Subprime Spillover
Foreclosures Cost Neighbors $223 Billion; 44.5 Million Homes Lose $5,000 on Average

¹ Ellen Schloemer, Wei Li, Keith Ernst, and Kathleen Keest, Losing Ground: Foreclosures in the Subprime Market and Their Cost to Homeowners, Center for Responsible Lending at 16 (December 2006), available at http://www.responsiblelending.org.

² Dan Immergluck and Geoff Smith, The External Costs of Foreclosure: The Impact of Single-Family Mortgage Foreclosures on Property Values, p. 57, 69, 72, 75Housing Policy Debate (17:1) Fannie Mae Foundation (2006) at http://www.fanniemaefoundation.org/programs/hpd/pdf/hpd_1701_immergluck.pdf.

³ Metropolitan statistical areasare geographic entities defined by the U.S. Office of Management and Budget for use by Federal statistical agencies in collecting, tabulating, and publishing Federal statistics. For more details, see http://www.census.gov/population/www/estimates/metroarea.html

⁴ In CRL’s “Losing Ground” paper, for loans originated during 2005 and 2006, we projected 1.1 million foreclosures in the US. In the current study, our foreclosure estimates limited to first lien, owner-occupied subprime loans reported in 2005 and 2006 HMDA as “higher-cost” loans, secured by a property located in a MSA. The resulting number of foreclosures in this slightly smaller geography is estimated at 923,034.

⁵ Specifically, Block Numbering Areas (BNAs), which are geographic entities similar to census tracts and delineated in counties (or the statistical equivalents of counties) without census tracts.

⁶ For the first time in 2004, lenders were required to report the spread between the APR of designated loans and the yield on a U.S. Treasury security of comparable maturity. Specifically, lenders submitted this information on first lien loans if the spread was at or above three percentage points. In this study, we refer to loans with APRs high enough to require the disclosure of this spread as subprime loans.

⁷ According to Census Bureau, "Census tracts are small, relatively permanent geographic entities within counties... Generally, census tracts have between 2,500 and 8,000 residents and boundaries that follow visible features." Since, "census tracts are to be as homogeneous as possible with respect to population characteristics, economic status, and living conditions", they are the observation units of our study.

⁸ All figures in this analysis cover only originated conventional home loans to owner-occupants, in a Metropolitan Statistical Area or Metropolitan Division, secured by a first-lien on a 1- to 4-unit home, as disclosed under HMDA.

⁹ Subprime foreclosure rates vary across geographical locations. In addition to the difference on risky loan terms, foreclosures are more likely in housing markets with lower house price growth. This geographical variation on foreclosure rates is incorporated into our study at MSA level. See Appendix 5 of Losing Ground (note 1), p. 48. CRL projected lifetime foreclosure rates for 378 MSAs. For the remaining 9 MSAs in the HMDA data, we assume the same foreclosure rate as the US average foreclosure rate (19.4%).

¹⁰ See Schloemer et al, note 1.

¹¹ Christopher L. Cagan, Mortgage Payment Reset: The Issue and the Impact, First American CoreLogic (March 19, 2007) and quoted in Subprime Borrowers to Lose Homes at Record Pace as Rates Rise, Bloomberg.com (September 19, 2007), available at http://www.facorelogic.com/uploadedFiles/Newsroom/Studies_and_Briefs/Studies/20070048MortgagePaymentResetStudy_FINAL.pdf and http://www.bloomberg.com/apps/news?pid=email_en&refer=finance&sid=akOEPec30TR4, respectively.

¹² Mortgage Finance Industry Overview, Lehman Brothers Equity Research (December 22, 2006).

¹³ Mark Zandi, Statement before the House Judiciary Committee United States House of Representatives (October 30, 2007) available at http://judiciary.house.gov/OversightTestimony.aspx?ID=1188

¹⁴ Market Tabs, Credit Suisse Fixed Income Research (October 26, 2007).

¹⁵ Data collected from the Summary File 3 database of 2000 Census.

¹⁶ See Immergluck et al, note 2.

¹⁷ For a census tract, let A be the area size in square miles, B be the number of foreclosed subprime loans, C be the number of housing units, and D be the median house price. Let G=64A/ƒÎ. Then the number of neighboring homes experiencing devaluation is given by

H= C if B is greater than or equal to G

CxB/G, if B is less than G (1)

The dollar amount of decrease in house value/tax base from foreclosure effect is given by

I = 0.009 xCxDxB/G (2).

¹⁸ In 2005 HMDA data, there are total of 387 MSAs. For these MSAs, there are total of 1,158 counties and 53,293 census tracts. In 2006 HMDA data, there are total of 387 MSAs. For these MSAs, there are total of 1,158 counties and 53,245 census tracts. Combining both years yields 56,777 census tracts.