Estimated Holdings of AAA CDO Tranches
VII. A House Prices Do not Rise
House prices were supposed to always go up. Between 2001 and 2005 homeowners enjoyed an average increase of 54.4 percent in the value of their houses, as measured by the Office of Federal Housing
53
When I say ―value‖ I usually mean to compute an expected loss or expected payoff using historical information. ―Marking-to-market‖ is another matter, briefly discussed later.
54
Enterprise Oversight (OFHEO).55 In terms of the two-year fixed rate part of a 2/28 subprime mortgage, from January 1997 to July 2007 every rolling two year period showed positive house price appreciation, according to the S&P/Case-Shiller (U.S. National) Index. In fact, from March 1998 to March 2007, every rolling two year period displayed double digit house price appreciation. There was no appreciation or depreciation in August 2007 and starting in September 2007 house price appreciation has been negative. The figure below shows a plot of the lagging two year house price appreciation.
But, then house prices declined. In fact, the S&P/Case-Shiller (U.S. National) quarterly home price index declined by 4.5 percent in Q3 2007 versus Q3 2006 – the largest drop since the index started recording data in 1988.56 Home prices, as measured in the 20 U.S. metropolitan areas, declined by 4.9 percent, the largest drop since the index was started in 2001, with 15 of the 20 cities showing year-on-year declines in prices. The two largest declines occurred in Tampa (-11.12% Y-o-Y) and Miami (-9.96% Y-o-Y). U.S. home prices declined 6.7 percent in October from a year earlier, a record drop for the ten city S&P/Case- Shiller index.57
Lagging Two-Year House Price Appreciation (%)
Source: S&P.
55
The calculation is the percentage change in the seasonally adjusted OFHEO repeat-sales house price index for purchase transactions only between the fourth quarters of 2000 and 2005. See www.ofheo.gov/HPLasp .
56
There are two indices that measure house price appreciation, S&P/Case-Shiller and the OFHEO House Price Index. Both of these indices are based on repeat sales. The two indices differ in important respects. Case-Shiller does not cover the entire U.S., and the omitted areas seem to be doing better than the included areas. Case-Shiller omits 13 states altogether and has incomplete coverage of 29 other states (see Leventis (2007)). The OFHEO index is not value-weighted and only includes homes with conforming mortgages.
57
The United States has not experienced large, nationwide decline in house prices since the Great Depression of the 1930s. In 1940 the median nonfarm housing value was 48.6 percent below the 1930 median value (based on the 1940 Housing Census). Over the same decade, the Consumer Price Index had fallen 17.4 percent and food prices had fallen 27 percent. In other words, even adjusting for the deflation during the period, housing prices had not recovered to the levels at the beginning of the Depression by 1940. See Fishback, Horrace and Kantor (2001).
The ability of subprime and Alt-A borrowers to sustain their mortgage payments depends heavily on house price appreciation because of the need for refinancing. When house prices did not appreciated to the same extent as in the past, and in many areas they have recently gone down, the ability of borrowers to refinance has been reduced. In fact, now because of the crisis, underwriting standards have become much tougher, and many lenders are in bankruptcy, meaning that the mortgage market for these borrowers to refinance has effectively closed.
Currently, almost all the major issuers of subprime mortgages are either out of business or have stopped making subprime loans unless they conform to GSE underwriting criteria. Problems in the Alt-A market are still mostly in the future, and it is likely that this market will also shut down. The unwillingness to originate subprime mortgages is significantly driven by the impossibility of a securitization take-out of the loans. This shutdown means that borrowers in the subprime and Alt-A mortgages will have a very difficult time refinancing when their hybrid ARMs are reset.
The shutdown of the subprime mortgage market is very important because of the number of borrowers who will soon reach their reset date, that is, the date at which the initial fixed teaser rate ends and the mortgage rate resets to a significantly higher floating rate. Evidence of the shutdown in the refinancing market comes from remittance data. Remittance data shows that the shutdown is dramatically reducing subprime prepayment speeds.58 A decline in prepayment speed means that borrowers cannot refinance either because they no longer can find a lender or because they have no equity built up on their houses. Delinquencies and foreclosures are the result.
58 The trustees for transactions make monthly reports known as remittance reports. Remittance reports details scheduled and unscheduled remittances of principal, servicer advances, loan repurchases, realized losses, delinquencies, and so on.
Delinquency Rates (%)
Source: Mortgage Bankers Association.
We now turn to the issue of how the information about house prices and delinquencies and foreclosures was linked to valuations of the various parts of the chain. Keep in mind that house price and mortgage performance information arrives with a lag, not in real time.