval, Jurek and Stafford (2009), which study the effects of local thinking[1] (the former) and poorly estimated parameters in structured products (the latter). They, thus, deemphasize the importance of private information in the finance industry during the crisis. However, I am not completely convinced that private information played an important role during this crisis. If one reads carefully the conclusion of the paper, the authors clearly state that their hypothesis holds mainly for the average securitization agent. More than that, in “Nada es Gratis” the author reviewing the paper goes on to assert that this is good news for the ones that think of the whole financial system as a “chicken running without a head not knowing where it goes”. Still, I was thinking that in books like Too Big To Fail or movies such as Margin Call, while most of the people were not aware about what structured products were, there were always very few people (within the financial system) that did and saw the problems coming. Then, a natural question would be, would that kind of private information shown up in the average securitization agent behavior? My feeling is that it would not. Furthermore, I am inclined to think that the greater is the amount of the private information, the worst are the consequences for the financial system in a crisis such as the 2008/2009 one[2] and the less it would show up in the average effect.

To sum up, I still consider that both avenues of research (local thinking and private information) should be pursued in trying to explain the global financial crisis and I think empirical studies such as the article by Cheng, Raina and Xiong (2012) should focus more on outliers within the securitization group and how they acted during the boom and bust of the housing bubble. The evidence coming from such studies could shed light about the size (if any) of private information during these bubbles.


Broner, F. 2008. “Discrete Devaluations and Multiple Equilibria in a First Generation Model of Currency Crisis”, Journal of Monetary Economics, Vol. 55 (3), pp. 592-605.

Cheng, I., Raina, S., and Xiong, W. 2012. “Wall Street and the Housing Bubble”, mimeo.

Coval, J., Jurek, J., and Stafford, E. 2009. “The Economics of Structured Finance”, The Journal of Economic Perspectives, Vol. 23 (1), pp. 3-26.

Gennaioli, N., Shleifer, A., and Vishny, R. “Neglected Risks”, Journal of Finance, forthcoming.

[1] Local thinking refers to the behavior of neglecting tail risks. In their model of shadow banking in Section III, the agents do not have rational expectations, but instead a form of local thinking: they only think about the two most likely states, but not in the more unexpected one.

[2] This thought is mainly due to Broner (2008), where a huge amount of private information during currency crisis leads to large swings in the price of assets.