| Measuring Risk in Residential Mortgages |
| Written by Jonathan Smoke | |
| 08.20.2007 | |
|
Discuss this article on the forums. (0 posts) As the housing downturn has worsened since March with the unwinding of the subprime market, the market has been fixated on the debt side of housing. Now with increasing concerns that the subprime meltdown will impact prime credit and even other asset classes, it has been a little surprising to see how nervous Mr. Market is about the level of risk in securitized mortgage pools and portfolios. The number of mortgage company casualties is definitely a sign that the market is nervous. According to MortgageDaily.com’s Mortgage Graveyard, there have been 84 failed companies in 2007 and 30 acquisitions with another 19 companies struggling. Granted, while this is a very fragmented market that grew rapidly and thus created part of its own future demise, the size of the failure has been astounding to me. Furthermore, since this market’s failure seems to be unnerving broad markets worldwide, it makes me wonder about the basis of the intelligence used to underwrite and estimate the risk of mortgage loans.Much has already been written about the potential failure of the rating companies to properly account for the risk. For one of the most comprehensive pieces written to date, see the Wall Street Journal’s “How Rating Firms’ Calls Fueled Subprime Mess.” As someone who believes the market eventually gets it right despite a few “Minsky Moments,” the level of this surprise and the potential risk it puts on the economy overall has been astounding. But I think I finally understand what’s so wrong about where the market is relative to where it ought to be. The market has been fed a lot of information that said the risk was X, but it has turned out that the risk may be 2X or 3X or maybe X x X or anything significantly larger than X. Because the market hates nothing more than being wrong and unsure at the same time, anything or anyone connected with these mortgage assets is being punished. What’s wrong with the current information about mortgage portfolios? It’s generally missing the information on the factors that are currently impacting the potential default risk: home price instability, local market economics, property market conditions, demographics and consumer credit. While prices always advanced these factors didn’t matter all that much, so the loan level data used to fuel most credit risk models focused only on data about the mortgages themselves. As we all know now, prices do fall and in so doing have created a bit of a perfect storm that in turn impacts several other factors. My background has led me to believe that to understand any real estate investment, you must understand the local market, demand and supply. Otherwise, you may as well just guess. An investor, developer or builder will determine much of their success in choosing where to invest and making the most out of what they do with that investment. So I would expect that to be successful in maximizing the performance of mortgage portfolios, one should consider all of the factors that would increase the likelihood of success or failure and make informed decisions accordingly. We are actively working on a way to provide intelligence that could be applied to a mortgage portfolio in order to properly account for key factors using the same methods and data we use to judge land investments. We will share what we discover with our readers, but we’d also like to hear from you about any ideas you have on the best way to analyze pools of mortgages and related credit risk. If you have opinions on this analysis or would like to share suggested metrics, please post a comment to this article or post a comment to the related topic in our forums, “What are the best metrics to judge mortgages?” |
| < Prev | Next > |
|---|

Granted, while this is a very fragmented market that grew rapidly and thus created part of its own future demise, the size of the failure has been astounding to me. Furthermore, since this market’s failure seems to be unnerving broad markets worldwide, it makes me wonder about the basis of the intelligence used to underwrite and estimate the risk of mortgage loans.


