Home Price Declines and Failures of US Financial Institutions, an Example of Macro-Factor Driven Default Risk

Location: Honolulu
Author:
Date: Monday, September 14, 2009
 

Our post on September 9 pointed out that the number of failed and rescued financial institutions in the current crisis has been far less than what we observed in the 1987-1992 period.  We all know that home prices are the number one driver of the 2007-2009 credit crisis, and recent data from the Federal Deposit Insurance Corporation makes that obvious.  That’s the focus of today’s post.

In our September 9 post, we showed a graph by month of the number of financial institution failures and FDIC “assistance” incidents for those months where there was at least one failure.  The graph below shows the results for all months, including those with zero failures, and the cyclicality of bank failures is starkly apparent.

How easy is it to trace the causes of bank failures to specific changes in the macro economy?  Formally, one would do that in the context of a comprehensive logistic regression default model like that offered by Kamakura Risk Information Services.  To prove how easy it is to show management that macro factors matter, however, we resort instead to a single graph. 

In the savings and loan crisis, remember that massive upward movements of interest rates combined with commercial real estate price declines to trigger widespread bank and savings and loan failures. Home prices were down modestly but were at best a secondary factor in those failures. Today, we concentrate on the period from January 1994 to August 2009.  We use the 10 city Case-Shiller home price index compiled by Fiserv and released by Standard & Poor’s Corporation.  We calculated the 2 year change in the Case-Shiller home price index and plotted the number of bank failures in a given month as a function of that 2 year home price change as of that month.  The results are stunning:

Banks do fail at all levels of home price changes.  That being said, as long as the two year home price decline is less than 20%, there is only one month since January 1994 where more than 2 banks have failed.  Once the price decline is greater than 20%, all hell breaks loose as the graph above shows quite starkly.

Two questions immediately come to mind.

Before home prices declined, if we stress tested a home price decline of 20%, would we have predicted that a bank could fail?

Working with many of the largest banks in the United States and national mortgage default data available only before the crisis, we find the answer to this question is yes.  The determining factor for a given bank is both the size and quality of the mortgage portfolio. Quality is measured loan by loan, including estimated current home price values.

Won’t the macro factors that drive the next crisis be different, surprising us and leading to another “black swan” crisis?

There’s a two part answer to this question. First, although the macro factors may be different, the macro factors that cause the next crisis are almost certain to come from this list of factors that have caused almost every other crisis in the last century:

• Home prices
• Commercial real estate prices
• Stock prices (which will reflect factors like economic growth and unemployment)
• Interest rates (and therefore inflation)
• Foreign exchange rates
• Commodity prices

Second, given that this is a very short list, Nassim Taleb will probably be the only person calling the next crisis a “black swan”!

To subscribe or visit go to:  http://www.riskcenter.com