Resolving The Foreclosure Crisis-Modification of Mortgages in Bankruptcy

LEVITIN - FINAL
4/23/2009 11:40 AM
RESOLVING THE FORECLOSURE CRISIS: MODIFICATION OF MORTGAGES IN BANKRUPTCY

ADAM J. LEVITIN
This Article empirically tests the economic assumption underlying the policy against bankruptcy modification of home-mortgage debtthat protecting lenders from losses in bankruptcy encourages them to lend more and at lower rates, and thus encourages homeownership. The data show that the assumption is mistaken; permitting modification would have little or no impact on mortgage credit cost or availability. Because lenders face smaller losses from bankruptcy modification than from foreclosure, the market is unlikely to price against bankruptcy modification. In light of market neutrality, the Article argues that permitting modification of home mortgages in bankruptcy presents the best solution to the foreclosure crisis. Unlike any other proposed response, bankruptcy modification offers immediate relief, solves the market problems created by securitization, addresses both problems of payment-reset shock and negative equity, screens out speculators, spreads burdens between borrowers and lenders, and avoids the costs and moral hazard of a government bailout. As the foreclosure crisis deepens, bankruptcy modification presents the best and least invasive method of stabilizing the housing market.
Introduction: Foreclosure, Bankruptcy, and Mortgages ............... 566 I. The Structure of the Mortgage Market ........................... 579 A. Treatment of Mortgages in Bankruptcy ..................... 579 B. Structure of the Modern Mortgage Market ................. 582 II. Bankruptcy-Modification Risk as Reflected in MortgageMarket Pricing ....................................................... 586 A. Mortgage-Interest-Rate Variation by Property Type ...... 586 1. Experiment Design ........................................ 586 2. Experiment Results ........................................ 589 B. Private-Mortgage-Insurance Rate Premiums ............... 593 C. Secondary-Market-Pricing Variation by Property Type .. 597
Associate Professor, Georgetown University Law Center. This Article was supported by grants from the Reynolds Family Fund at the Georgetown University Law Center. This Article has benefited from comments and suggestions from Amy Crews-Cutts, William Bratton, Michael Diamond, Anna Gelpern, Joshua Goodman, Rich Hynes, Gregory Klass, Richard Levin, Sarah Levitin, Ronald Mann, Katherine Porter, Mark Scarberry, Eric Stein, Dom Sutera, Fred Tung, Tara Twomey, William Vukowich, Susan Wachter, and Elizabeth Warren. The Article has also benefited from presentations at the Harvard/University of Texas Conference on Commercial Realities, the Georgetown University Law Center Faculty Workshop, the Research and Statistics Seminar of the Board of Governors of the Federal Reserve System, the University of Virginia School of Law Faculty Workshop, and the American Law and Economics Associations Annual Convention. Special thanks to Robert P. Enayati, Tai C. Nguyen, and Galina Petrova for research assistance. Please direct comments to alevitin@law.georgetown.edu.
Electronic copy available at: http://ssrn.com/abstract=1071931
LEVITIN - FINAL
4/23/2009 11:40 AM
566
WISCONSIN LAW REVIEW
D. Historical Impact of Permitting Strip-Down ............... 598 III. Explaining Mortgage-Market Indifference to Bankruptcy Strip-Down ........................................................... 600 A. The Baseline for Loss Comparison: Foreclosure Sales ... 603 B. Projecting Lender Losses in Bankruptcy ................... 606 1. 2001 Consumer Bankruptcy Project Database ......... 607 2. 2007 Riverside-San Bernardino Database .............. 611 IV. Policy Implications .................................................. 618 A. Voluntary Versus Involuntary Modification of Mortgages ....................................................... 618 B. Bankruptcy Modification Compared with Other Policy Responses ....................................................... 626 1. Laissez-Faire Market Self-Correction .................. 627 2. Coordinated Voluntary-Workout Efforts ............... 627 3. Procedural Requirements to Encourage Consensual Workouts .................................................... 627 4. Foreclosure and Rate-Increase Moratoria and Other Limitations on the Foreclosure Process ................ 628 5. Government Modification Following EminentDomain Takings............................................ 631 6. Government Refinancing, Guaranty, or Insurance of Mortgages ................................................... 631 a. FHASecure and HOPE for Homeowners Act ...... 634 b. Homeowner Affordability and Stability Program .. 636 7. The Advantages of Bankruptcy Modification .......... 640 Conclusion .................................................................... 647 Postscript...................................................................... 649 Appendix A: Mortgage-Origination Rate Quotes (Selected) .......... 651 Appendix B: Evaluating the Mortgage Bankers Associations Modification-Impact Claim......................................... 654 INTRODUCTION: FORECLOSURE, BANKRUPTCY, AND MORTGAGES The United States is in the midst of an unprecedented homeforeclosure crisis. At no time since the Great Depression have so many Americans lost their homes, and many millions more are in jeopardy of foreclosure. Nearly 1.7 million homes entered foreclosure in 2007,1
1. HOPE NOW, MORTGAGE LOSS MITIGATION STATISTICS: INDUSTRY EXTRAPOLATIONS (QUARTERLY FOR 2007 AND 2008) (2009), available at http://www.hopenow.com/upload/data/files/HOPE%20NOW%20Loss%20Mitigation% 20National%20Data%20July07%20to%20February09.pdf; see also Press Release, RealtyTrac, Inc., U.S. Foreclosure Activity Increases 75 Percent in 2007 (Jan. 29,
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
567
and another 2.2 million entered in the first three quarters of 2008.2 Over half a million homes were actually sold in foreclosure or otherwise surrendered to lenders in 2007,3 and over 900,000 were sold in foreclosure in 2008.4 At the end of 2008, more than one in ten homeowners were either past due or in foreclosure, the highest levels on record.5 By 2012, Credit Suisse predicts around 8.1 million homes, or 16 percent of all residential borrowers, could go through foreclosure.6 Expressed differently, one in every nine homeowners and one in six households that have a mortgagewill lose their home to foreclosure.
2008), available at http://www.realtytrac.com/ ContentManagement/pressrelease.aspx? ChannelID=9&ItemID=3988&accnt=64847 (providing a lower number). 2. HOPE NOW, supra note 1; see also Chris Mayer et al., The Rise in Mortgage Defaults, 23 J. ECON. PERSP. (forthcoming 2009) (noting 1.2 million foreclosure starts in first half of 2008). 3. HOPE NOW, supra note 1; see also E-mail from Daren Blomquist, RealtyTrac, Inc., to author (Mar. 7, 2008) (on file with author). 4. HOPE NOW, supra note 1. 5. See Press Release, Mortgage Bankers Assn, Delinquencies Continue to Climb in Latest MBA National Delinquency Survey (Mar. 5, 2009), available at http://www.mbaa.org/NewsandMedia/PressCenter/68008.htm. Approximately 3.30 percent of all one-to-four-family residential mortgages outstanding were in the foreclosure process in the first quarter of 2008, and 7.88 percent were delinquent. Id.; see also Vikas Bajaj & Michael Grynbaum, A Rising Tide of Mortgage Defaults, Not All on Risky Loans, N.Y. TIMES, June 6, 2008, at C1. Because of the steadily increasing level of homeownership in the United States, see U.S. CENSUS BUREAU, HOUSING VACANCIES AND HOMEOWNERSHIP (CPS/HVS), at tbl. 14 (2008), available at http://www.census.gov/hhes/www/housing/hvs/historic/histt14.html, higher percentages of past-due and foreclosed mortgages mean that an even greater percentage of Americans are directly affected by higher delinquency and foreclosure rates. 6. CREDIT SUISSE FIXED INCOME RESEARCH, FORECLOSURE UPDATE: OVER 8 MILLION FORECLOSURES EXPECTED 1 (Dec. 4, 2008), available at http://www.chapa.org/pdf/ForeclosureUpdateCreditSuisse.pdf. Even Credit Suisses best-case scenario still involves 6.3 million foreclosures. Id. at 7.
LEVITIN - FINAL
4/23/2009 11:40 AM
568
3.5% 3.0% 2.5% 2.0% 1.5% 1.0% 0.5% 0.0%

93 95 96 97 98 00 01 02 03 04 05 06 07 20 94 19 19 19 19 19 19 19 99 20 20 20 20 20 20 20 20 08
Chart 1: Percentage of One-to-Four-Family Residential Mortgages in Foreclosure Process7 Both the increase in, and the sheer number of, foreclosures should be alarming, because foreclosures create significant deadweight loss and have major third-party externalities.8 Historically, lenders are estimated to lose from 40 to 50 percent of their investment in a foreclosure situation,9 and in the current market, even greater losses are expected.10 Borrowers lose their homes and are forced to relocate, often
7. Mortgage Bankers Assn, National Delinquency Surveys (on file with author). 8. Anthony Pennington-Cross, The Value of Foreclosed Property, 28 J. REAL ESTATE RES. 19495 (2006) (surveying estimates of deadweight loss on foreclosure). 9. See id.; see also Gretchen Morgenson, Cruel Jokes, and No One Is Laughing, N.Y. TIMES, Jan. 13, 2008 (citing historical foreclosure-loss rates of 20 to 40 percent); Posting of Treasury Secretary Henry Paulson to http://www.etrunk.kiev.ua/ask/20071207.html (Dec. 7, 2007). Actual foreclosure losses are difficult to measure consistently because loss reporting may or may not include loss of junk fees and matured interest, rather than a more consistent baseline of unpaid principal. Because most mortgages are held by securitization trusts, the losses to holders of trust securities will vary by tranche. See generally CREDIT SUISSE EQUITY RESEARCH, MORTGAGE LIQUIDITY DU JOUR: UNDERESTIMATED NO MORE (Mar. 12, 2007), available at http://billcara.com/CS%20Mar%2012%202007%20 Mortgage%20and%20Housing.pdf. Some tranches may experience no losses, while other tranches may have complete losses. See generally id. 10. FITCH RATINGS, RESIDENTIAL MORTGAGE CRITERIA REPORT: REVISED LOSS EXPECTATIONS FOR 2006 AND 2007 SUBPRIME VINTAGE COLLATERAL 1 (Mar. 25, 2008), available at http://billcara.com/CS%20Mar%2012%202007%20Mortgage%20and% 20Housing.pdf.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
569
to new communities, a move that can place extreme stress on borrowers and their families.11 Foreclosure is an undesirable outcome for borrowers and lenders. Foreclosures also impose costs on third parties. When families have to move to new homes, community ties are rent asunder. Friendships, religious congregations, schooling, childcare, medical care, transportation, and even employment often depend on geography.12 Foreclosures also depress housing and commercial-realestate prices throughout entire neighborhoods. For example, a study on foreclosures in Chicago in the late 1990s concluded that a single foreclosure depressed neighboring properties values between $159,000 and $371,000, or between 0.9% and 1.136% of the property value of all the houses within an eighth of a mile.13 For Chicago, which has a housing density of 5,076 houses per square mile,14 or around 79 per square eighth of a mile, this translates into a single foreclosure costing each of 79 neighbors between $2,012 and $4,696.
11. See, e.g., MINDY THOMPSON FULLILOVE, ROOT SHOCK 1120 (2005) (noting emotional harms from urban renewal); Lorna Fox, Re-Possessing Home: A Reanalysis of Gender, Homeownership and Debtor Default for Feminist Legal Theory, 14 WM. & MARY J. WOMEN & L. 423, 434 (2008) (The impact of losing ones home on an individual occupiers quality of life, social and identity status, personal and family relationships, and for his or her emotional, psychological and physical health and wellbeing have been well established in housing and health literature.); Eric S. Nguyen, Parents in Financial Crisis: Fighting to Keep the Family Home, 82 AM. BANKR. L.J. 229 (2008); Margaret Jane Radin, Property and Personhood, 34 STAN. L. REV. 957, 95859 (1982) (observing that control over property like homes that are bound up with ones self is essential for psychological well-being). For a critical review of the literature on homes and psychology see Stephanie Stern, Residential Protectionism and the Legal Mythology of Home, 107 MICH. L. REV. (forthcoming 2009); see also Michael Levenson, The Anguish of Foreclosure: Fearing Sale of House, Woman Kills Herself Before the Auction, BOSTON GLOBE, July 24, 2008, at B1; Ohio Woman, 90, Attempts Suicide After Foreclosure, REUTERS, Oct. 3, 2008. 12. See Phillip Lovell & Julia Isaacs, The Impact of the Mortgage Crisis on Children, FIRST FOCUS, May 2008, at 1, 1, available at http://www.firstfocus.net/Download/HousingandChildrenFINAL.pdf (estimating two million children will be impacted by foreclosures, based on a projection of 2.26 million foreclosures). 13. Dan Immergluck & Geoff Smith, The External Costs of Foreclosure: The Impact of Single-Family Mortgage Foreclosures on Property Values, 17 HOUS. POLY DEBATE 57, 58 (2006); see also MARK DUDA & WILLIAM C. APGAR, MORTGAGE FORECLOSURES IN ATLANTA: PATTERNS AND POLICY ISSUES, at ii (Dec. 15, 2005), available at http://www.nw.org/network/neighborworksProgs/foreclosuresolutions OLD/documents/foreclosure1205.pdf. 14. City-data.com, Chicago, IL (Illinois) Housing and Residents, www.citydata.com/housing/houses-Chicago-Illinois.html (last visited Mar. 24, 2009).
LEVITIN - FINAL
4/23/2009 11:40 AM
570
The property-value declines caused by foreclosure hurt local businesses and erode state and local government tax bases.15 Condominium and homeowner associations likewise find their assessment base reduced by foreclosures, leaving the remaining homeowners with higher assessments.16 Foreclosed properties also impose significant direct costs on local governments and foster crime.17 A single foreclosure can cost the city of Chicago over $30,000.18 Moreover, foreclosures have a racially disparate impact because African-Americans invest a higher share of their wealth in their homes19 and are also more likely than financially similar whites to have subprime loans.20 In short, foreclosure is an inefficient outcome that is bad not only for lenders and borrowers, but for society at large. Traditionally, bankruptcy is one of the major mechanisms for resolving financing distress. Bankruptcy creates a legal process through which the market can work out the problems created when parties end up with unmanageable debt burdens. Although the process can be a painful one for all parties involved, bankruptcy allows an orderly forum for creditors to sort out their share of losses and return the deleveraged
15. Laura Johnston, Foreclosure Study Says Vacant Properties Cost Cleveland $35+ Million, BLOG.CLEVELAND.COM, Feb. 19, 2008, http://blog.cleveland.com/ metro/2008/02/foreclosure_study_says_vacant.html; see also GLOBAL INSIGHT, THE MORTGAGE CRISIS: ECONOMIC AND FISCAL IMPLICATIONS FOR METRO AREAS 2 (Nov. 26, 2007), available at http://www.vacantproperties.org/resources/documents/ USCMmortgagereport.pdf. 16. Christine Haughney, Collateral Foreclosure Damage, N.Y. TIMES, May 15, 2008, at C1. 17. WILLIAM C. APGAR & MARK DUDA, COLLATERAL DAMAGE: THE MUNICIPAL IMPACT OF TODAYS MORTGAGE FORECLOSURE BOOM (2005), available at http://www.995hope.org/content/pdf/Apgar_Duda_Study_Short_Version.pdf; Dan Immergluck & Geoff Smith, The Impact of Single-Family Mortgage Foreclosures on Neighborhood Crime, 21 HOUS. STUD. 851, 85154 (2006). 18. APGAR & DUDA, supra note 17, at 4. 19. MELVIN L. OLIVER & THOMAS M. SHAPIRO, BLACK WEALTH/WHITE WEALTH: A NEW PERSPECTIVE ON RACIAL INEQUALITY 66 (2006) (noting that housing equity accounted for 62.5 percent of all black assets in 1988, but only 43.3 percent of white assets, even though black homeownership rates were 43 percent and white homeownership rates were 65 percent); see also Brian K. Bucks et al., Recent Changes only a $35,000 difference in median home equity between whites and nonwhites/Hispanics in 2004, there was a $115,900 difference in median net worth and a $33,700 difference in median financial assets; this suggests that for minority homeowners, wealth is disproportionately invested in the home); Kai Wright, The Subprime Swindle, NATION, July 14, 2008, at 11, 1112. 20. Mary Kane, Race and the Housing Crisis, WASH. INDEP., July 25, 2008; Bob Tedeschi, Subprime Loans Wide Reach, N.Y. TIMES, Aug. 3, 2008.
in U.S. Family Finances: Evidence from the 2001 and 2004 Survey of Consumer Finances, FED. RES. BULL., at A1, A8, A12, A23 (2006) (noting that while there was
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
571
debtor to productivity; a debtor hopelessly mired in debt has little incentive to be economically productive because all of the gain will go to creditors. Moreover, the existence of the bankruptcy system provides a baseline against which consensual debt restructurings can occur. Thus, for over a century bankruptcy has been the social safety net for the middle class, joined later by Social Security and unemployment benefits. The bankruptcy system, however, is incapable of handling the current home-foreclosure crisis because of the special protection it gives to most residential-mortgage claims. While debtors may generally modify all types of debts in bankruptcyreducing interest rates, stretching out loan tenors, changing amortization schedules, and limiting secured claims to the value of collateralthe Bankruptcy Code forbids any modification of mortgage loans secured solely by the debtors principal residence.21 Defaults on such mortgage loans must be cured and the loans then paid off according to their original terms, including all fees that have been levied since default, or else the bankruptcy stay on collection actions will be lifted, permitting the
21. 11 U.S.C. 1322(b)(2) (2006); cf. id. 1123(b)(5) (parallel residential mortgage antimodification provision for Chapter 11). Section 1123(b)(5) provides that a plan of reorganization may modify the rights of holders of secured claims, other than a claim secured only by a security interest in real property that is the debtors principal residence. Id. 1123(b)(5). Since 2005, section 101(13A) of the Bankruptcy Code has defined debtors principal residence as a residential structure, including incidental property, without regard to whether that structure is attached to real property . . . and . . . includes an individual condominium or cooperative unit, a mobile or manufactured home, or trailer. Id. 101(13A). State law, however, still determines what real property is. Modification of a principal residence is even permitted per 11 U.S.C. 1322(c)(2) in cases where the last payment on the contractual payment schedule is due before the final payment on the plan. Am. Gen. Fin., Inc. v. Paschen (In re Paschen), 296 F.3d 1203 (11th Cir. 2002); First Union Mortgage Corp. v. Eubanks (In re Eubanks), 219 B.R. 468 (B.A.P. 6th Cir. 1998). But see Witt v. United Cos. Lending Corp. (In re Witt), 113 F.3d 508 (4th Cir. 1997) (holding that 11 U.S.C. 1322(c)(2) does not permit bifurcation of claims despite its enactment subsequent to the Supreme Court of the United Statess Nobelman decision). It is unclear whether the antimodification provision prevents an undersecured mortgagee from receiving postpetition interest and fees under 11 U.S.C. 506(b). Compare Campbell v. Countrywide Home Loans, Inc. (In re Campbell), 361 B.R. 831, 850 (Bankr. S.D. Tex. 2007) (noting that section 1322(b)(2) trumps section 506(b)), with Citicorp Mortgage, Inc. v. Hunt, No. 5:92CV56(JAC), 1994 U.S. Dist. LEXIS 13146, at *89 (D. Conn. 1994) (noting that section 1322(b)(2) does not vitiate section 506(b)). It would seem, however, that the legal fiction engendered by Nobelman v. American Savings Bank s interpretation of section 1322(b)(2) does not require anything beyond treating a principal-home mortgage as fully secured; it need not be treated as oversecured, and if fully secured to the exact amount, postpetition interest and fees could not accrue. 508 U.S. 324 (1993).
LEVITIN - FINAL
4/23/2009 11:40 AM
572
mortgagee to foreclose on the property.22 As a result, if a debtors financial distress stems from a home mortgage, bankruptcy is unable to help the debtor retain her home, and foreclosure will occur. The absence of a bankruptcy-modification option also reduces the incentive for creditors to engage in consensual nonbankruptcy debt restructuring. Because of bankruptcys special treatment of principal residential mortgages, the legal mechanism on which the market depends for sorting through debt problems cannot function properly, and this is exacerbating the impact of the mortgage crisis. The Bankruptcy Codes special protection for home-mortgage lenders reflects a hitherto unexamined economic assumption. The assumption is that preventing modification of home-mortgage loans in bankruptcy limits lenders losses and thereby encourages greater mortgage credit availability and lower mortgage credit costs,23 in turn encouraging the homeownership that has been a major goal of federal economic policy for the past half century.24 As Justice John Paul Stevens noted when the Supreme Court of the United States addressed the Bankruptcy Codes antimodification provision in 1993:
22. See 11 U.S.C. 362(d)(1) (authorizing the lifting of an automatic stay for cause if adequate protection cannot be provided). Bankruptcy allows the homeowner to unwind any acceleration on the loan, however. Id. 1322(c). Therefore, if the homeowners problems stem not from a generally unaffordable mortgage payment, but from a temporary loss of income or unexpected one-time expense, bankruptcy can still provide the homeowner with the breathing space to straighten out her finances, deaccelerate, cure, and reinstate the mortgage. 23. See, e.g., Donald C. Lampe et al., Introduction to the 2008 Annual Survey of Consumer Financial Services Law, 63 BUS. LAW. 561, 568 (2008) (Solutions designed to prevent future problems by reducing the availability of credit to marginal borrowers may (in addition to affecting adversely those future borrowers) worsen the current plight of existing marginal borrowers who need to refinance their homes. Direct relief for troubled borrowers, e.g., a foreclosure moratorium or expanded bankruptcy relief, may have the same effect. To some extent this has already happened. The tightening of mortgage law requirements and regulatory restrictions over the past few years in response to allegations of predatory lending have probably contributed to the dramatic increase in foreclosures by making it more difficult for troubled borrowers to refinance. A significant further tightening of these restraintswe have heard this further tightening referred to as more robust regulationmay worsen the problem and increase the number of consumers facing foreclosure as a result.). 24. See, e.g., Consumer Issues in Bankruptcy: Hearing Before the Subcomm. on Economic and Commercial Law of the H. Comm. on the Judiciary, 102d Cong. 56 57 (1992) (statement of Willard Gourley, Jr., BarclaysAmerican Mortgage Corp., on behalf of the Mortgage Bankers Association of America); Douglas G. Baird, Technology, Information, and Bankruptcy, 2007 U. ILL. L. REV. 305, 307 (2007); J. Peter Byrne & Michael Diamond, Affordable Housing, Land Tenure, and Urban Policy: The Matrix Revealed, 34 FORDHAM URB. L.J. 527, 542 (2007); Kenya Covington & Rodney Harrell, From Renting to Homeownership: Using Tax Incentives to Encourage Homeownership Among Renters, 44 HARV. J. ON LEGIS. 97, 99 (2007).
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
573
At first blush it seems somewhat strange that the Bankruptcy Code should provide less protection to an individuals interest in retaining possession of his or her home than of other assets. The anomaly is, however, explained by the legislative history indicating that favorable treatment of residential mortgagees was intended to encourage the flow of capital into the home lending market.25 According to Justice Stevens, Congress intended to promote mortgage lending by limiting lender losses in bankruptcy. Justice Stevenss assertion has scant support in the legislative history,26 but has
25. Nobelman, 508 U.S. at 332 (Stevens, J., concurring). 26. The legislative history that Justice Stevens relied on in his Nobelman concurrence was roughly outlined in Grubbs v. Houston First American Savings Assn, 730 F.2d 236, 24546 (5th Cir. 1984). The legislative history of 11 U.S.C. 1322(b)(2) is scant. The Bankruptcy Reform Act of 1978, Pub. L. No. 95-598, 92 Stat. 2549, was a compromise between House and Senate bills. The House bill, H.R. 8200, permitted modification of all secured claims, while the Senate bill, S.B. 2266, contained a provision barring any modification of claims secured by real estate. Grubbs, 730 F.2d at 245. The bills were reconciled through a series of floor amendments, id. at 246, which resulted in a ban on modification of loans secured solely by the debtors principal residence. There was no discussion on the issue in the Congressional Record. It was, however, raised in Senate hearings, most notably in a dialogue amongst Edward J. Kulik, Senior Vice President, Real Estate Division, Massachusetts Mutual Life Insurance Co.; his counsel, Robert E. OMalley, of Covington & Burling; and Senator Dennis DiConcini (D-Ariz.). The dialogue is worth reproducing because it is virtually the only evidence of congressional intent and is less than overwhelming: MR. KULIK: These provisions may cause residential mortgage lenders to be extraordinarily conservative in making loans in cases where the general financial resources of the individual borrower are not particularly strong. Serious consideration should be given to modifying both bills so that, at the least: One, a mortgage on real property other than investment property may not be modified . . . . SENATOR DICONCINI: If [the cramdown provision was] not change[d], you do not really suggest that savings and loans and mortgage bankers will stop lending money, do you? MR. KULIK: Mr. Chairman, I would have to speak for the life insurance industry. I think we would channel more of our funds into direct placements and bond purchases and stay away from mortgages, particularly where limited partnerships are the borrowers. SENATOR DICONCINI: But not as to individuals? MR. KULIK: Let me ask counsel, but I believe individuals are in the same category as limited partnerships. Counsel tells me it is not as serious for individuals.
LEVITIN - FINAL
4/23/2009 11:40 AM
574
SENATOR DICONCINI: But notwithstanding, there have been times that savings and loan associations in Arizona have really been anxious to loan money, notwithstanding the present law. That happens to be the case right now. That has not always been the case. So, I wonder what really detrimental effect there is. The last part of your statement left me with the indication that this is so severe that if we do not do something, it will severely strangle the home loan mortgage business. I realize the severity of your problem. Your statement is excellent, but I challenge the fact that it is as severe as you left me with the last closing statement that you made. MR. KULIK: Mr. Chairman, counsel has asked to reply. MR. OMALLEY: Mr. Chairman, would you indulge me to speak to that point? SENATOR DICONCINI: Certainly. MR. OMALLEY: With respect to the savings and loans, in particular, and the future prospects for loans to individuals under the proposed bills, there is really only one basic problem. That is, the provision in both bills that provides for modification of the rights of the secured creditor on residential mortgages, a provision that is not contained in present law. I think the answer to your question is that, of course, savings and loans will continue to make loans to individual homeowners, but they will tend to be, I believe, extraordinarily conservative and more conservative than they are now in the flow of credit. It seems to me they will have to recognize that there is an additional business risk presented by either or both of these two bills if the Congress enacts chapter XIII in the form proposed, thus providing for the possibility of modification of the rights of the secured creditor in the residential mortgage area. I think the answer is that they will be much more conservative than they have been in the past.
Hearings on S. 2266 and H.R. 8200 Before the Subcomm. on Improvements in Judicial Mach. of the S. Comm. on the Judiciary, 95th Cong. 71415 (1977) (testimony
of Edward J. Kulik, Senior Vice President, Real Estate Division, Massachusetts Mutual Life Insurance Co., accompanied by Robert E. OMalley, Attorney, Covington & Burling). See also Hearings Before the Subcomm. on Improvements of the Judicial Mach. of the S. Comm. on Judiciary, 94th Cong. 12728, 130, 13234, 13738 (1975) (statement of Walter Vaughan, American Bankers Association); id. at 14142, 16768, 17680 (statement of Alvin Wiese, National Consumer Finance Association); id. at 64 (written testimony of Conrad Cyr, a bankruptcy judge in Maine (now a federal appeals court judge)) (arguing that modification of real-property mortgages should be permitted in Chapter 13). Despite Senator DeConcinis apparent incredulity, based on his unfortunately intimate knowledge of the savings and loan industry, this dialogue appears to be the basis for assuming a particular policy basis for the antimodification provision. As Grubbs noted, This limited bar [on modification] was apparently in response to perceptions, or to suggestions advanced in the legislative hearings . . . that, home-
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
575
nonetheless become the dominant explanation for the Bankruptcy Codes mortgage antimodification provision. Underlying the economic assumption embedded in the Bankruptcy Codes antimodification provision is another assumptionthat mortgage markets are sensitive to bankruptcy-modification risk. This Article empirically tests the policy assumption behind the Bankruptcy Codes prohibition on the modification of single-family primary-residence mortgages. It marshals a variety of original empirical evidence from mortgage origination, insurance, and resale markets to show that mortgage markets are indifferent to bankruptcy-modification risk.
mortgagor lenders, performing a valuable social service through their loans, needed special protection against modification thereof . . . . 730 F.2d at 246. The evidence for a deliberate policy preference by Congress is very limited. The relationship of section 1322(b)(2) to pre-1978 bankruptcy law is consistent with either an interpretation of it as an explicit and deliberate policy preference by Congress or a special-interest provision. Section 1322(b)(2) appears, at first blush, to continue pre-1978 bankruptcy law, which functionally forbade any modification of a mortgage. But an examination of the details of pre-1978 law shows that the story is more complex. Chapter XIII of the Bankruptcy Act of 1898 was the predecessor of the modern Chapter 13. (Chapter XIII was enacted as part of the Chandler Act on June 22, 1938. Ch. 575, 52 Stat. 840). In a Chapter XIII wage earners plan, there were no statutory limitations on the ability to modify a secured debt, including a mortgage. 11 U.S.C. 646 (1976) (repealed 1978). However, a wage earners plan that affected a secured debt could not be confirmed without the consent of the affected secured creditor. Id. 1052(1) (repealed in 1978). Therefore, it was impossible to modify a mortgage or any other secured debt without the consent of the impaired creditor. In both the House and Senate versions of the Bankruptcy Reform Act, there was no blanket prohibition on the modification of secured debt. In the House version, all secured debts could be modified, just as in Chapter XIII, H.R. 8200, 95th Cong. 132223 (1978), while the Senate version generally permitted the modification for secured debts, but with an exclusion for mortgages. S. 2266, 95th Cong. 132223 (2d Sess. 1978). In both versions, the impaired secured creditors veto was eliminated. Instead, under the final version of Chapter 13, the plan is confirmable if it meets certain statutory requirements for the treatment of secured debts, regardless of the affected secured creditors consent. This history shows that the mortgage antimodification provision is not a continuation of pre-Code bankruptcy law. Pre-Code law permitted modification, subject to a creditor veto, but that was for all secured debts. 11 U.S.C. 1046, 1052(1) (repealed 1978). The Bankruptcy Code instead permits modification without any creditor veto, but carves out one particular class of debts that cannot be modified under a plan, regardless of creditor consent. 11 U.S.C. 1322(b)(2) (2006). Had Congress wished to continue pre-Code law just for mortgages, it could have retained a veto for mortgage creditors. It did not. The history of the antimodification provision and the change from the 1898 Acts veto to the Codes prohibition indicates that section 1322(b)(2) is actually something new, not a continuation of prior practice. While Grubbs assumes otherwise, there is no conclusive evidence in the legislative history that section 1322(b)(2) was intended to lower the cost of mortgage credit or increase its availability. At best, we can say that it was the result of a legislative compromise that gave a subset of mortgage creditors a more favorable position relative to other creditors than they had under the 1898 Act.
LEVITIN - FINAL
4/23/2009 11:40 AM
576
The Article explains this indifference by reference to data on the relative losses lenders incur in modification and foreclosure, and argues that as long as lenders face larger losses in foreclosure than modification, the mortgage market will not price and ration credit based on bankruptcy-modification risk. Accordingly, this Article argues that the Bankruptcy Code should be amended to permit debtors to modify all mortgages. Such an amendment would provide the most effective, fair, immediate, and tax-payer-cost-free tool for resolving the homemortgage crisis. In a perfectly functioning market without agency and transaction costs, lenders would be engaged in large-scale modification of defaulted or distressed mortgage loans, as the lenders would prefer a smaller loss from modification than a larger loss from foreclosure. Voluntary modification, however, has not been happening on a large scale27 for a variety of reasons,28 most notably contractual impediments,29 agency costs, practical impediments, and other transaction costs.30 If all distressed mortgages could be modified in bankruptcy, it would provide a method for bypassing the various contractual, agency, and other transactional inefficiencies. Permitting bankruptcy modification would give homeowners the option to force a workout of the mortgage, subject to the limitations provided by the Bankruptcy Code. Moreover, the possibility of a bankruptcy modification would encourage voluntary modifications, as mortgage lenders would prefer to exercise more control over the shape of the modification. An involuntary public system of mortgage modification would actually help foster voluntary, private solutions to the mortgage crisis. Unlike programs for government refinancing or insurance of distressed mortgages, the bankruptcy system is immediately available to resolve the mortgage crisis. Government refinancing or insurance plans would take months to implement, during which time foreclosures would continue. In contrast, bankruptcy courts are experienced, up-and27. See, e.g., COMPTROLLER OF THE CURRENCY, OCC MORTGAGE METRICS REPORT: ANALYSIS AND DISCLOSURE OF NATIONAL BANK MORTGAGE LOAN DATA 1, 8 (2008), available at http://www.occ.treas.gov/ftp/release/2008-65b.pdf; see also infra Charts 6 and 7. 28. See generally Kurt Eggert, Limiting Abuse and Opportunism by Mortgage Servicers, 15 HOUS. POLY DEBATE 753 (2004). 29. See generally Anna Gelpern & Adam J. Levitin, Rewriting Frankenstein Contracts: The Workout Prohibition in Residential Mortgage-Backed Securities, 82 S. CAL. L. REV. (forthcoming 2009), available at http://papers.ssrn.com/sol3/ papers.cfm?absabstr_id=1323546. See generally CONG. OVERSIGHT PANEL, THE FORECLOSURE CRISIS: 30. WORKING TOWARD A SOLUTION 3748 (2009) (discussing reasons for failure of private modification efforts).
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
577
running, and currently overstaffed relative to historic caseloads.31 Moreover, the bankruptcy automatic stay would immediately halt any foreclosure action in process upon a homeowners filing of a bankruptcy petition.32 And, unlike government guarantees or refinancing, bankruptcy modification of all mortgages would not involve the public fisc. Bankruptcy modification would also avoid the moral hazard for lenders and borrowers of a bailout. Lenders would incur costs for having made poor lending decisions through limited recoveries. Borrowers would face the shame of filing, their finances becoming a matter of public record, the requirement of living for three to five years on a court-supervised budget in which all disposable income goes to creditors, a damaged credit rating, and the inability to file for bankruptcy for a number of years. Bankruptcy modification also provides an excellent device for sorting out types of mortgage debtors. It can correct the two distinct mortgage problems in the current crisispayment-reset shock from resetting adjustable rate mortgages (ARMs), and negative equity from rapidly depreciating home priceswhile preventing speculators and vacation-home purchasers from enjoying the benefits of modification. And, by providing an efficient and fair system for restructuring debts and allocating losses, bankruptcy will help stabilize the housing market. Making bankruptcy a forum for distressed homeowners to restructure their mortgage debts is both the most moderate and the best method for resolving the foreclosure crisis and stabilizing mortgage markets. Permitting modification of all mortgages in bankruptcy would thus create a low-cost, effective, fair, and immediately available method for resolving much of the current foreclosure crisis without imposing costs on the public fisc or creating a moral hazard for borrowers or lenders.
31 See Michelle J. White, Bankruptcy: Past Puzzles, Recent Reforms, and the Mortgage Crisis 18 (Natl Bureau Econ. Research, Working Paper No. 14549, 2008), available at http://www.nber.org/papers/w14549. In 2005, there were 2,078,415 total bankruptcy filings, the highest level on record. USCourts.gov, Bankruptcy Statistics, http://www.uscourts.gov/bnkrpctystats/statistics.htm#june (last visited Mar. 24, 2009). Most of these filings came in a spike before the effective date of the Bankruptcy Abuse Prevention and Consumer Protection Act of 2005 (BAPCPA), Pub. L. No. 109-8, 119 Stat. 23. BAPCPA added twenty-eight new bankruptcy judgeships to the existing 296. Id. 1223. In 2005, the courts were able to handle a significant increase in case filings with fewer judges. The increase in the number of bankruptcy judges means that the courts are better equipped today to handle another surge in filings. 32. 11 U.S.C. 362(a).
LEVITIN - FINAL
4/23/2009 11:40 AM
578
This Article proceeds in four Parts. Part I reviews the state of the law on mortgage modification in bankruptcy and the structure and cast of characters in the mortgage market. Part II tests the economic assumption behind the Bankruptcy Codes prohibition on single-family, primary-residence mortgage modification in two ways. First, it examines whether current mortgage-market pricing from the origination market, the secondary market, and the private-mortgage-insurance (PMI) market reflects the risk of modification in bankruptcy that attaches to multifamily properties, vacation homes, and investor properties, but not single-family, owner-occupied principal residences. Second, Part II examines historical mortgage and bankruptcy-filing data from a period when strip-down, a particularly significant type of modification, was permitted in approximately half of the federal judicial districts. Taken together, the current market pricing and the historical data indicate that mortgage markets are largely indifferent to bankruptcy-modification outcomes. The current market data suggest almost complete indifference, whereas the historical data show some sensitivity, particularly for higher-price and higher-loan-to-value-ratio (i.e., riskier) borrowers. Part II includes a significant amount of technical, detailed mortgage-rate analysis that provides important evidence for the Articles argument, but which may not be of interest to all readers. Readers who are not concerned with the technical analysis should skip to Part III. Part III addresses why mortgage markets are so indifferent to bankruptcy-modification risk. Using data from Chapter 13 bankruptcy filings, it examines the impact of permitting strip-down on mortgage lenders and shows that its impact is usually far less than the lenders would lose in foreclosure. Because lenders would generally fare better in bankruptcy than in foreclosure, they do not price adversely to a bankruptcy-modification option. Part IV considers the policy implications of market indifference to bankruptcy-modification risk and compares bankruptcy modification to other proposed solutions to the mortgage crisis, establishing that bankruptcy modification offers unparalleled advantages over other potential solutions. This Article concludes by positing a new theory of the relationship of consumer finance and bankruptcy law, namely that changes in the structure of the lending market mean that bankruptcy law has little effect on consumer credit, so consumer-bankruptcy policy should not be guided by concerns over credit constriction. Two appendices contain illustrative data and a debunking of the Mortgage Bankers Associations claims about the likely impact of bankruptcy modification on mortgage interest rates and credit availability.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
I. THE STRUCTURE OF THE MORTGAGE MARKET
579
A. Treatment of Mortgages in Bankruptcy

There are two main types of consumer bankruptcy, Chapter 7 liquidations and Chapter 13 repayment plans.33 In Chapter 7, the debtor surrenders all nonexempt assets for distribution to creditors.34 In most circumstances, this means that a Chapter 7 debtor will not be able to retain her home. In Chapter 13, in contrast, the debtor retains all of her property, but must devote all disposable personal income for the next three or five years to repaying creditors under a court-supervised repayment plan and budget.35 Chapter 13, then, is the type of bankruptcy generally suited for a debtor seeking to retain major property, such as a residence.36 Debtors in Chapter 13 repayment-plan bankruptcies are able to modify almost all types of debts. This means they can change interest rates, amortization, and terms of loans.37 They can also strip-down debts secured by collateral to the value of the collateral.38 Strip-down bifurcates an undersecured 39 lenders claim into a secured claim for the value of the collateral and a general unsecured claim for the deficiency. In Chapter 13, a creditor is guaranteed to receive the value of a secured
33. Consumers are also eligible for Chapter 11, but few who are eligible for Chapter 13, see id. 109(e), use Chapter 11 because of greater creditor control in Chapter 11 through creditor voting rights on plans of reorganization. 34. Id. 522(b), 541. 35. Id. 1325(b). 36. A debtor may enter into a court-approved reaffirmation agreement with a creditor and retain nonexempt property in exchange for continuing to make payments to the lender. Id. 524(c). Alternatively, a debtor may redeem collateral in bankruptcy by paying the lender the value of the property. Id. 722; cf. UCC 9-623 (2008) (illustrating that redemption at state law requires paying the full amount outstanding on the loan plus reasonable lender expenses). 37. 11 U.S.C. 1322(b)(2). 38. See id. 506. Strip-down is synonymous with lien-stripping and cramdown. Because cramdown has a distinct meaning in the context of Chapter 11 reorganizationsthe confirmation of a plan of reorganization absent approval of all impaired classes of creditors and equity-holders under the provisions of 11 U.S.C. 1129(b)I use the term strip-down. 39. A loan is undersecured if the amount owned on the loan is more than the value of the collateral securing the loan. If there is no collateral securing the loan, the loan is unsecured. Undersecured lenders and loans are also referred to as upsidedown or underwater. The homeowner in such a situation has negative equity. If there are multiple mortgages on the property, it is possible for the homeowner to have negative equity even though the senior mortgage is still oversecured.
LEVITIN - FINAL
4/23/2009 11:40 AM
580
claim.40 In contrast, general unsecured claims are guaranteed only as much as would be paid out in a Chapter 7 liquidation, which is often mere cents on the dollar or nothing at all.41 Therefore, strip-down can function like a reduction in the principal amount owed on a debt. Strip-down is thus the most significant type of modification because it affects the treatment of the principal amount of the creditors claim, not just the interest. Because of compound interest, a strip-down of X percent of the principal will have a larger impact on the total return than a modification of X percent of the interest rate. Moreover, because most residential mortgages are prepaid (typically through refinancing) in their first seven years,42 an adjustment to principal is more significant than an adjustment to interest rates. A reduction in secured principal has an immediate impact on the loan, whereas a reduction in the interest rate is spread out for the full term of the loan, so if the loan is prepaid, the reduction in interest income is curtailed.
40. Id. 1325(a)(5)(B). Section 1325(a)(5) only applies to allowed secured claims provided for by the plan. Therefore, it is possible for a debtor to deal with a secured claim outside of a plan and not subject to 1325(a)(5)s requirement. It is also important to note that courts have split regarding whether section 1325(a)(5) requires that a debt be paid off under a plan. For non-principal-residence mortgages, debtors will sometimes propose to modify the mortgage through strip-down and then maintain payments on it pursuant to the original term of the loan (reamortized or not). Compare Enewally v. Wash. Mut. Bank (In re Enewally), 368 F.3d 1165, 1172 (9th Cir. 2004) (holding that a modified mortgage on rental property must be paid off during a plan), with In re McGregor, 172 B.R. 718, 721 (Bankr. D. Mass. 1994) (suggesting that a debtor might amend a plan to provide for payments on stripped-down principal at the original contract interest rate in the amount called for by the mortgage contract until the total principal payments equaled the allowed amount of the secured claim). An argument that courts do not appear to have considered is that section 1325(a)(5)(B)(ii) merely requires distribution of value equal to the allowed amount of a secured claim, and that this value could be in the form of a note with a term longer than the duration of the plan. As Justice Thomas observed in his plurality concurrence in Till v. SCS Credit Corporation, 541 U.S. 465 (2004), an auto-loanstrip-down case, nothing in 1325 suggests that property is limited to cash. Rather, property can be cash, notes, stock, personal property or real property; in short, anything of value. Id. at 488 (Thomas, J., concurring). The payment of a secured debt with a new note is frequently done in Chapter 11 under the parallel provision for distributions to secured creditors, section 1129(a)(7)(A)(ii). Perhaps the best evidence that cash is not required is that the Bankruptcy Code specifically calls for cash distributions in other circumstances, such as requiring cash payments for administrative expense claims. See 11 U.S.C. 1129(a)(9), 1322(a)(2). If cash payment is not required, then distributions could be made during the plan that would essentially transform the old debt into a new one with a duration longer than the plan. 41. Id. 1325(a)(4). See Wesley Phoa, A Practical Guide to Relative Value for Mortgages, in 42. MANAGING FIXED INCOME PORTFOLIOS 357, 363 (Frank J. Fabozzi ed., 1997).
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
581
Chapter 13 provides debtors with a very broad ability to restructure their debts. There is a significant limitation, however, for certain home mortgages. Section 1322(b)(2) of the Bankruptcy Code provides that a Chapter 13 repayment plan may modify the rights of holders of secured claims, other than a claim secured only by a security interest in real property that is the debtors principal residence.43 Section 1322(b)(2) thus prevents modification only of mortgages secured solely by real property that is the debtors principal residence.44 Under current law, debtors can modify mortgages on vacation homes, investor properties, and multifamily residences in which the owner occupies a unit.45 Debtors can also currently modify wholly
43. 11 U.S.C. 1322(b)(2). For the history of this provision, see Veryl Victoria Miles, The Bifurcation of Undersecured Residential Mortgages Under 1322(b)(2) of the Bankruptcy Code: The Final Resolution, 67 AM. BANKR. L.J. 207 (1993). 44. 11 U.S.C. 1322(b)(2). Sections 1322(b)(2) and 1325(a)(5) place limitations on the modification of all mortgages. It is important to note that the protections given mortgage holders depend on owner-occupancy status, so mortgage holders protections are dependent upon debtor cooperation, a factor upon which mortgage holders cannot justifiably rely. 45. See, e.g., Scarborough v. Chase Manhattan Mortgage Corp. (In re Scarborough), 461 F.3d 406, 408, 41213 (3d Cir. 2006) (permitting strip-down on a two-unit property in which the debtor resided); Chase Manhattan Mortgage Corp. v. Thompson (In re Thompson), 77 Fed. Appx. 57, 58 (2d Cir. 2003) (permitting stripdown on a three-unit property in which the debtor resided); Lomas Mortgage, Inc. v. Louis, 82 F.3d 1, 7 (1st Cir. 1996) (permitting strip-down on a three-unit property in which the debtor resided); First Nationwide Mortgage Corp. v. Kinney (In re Kinney), No. 3:98CV1753(CFD), 2000 U.S. Dist. LEXIS 22313, *1113 (D. Conn. Apr. 12, 2000) (permitting modification of a two-unit property in which the debtor resided); In re Stivender, 301 B.R. 498, 500 n.2 (Bankr. S.D. Ohio 2003) (permitting bifurcation on a two-unit property containing the debtors residence); Enewally v. Wash. Mut. Bank (In re Enewally), 276 B.R. 643, 652 (Bankr. C.D. Cal. 2002) (holding a mortgage on a rental property that is not the debtors residence may be modified); In re Kimbell, 247 B.R. 35, 38 (Bankr. W.D.N.Y 2000) (permitting bifurcation on a twounit property containing the debtors residence); Ford Consumer Fin. Co. v. Maddaloni (In re Maddaloni), 225 B.R. 277, 278 (D. Conn. 1998) (permitting bifurcation on a two-unit property containing the debtors residence); In re Del Valle, 186 B.R. 347, 34850 (Bankr. D. Conn. 1995) (permitting modification of a two-unit property, where the debtor lived in one unit and rented the other); Adebanjo v. Dime Sav. Bank, FSB (In re Adebanjo), 165 B.R. 98, 100 (Bankr. D. Conn. 1994) (permitting bifurcation on a three-unit property containing the debtors residence); In re McGregor, 172 B.R. 718, 721 (Bankr. D. Mass. 1994) (permitting modification of a mortgage of a four-unit apartment building in which the debtor resided); Zablonski v. Sears Mortgage Corp. (In re Zablonski), 153 B.R. 604, 606 (Bankr. D. Mass. 1993) (holding that a mortgage encumbering a two-family home was not protected from modification); In re McVay, 150 B.R. 254, 25657 (Bankr. D. Or. 1993) (failing to protect a mortgage encumbering a bed and breakfast, which was the debtors principal residence but which had inherent income producing power, from modification).
LEVITIN - FINAL
4/23/2009 11:40 AM
582
unsecured second mortgages on their principal residences,46 as well as loans secured by yachts, aircraft, jewelry, household appliances, furniture, vehicles, or any other type of personalty.47 The Bankruptcy Code, however, forbids the modification of mortgage loans secured solely by the debtors principal residence.48 Such mortgage loans must be cured and then paid off according to their original terms, including all fees that have been levied since default, or else the bankruptcy automatic stay will be lifted, permitting the mortgagee to foreclose on the property.49 As a result, if a debtors financial distress stems from an unaffordable home mortgage, bankruptcy is unable to help the debtor retain her home, and foreclosure will occur.
B. Structure of the Modern Mortgage Market

Mortgage markets reaction to bankruptcy-modification risk is shaped by the structure of modern mortgage markets. Traditionally, mortgage lenders looked like the Bailey Building & Loan Association of Bedford Falls (Bailey) in Frank Capras classic Christmas film, Its a Wonderful Life.50 Bailey would make mortgage loans to borrowers in the Bedford Falls community and keep the loans on its books as
46. Every federal circuit court of appeals addressing the issue has held that modification, including strip-down, of wholly unsecured second mortgages on principal residences is permitted. See, e.g., Lane v. W. Interstate Bancorp (In re Lane), 280 F.3d 663, 669 (6th Cir. 2002); Zimmer v. PSB Lending Corp. (In re Zimmer), 313 F.3d 1220, 1227 (9th Cir. 2002); Pond v. Farm Specialist Realty (In re Pond), 252 F.3d 122, 126 (2d Cir. 2001); Bartee v. Tara Colony Homeowners Assn (In re Bartee), 212 F.3d 277, 288 (5th Cir. 2000); McDonald v. Master Fin., Inc. (In re McDonald), 205 F.3d 606, 608 (3d Cir. 2000); Tanner v. FirstPlus Fin., Inc. (In re Tanner), 217 F.3d 1357, 1360 (11th Cir. 2000); Lam v. Investors Thrift (In re Lam), 211 B.R. 36, 41 (B.A.P. 9th Cir. 1997). These are known as the Son of Strip-down cases. 47. 11 U.S.C. 1322(b)(2). Until 2005, loans secured by all vehicles could be stripped down. Since October 17, 2005, purchase money loans secured by motor vehicles may not be stripped down in their first two-and-a-half years, and other purchase money secured loans may not be stripped down in their first year. Id. 1325(a). 48. See supra note 21. 49. See 11 U.S.C. 362(d). Bankruptcy allows the homeowner to unwind any acceleration on the loan, however. Id. 1322(c). Therefore, if the homeowners problems stem not from a generally unaffordable mortgage payment, but from a temporary loss of income or unexpected one-time expense, bankruptcy can still provide the homeowner with the breathing space to straighten out her finances, deaccelerate, and reinstate the mortgage. 50. RKO Radio Pictures 1946.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
583
assets.51 Bailey had a portfolio of Bedford Falls mortgages and longstanding banking relationships with its borrowers.52 Bailey faced two key problems with its business model. First, the mortgage loans held by Bailey were largely illiquid, long-term assets, so Bailey could not sell them to improve its short-term cash flow.53 Second, due to interstate-banking restrictions, all the loans were made in the Bedford Falls area, which meant they were not diversified, making Bailey heavily exposed to the overall economic conditions of Bedford Falls.54 As a result, Bailey had to be careful about its lending, but having community-based relationships with its borrowers gave it additional information for sound loan underwriting.55 Over the past quarter century or so, the mortgage industrys traditional relational-portfolio-lending model has been replaced with an originate-to-distribute (OTD) model designed to increase liquidity and portfolio diversification in mortgage lending.56 The OTD mortgage market involves a cast of several players: originators; private mortgage insurers; and secondary-market securitizers, including governmentsponsored entities (GSEs), mortgage-pool insurers, mortgage-backed security investors, and servicers. First, there are the financial institutions that advance (or, in mortgage-industry parlance, originate) the mortgage loan to the homeowner, sometimes directly, and sometimes through mortgage brokers. These institutions include commercial banks, savings banks, credit unions, finance companies, and mortgage banks. If the loan-tovalue (LTV) ratio on the mortgage exceeds 80 percent, mortgage insurance will generally be required to make the loan marketable on the secondary market.57 Usually the mortgage insurance is purchased by the
51. Id. Before 1994, there was no interstate branch banking, so most lending was local. See Riegle-Neal Interstate Banking and Branching Efficiency Act of 1994, Pub. L. No. 103-328, 109(c), 108 Stat. 2338, 2338 (codified at 12 U.S.C. 1835(a)). 52. ITS A WONDERFUL LIFE (RKO Radio Pictures 1946). 53. Id. Id. 54. See id. 55. 56. See infra Chart 2. 57. Fannie Mae and Freddie Mac, the government-sponsored entities that dominate the secondary mortgage market, will not purchase standard mortgages with LTVs of more than 80 percent without mortgage-insurance coverage. FANNIE MAE, SINGLE FAMILY 2007 SELLING GUIDE, at pt. V, 101.01 (2007), available at https://www.efanniemae.com/sf/guides/ssg (under Online Access click Access AllRegs and choose the appropriate part and section); FREDDIE MAC, SINGLE-FAMILY SELLER/SERVICER GUIDE, at 27.1, (2009), available at http:// www.freddiemac.com/sell/guide/# (under Access the Guide click AllRegs and choose the appropriate chapter).
LEVITIN - FINAL
4/23/2009 11:40 AM
584
homeowner, with the mortgagee as the payee, but it can also be lender purchased.
100% Market Share ($ Outstanding) 80% 60% 40% 20% 0%
1 Q 20 08
80
84
88
60
64
68
72
76
92
52
56
96
00
19
19
19
19
19
19
19
19
19
19
19
19
20
Private Label MBS Pools GSEs Savings Institutions and Credit Unions Household Sector
Agency & GSE MBS Pools Commercial Banks Insurance Companies Other
Chart 2: Holders of Residential One-to-Four-Family Mortgages by Entity Type58 Sometimes originators hold the mortgage loans on their own books, but often (indeed, almost always, in the case of mortgage banks) they sell them into the secondary market. Sometimes this is done through a direct securitization, in which the originator sells a pool of mortgage loans to a specially created entity (special purpose vehicle or SPV), typically a trust (a securitization trust).59 The SPV pays for the mortgage loans by selling securities, secured by the pooled mortgages, to capital-market investors.60 These securities are the mortgage-backed securities (MBS) because they are collateralized or backed by the mortgages in the pool. Often, however, the originator does not securitize the loans directly. Instead, the originator sells the loans to either a GSE like Fannie Mae or Freddie Mac or to a private securitization conduit, such
58. See Fed. Reserve Bank, Statistical Release, No. Z.1, Flow of Funds Accounts of the United States, at tbl.L.217 (Dec. 11, 2008), available at http://www.federalreserve.gov/releases/z1/Current/accessible/1217.htm. 59. VINOD KOTHARI, SECURITIZATION: THE FINANCIAL INSTRUMENT OF THE FUTURE 11 (2006). Id. 60.
20
04
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
585
as an investment bank.61 Sometimes these entities retain the mortgages in their own portfolios, but they generally pool mortgages originated by many different originators and undertake a multiconduit securitization. Sometimes MBS are themselves pooled and securities are issued against a pool of MBS, in what is called a collateralized debt obligation (CDO).62 It is estimated that 75 percent of outstanding first-lien residential mortgages are held by securitization trusts, and that twothirds are in GSE MBS.63 The MBS issued by the SPVs are typically divided into slices or tranches of various risk, based on senior or subordinated status.64 The price of the various MBS tranches is largely determined by the ratings given to them by rating agencies like Moodys, Standard and Poors, and Fitch Ratings.65 The credit ratings of various MBS tranches are often enhanced through myriad credit-enhancement mechanisms, such as overcollateralization or pool-levelstop-loss bond insurance.66 Although the securitizer does not carry the mortgages it has sold to the SPV on its books, it frequently keeps a relationship with the mortgages it has sold by entering into a pooling and servicing agreement (PSA) with the SPV.67 Because the SPV exists only on paper, it needs an agent to manage and collect (or service) the loans it owns.68 The PSA is the contract that creates the principal-agent relationship between the SPV and the servicer. The originator will thus service the loans even though it does not hold them on its books, creating a potential moral hazard as the servicer does not internalize the full costs and benefits of its actions. A homeowners mortgage may thus be transferred several times during its lifetime, even as the servicer remains the same (or the servicer may change while the ultimate ownership of the mortgage remains constant).
61. Id. at 32829. Id. at 41314. 62. 63. See CREDIT SUISSE EQUITY RESEARCH, supra note 9, at 28. 64. KOTHARI, supra note 59, at 57, 211. 65. Id. at 208. Id. at 21130 (discussing various credit-enhancement techniques). 66. 67. Id. at 209. Typically securitization originators retain the last-out equity tranche of the trust, which allows them to retain any excess spread over the trusts payout. Id. at 289. Id. at 209. 68.
LEVITIN - FINAL
4/23/2009 11:40 AM
586

II. BANKRUPTCY-MODIFICATION RISK AS REFLECTED IN MORTGAGE-MARKET PRICING
Because only certain types of residential mortgages are subject to the Bankruptcy Codes antimodification provision,69 it is possible to examine different types of mortgage-market pricing to see if they reflect differences in bankruptcy-modification risk. One would expect that if the market were sensitive to bankruptcy modification, there would be a risk premium for mortgages on the types of property that can currently be modified in bankruptcymortgages on vacation homes, multifamily homes, and investment propertiesand that this premium would not exist for single-family, owner-occupied, principalresidence mortgages, which cannot be modified. To test this hypothesis, I examined three different pricing measures in mortgage markets: effective mortgage interest rates (annual percentage rates or APRs), PMI rates, and secondary-mortgage-market pricing from Fannie Mae and Freddie Mac. In each market I examined rate variation by property type in order to isolate the expected risk premium for bankruptcy-modification risk on non-single-family, owneroccupied properties. Astonishingly, all three measures indicate that mortgage markets are indifferent to bankruptcy-modification risk, at least in terms of pricing.70
A. Mortgage-Interest-Rate Variation by Property Type

1. EXPERIMENT DESIGN Using online rate-quote generators, I pricing on six types of properties to see variations in bankruptcy-modification risk:71 family, principal residences; single-family tested current mortgage if the pricing reflected owner-occupied, singlesecond homes; owner-
69. See supra text accompanying notes 4346. 70. It is possible that there is simply less available credit for modifiable properties. I was unable to test this possibility, however. 71. The reliability of online quotes was confirmed in interviews with veteran mortgage brokers. E.g., Telephone interview with Dom Sutera, The Sutera Group (Jan. 23, 2008). In response to my congressional testimony based on a draft of this Article, David M. Kittle, the President of the Mortgage Bankers Association, declared that online rate quotes were unreliable. Helping Families Save Their Homes: The Role of Bankruptcy Law, Hearing Before the S. Judiciary Comm., 110th Cong. (1998) (testimony of David Kittle). Mr. Kittles surprising admission of mortgage industry bait-and-switch tactics aside, the rate-quote generators results are consistent with other, more reliable data sources, see infra Sections II.B., C, as well as with quotes produced by the Mortgage Bankers Association in Mr. Kittles own testimony.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
587
occupied, two-family residences; owner-occupied, three-family residences; owner-occupied, four-family residences; and investor properties. I obtained the quotes from four major mortgage lenders: eLoan, IndyMac, Chase, and Wachovia.72 These lenders were selected because their online quote generators did not require disclosure of personal information. The quotes were generated between January 17, 2008 and January 27, 2008.73 Using the online quote generators, I tested 530 mortgage rate quotes from eleven states. The quotes were divided into two subsamples. First I took a standardized sampling of 288 quotes in three states: California, Massachusetts, and Pennsylvania. I chose Massachusetts and Pennsylvania because of the clarity of the law in those states, which are located in the jurisdictions of the United States Courts of Appeals for the First and Third Circuits, respectively. There is unambiguous circuit-level law in both the First and Third Circuits permitting the strip-down of mortgages on all multiunit residences.74 I included California both because it is the largest single state mortgage market and because it has been hit particularly hard by the mortgage crisis.75 For this three-state sample, I obtained 288 quotes for 30-year fixed-rate, first-lien purchase money mortgages, the most common traditional-mortgage product. I tested assuming a LTV ratio of 80 percent, representing a 20 percent down payment. Half of the quotes obtained were for loan amounts within the GSE-conforming limits, and half were for nonconforming jumbos.76 The conforming quotes were
72. Chases online rate-quote generators can be found at http://mortgage.cha se.com/pages/purchase/crq_p_landing.jsp. Since the time I gathered the rate quotes, IndyMac and Wachovia both failed, so their quote generators are no longer available, and eLoan has outsourced its rate quotes to Lending Tree. 73. IndyMac was placed in an FDIC conservatorship on July 11, 2008. Louise Story, Regulators Seize IndyMac After a Run on the Bank, N.Y. TIMES, July 12, 2008, at C5. 74. See, e.g., Scarborough v. Chase Manhattan Mortgage Corp. (In re Scarborough), 461 F.3d 406, 413 (3d Cir. 2006); Lomas Mortgage, Inc. v. Louis, 82 F.3d 1, 7 (1st Cir. 1996). 75. Kevin Yamamura, Schwarzenegger Vetoes Mortgage Broker Restrictions, SACRAMENTO BEE (Cal.), Sept. 26, 2009, at A3 (California is the epicenter of the national foreclosure crisis.) (quoting Paul Leonard, California Office Director, Center for Responsible Lending). 76. Federal law limits the size of loans that GSEs may purchase. 12 U.S.C. 1454(a)(2)(C) (2006). Loans above the conforming limit are known as jumbos. Bob Tedeschi, The Race Is to the Swift, N.Y. TIMES, Aug. 10, 2008. The limits are adjusted annually. 12 U.S.C. 1454(a)(2)(C). For 2007, the limit was $417,000. Press Release, Office of Federal Housing Enterprise Oversight, 2007 Conforming Loan Limit to Remain at $417,000 (Nov. 28, 2006), available at www.ofheo.gov/media/pdf/ PRConfLoan07.pdf; OFHEO.gov, Supervision & Regulations: Conforming Loan
LEVITIN - FINAL
4/23/2009 11:40 AM
588
for loan amounts based on the average mortgage loan amount in the state. The quotes for the jumbos were for loan amounts slightly higher than the conforming limit for a three-family residence.77 For each of the six types of residences, I recorded the quoted interest rate, points, and APR for the lowest APR quotation. For IndyMac and eLoan, I obtained a full set of quotes for each of three different credit scores: 760, 660, and 560, representing prime, Alt-A, and subprime borrowers, respectively.78 For Chase and Wachovia, I was not able to test for specific credit scores and have assumed that the single set of quotes generated are for prime borrowers, based on rate comparisons with IndyMac and eLoan.79 Accordingly, in each state I tested thirty-six quotes for IndyMac and eLoan, and twelve for Chase and Wachovia, for a total of 96 quotes per state and 288 quotes total. Appendix A provides an illustrative example of the data. Appendix A shows the rate quotes generated by IndyMac on January 27, 2008, for mortgages in California within conforming limits at 20 percent and 10 percent down (Tables A1A3 and A4A5, respectively) with variations by credit scores. As a cross-check on the ability to extrapolate from 30-year-fixed, first-lien, purchase-money-mortgage rate quotes in California, Massachusetts, and Pennsylvania, I also tested an additional
Limits, http://www.ofheo.gov/Regulations.aspx?Nav=128 (last visited Jan. 27, 2009). In 2007, the limit was temporarily raised for selected metropolitan statistical areas by different amounts. See OFFICE OF FEDERAL HOUSING ENTERPRISE OVERSIGHT, METROPOLITAN STATISTICAL AREAS, MICROPOLITAN STATISTICAL AREAS AND RURAL COUNTIES WITH NEW LOAN LIMITS (2008), available at http:// www.ofheo.gov/media/hpi/AREA_LIST_5_2008.pdf. As of July 30, 2008, OFHEO was replaced by the Federal Housing Finance Agency (FHFA). Housing and Economic Recovery Act of 2008, Pub. L. No. 110-289, 1101, 130104, 122 Stat. 2654. 77. By testing just above the conforming limit for three-family residences, all of the four-family-residence quotes ended up being for conforming properties because of the higher conforming loan limit for four-family residences. I tested just above the three-family limit out of concern that the loan amount necessary for a four-family jumbo might be so large as to distort the results for single- and two-family properties. Since there is no difference in legal treatment of three-family and four-family residences, I do not believe that the absence of four-family jumbos from our sampling is significant. 78. There is no standard definition of subprime credit score; cutoffs of 600, 620, or even 650 are commonly used. There is also no standard definition of Alt-A; it can range between 620 to 800+. See Jody Shenn, Subprime Meltdown Snares Borrowers with Better Credit, BLOOMBERG.COM, Mar. 22, 2009, http://www.bloomberg.com/apps/news?pid=20601087&refer=home&sid=a8ilcv.eOx Mc. 79. Chase permits specification of credit by characterization (excellent, good, fair, etc.), but not by score. See Mortgage.chase.com, Check Rates, http://mortgage.chase.com/pages/refinance/crq_r_landing.jsp (last visited Mar. 25, 2009). I used excellent as the assumption.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
589
nonscientific sample of 242 quotes from those three states, as well as from eight additional states: Illinois, Florida, Maryland, Michigan, Missouri, Ohio, Nevada, and Texas. In this sample I tested at a variety of credit scores, ranging from 540 to 760, a range of LTV ratios from 90 percent to 70 percent, a variety of property values, and other mortgage products such as 15-year-fixed mortgages, 2/1 and 5/1 LIBOR adjustable rate mortgages (ARMs), and interest-only mortgages. 2. EXPERIMENT RESULTS The samplings produced three general rate-quote patterns that did not vary by either state or mortgage product type. First, for all conforming mortgage loans with 20 percent down payments from eLoan, IndyMac, and Wachovia, there was no difference in credit scores between the quotes offered for single-family primary residences, vacation homes, or any multifamily unit in which one unit is owner occupied.80 Interest rates, points, and APRs were identical for these property types, despite the variation in bankruptcy-modification risk. Uniformly, however, investor properties had higher interest rates and points. The rate-spread patterns in the quotes make sense intuitively. Investor properties share the same bankruptcy-modification risk as vacation homes and multifamily residences. Therefore, the rate premium on investor-property mortgages cannot be attributed to bankruptcy-modification risk because there is not also a corresponding rate premium for vacation homes or multifamily homes. Instead, to the extent that investor-property mortgages have a premium over mortgages on single-family, owner-occupied properties, it reflects risks distinct from bankruptcy modification. It is unsurprising that vacation homes have the same rates as single-family principal residences. Vacation homes reputedly have lower default rates than principal residences because typically only well-heeled buyers purchase them. They do not have tenant risks such as vacancy, nonpayment, or damage, and they are typically well maintained because of the pride-of-ownership factor. Arguably neither vacation-home nor investor-property mortgages reflect a premium for bankruptcy modification because neither is likely
80. Chase rate quotes for conforming 20-percent-down mortgages presented a variation on this pattern. Single-family principal residences, vacation homes, and fourfamily residences had identical quotes, but two- and three-family residences were priced around twenty-five basis points higher, and 30-year-fixed quotes were unavailable for investor properties.
LEVITIN - FINAL
4/23/2009 11:40 AM
590
to be modified in bankruptcy. A mortgage-loan modification in bankruptcy can occur only as part of a plan.81 The automatic stay would likely be lifted on an investment property (or vacation home) before a plan could be confirmed. The Bankruptcy Code provides that the automatic stay shall be lifted for cause, including either lack of adequate protection of a secured creditors interest in the propertythat is, payments to compensate the secured creditor for depreciation in its collateral during the bankruptcyor the debtors lack of equity in the property, if the property is not necessary for an effective reorganization.82 Therefore, debtors with positive equity who could not handle mortgage payments would be unlikely to be able to make the adequate-protection payments necessary to prevent the lifting of the stay,83 and debtors with negative equity would find the stay lifted because investment properties and second homes are not essential to their reorganizations.84 Thus, investor-property and vacation-home mortgages do not provide particularly meaningful comparisons to single-family, owneroccupied properties when considering bankruptcy-modification risk because both are unlikely to be modified. In contrast, multifamily residences in which the debtor occupies a unit are essential to a debtors successful reorganization, and therefore present a valid comparison to single-family, owner-occupied properties. Multifamily residences in which the owner resides carry the same tenant risks as investor properties. I do not have default-rate data on multifamily residences, but owner residency likely reduces default risk and ensures reasonable property maintenance. Significantly, for all four lenders, single-family primary residences, which are not modifiable in bankruptcy, were priced the same as both vacation homes and at least one of the multifamily residence types, property types that are modifiable in bankruptcy. The expected rate-premium differential among property types does not exist.85 When I reduced the down payment to 10 percent on conforming mortgages, a slightly different pattern emerged. First, rate quotes were
81. 11 U.S.C. 1321, 1322(b). 82. Id. 362(d). See id. 362(d)(1). 83. 84. See id. 362(d)(2). 85. In theory, the lack of the differential could be explained by the market anticipating bankruptcy-reform legislation, which would impose the same basic modification risk on all properties, even if the risk would actually be less for investor properties and vacation homes because of the greater likelihood that the stay would be lifted on those properties.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
591
not always available with subprime credit scores (560 and 540). Second, for prime and Alt-A credit scores, there were four tiers of pricing by property type. Single-family principal residences and twofamily, owner-occupied properties were priced identically. Vacation homes also had the same interest rates and points, but APRs were about ten basis points higher because of additional PMI premiums.86 Investor properties and three- and four-family, owner-occupied residences had significantly higher APRs (around 150 and 250 basis points, respectively).87 This makes sense because three- and four-family properties have significant tenant risk, much like investor properties. When I tested jumbos, Wachovia followed its price pattern for conforming loans at 80 percent LTV, and did not differentiate among property types except for investor properties. At 90 percent LTV, investor-property quotes were unavailable, and the interest rate and points were the same for all other property types. The APR, however, was lower for single-family properties at a 90 percent LTV ratio because of the higher closing costs for the other property types due to items such as higher appraisal fees. For Chase, jumbo quotes were only available at 80 percent LTV. For single-family principal residences, the quotes were identical to those for two-family residences. Vacation homes were quoted slightly higher, and three- and four-family and investor-property quotes were unavailable from Chase for jumbos. IndyMac and eLoan had a different pattern for jumbos. First, quotes were simply unavailable for subprime credit scores with 10percent or 20-percent down payments, and for some Alt-A products. I was only able to generate quotes when I significantly increased down payments. Second, a three-tier rate spread emerged for prime borrowers depending on property type. Single-family principal residences were priced the lowest. Vacation homes and two-family properties were priced with slightly higher interest rates, but lower points, and APRs (the unit price) that were approximately eight to twelve basis points higher.88 Finally I was unable to obtain rate quotes for jumbo mortgages on three- or four-family properties or investor
86. A basis point is 1/100th of a percent (0.01 percent). 87. Again, Chase rate quotes were different. At 10 percent down, rate quotes were still unavailable for investor properties for 30-year-fixed mortgages. Rates for vacation-home mortgages were actually slightly lower (five basis points) than for single-family principal residences. Notably, interest rates and points for two-family residences were the same as for single-family principal residences, but APRs were higher by thirty-eight basis points. I was unable to ascertain the source of the APR variation. 88. On a $500,000, 30-year, 6-percent fixed mortgage, this translates into an additional $25.97$38.89 per month.
LEVITIN - FINAL
4/23/2009 11:40 AM
592
properties with 20 percent down or less. As with the subprime and AltA mortgages, I was able to get quotes when I decreased the LTV ratio. The major insight from these rate quotes is that current mortgage rates (in January 2008) evinced a marked indifference to bankruptcymodification risk, at least among conforming loans. Regardless of the LTV ratio, there was no difference among conforming loans between the rates for single-family, owner-occupied properties, which may not currently be modified in bankruptcy, and those for two-family, owneroccupied properties, which may currently be modified. This means that the rate differences that emerge at 90 percent LTV ratios between oneand two-family, owner-occupied residences and other property types are not attributable to bankruptcy-modification risk. For both conforming and jumbo products, the higher interest rates for three- and four-family properties and investor properties are a function of risks other than bankruptcy modification. Mortgages on three- and four-family residences may carry higher prices at low LTV ratios because of higher default rates, given the difficulties in managing income-producing properties for amateur landlords and the extremely limited foreclosure-sale market for these properties outside of a few urban areas.89 Likewise, the higher interest rates and points required on investor properties at all LTV ratios are explained by higher default rates on investor properties, the greater likelihood of investor properties being nonrecourse, and the more limited secondary market for investorproperty mortgages. Investor properties have inherently greater default risk in part because an investor has the additional rent or mortgage expense that an owner-occupier does not. Investor properties also carry a variety of tenant risks: vacancy, nonpayment, and damage. Because investor-property mortgages are often financed through rental payments, tenant risk adds to the default risk. GSE-conforming mortgages have the same bankruptcymodification risk as jumbos. Therefore, it seems unlikely that the small difference in the APR between single- and two-family, owner-occupied properties for some lenders jumbo mortgages relate to bankruptcymodification risk. I suspect it is a function of the significantly smaller secondary market for jumbos, particularly for two-family, owneroccupied properties. While there is variation in rate-quote patterns among the four lenders surveyed, all four lenders provided identical quotes for singlefamily, owner-occupied properties, which cannot be modified in bankruptcy, and for certain types of multifamily properties, all of
89. regard. Chases outlier pricing for four-family conforming loans is puzzling in this
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
593
which can be modified in bankruptcy. This indicates that current mortgage-pricing variations by property type do not reflect bankruptcymodification risk, but other risk factors. Current mortgage-origination rates indicate that mortgage-lending markets are indifferent to bankruptcy-modification risk, a conclusion confirmed by PMI pricing.
B. Private-Mortgage-Insurance Rate Premiums

Another way to examine mortgage-market sensitivity to bankruptcy modification is through PMI premiums. PMI is generally required for all mortgages on which there is less than a 20 percent down payment.90 The borrower pays the PMI premiums, but the lender is the insurance payee. Private mortgage insurers stand in the mortgage lenders shoes and assume all the risks that the mortgage lender bears, with three exceptions: PMI policies typically exclude coverage for bankruptcy modification (including strip-down), fraud, and special hazards, such as earthquakes and floods.91
90. Fannie Mae and Freddie Mac, the GSEs that dominate the secondary mortgage market, will not purchase standard mortgages with LTVs of more than 80 percent without mortgage-insurance coverage. See FANNIE MAE, supra note 57, pt. V, 101.01; FREDDIE MAC, supra 57, 27.1. As a result, most mortgage originators require some form of mortgage-insurance coverage, typically from a private mortgage insurer, in order to access the full secondary market. See FREDDIE MAC, MORTGAGE INSURANCE COVERAGE OPTIONS MATRIX (2008), available at http://www.freddiemac.com/learn/pdfs/us/flexmi.pdf. Lender PMI-coverage requirements are required to terminate when the LTV ratio reaches 78 percent. 12 U.S.C. 4901(18), 4902(b) (2006). For a detailed examination of private mortgage insurance, see Quintin Johnstone, Private Mortgage Insurance, 39 WAKE FOREST L. REV. 783 (2004). 91. ANDREW LIPTON & SHIV RAO, MOODYS INVESTOR SERV., VALUING LENDER-PAID MORTGAGE INSURANCE IN MBS AND ABS TRANSACTIONS 5 (Feb. 9, 2001), available at http://www.natlaw.com/seminar/doc34.pdf. Notably, Radian Guaranty Co. does not exclude bankruptcy losses from its coverage and specifically covers losses from strip-down. RADIAN GUARANTY INC., MASTER POLICY 16 (2006), available at http://www.radian.biz/pdf/master_policy.pdf. Professor Scarberry has observed that lenders are uniquely vulnerable because of the PMI exclusion. The Looming Foreclosure Crisis: How To Help Families Save Their Homes, 110th Cong. (Dec. 5, 2007) (Statement of Mark S. Scarberry), available http://judiciary.senate.gov/hearings/testimony.cfm?id=3046&wit_id=6808. Some private mortgage insurers, however, do not exclude bankruptcy strip-down from their master policies. See, e.g., RADIAN GUARANTY INC., supra, at 16; STATE OF NEW YORK MORTGAGE AGENCY, PRIMARY MORTGAGE INSURANCE MASTER POLICY 24 (2003), available at http://nyhomes.org/docs/pmigenc3pol.pdf. Thus lack of PMI coverage for strip-down from major private mortgage insurers seems to be attributable to lack of lender demand, as indicated in the lenders own pricing. Even when there is an exclusion for bankruptcy strip-down, however, the exclusion applies not just to currently nonmodifiable mortgages, but to all types of
at
LEVITIN - FINAL
4/23/2009 11:40 AM
594
Private mortgage insurers are in the business of pricing for risk, so the slight coverage difference between the risks covered by PMI policies and the risks incurred by mortgage lenders creates a natural experiment for testing the sensitivity of mortgage originators to bankruptcy modification using nontemporal differences in differences analysis. We can examine the differences in the spread between PMI rates for certain property types and the spread in mortgage interest rates between the same property types. In other words, by subtracting the additional PMI premium on two-, three-, or four-family properties, vacation homes, or investor properties relative to single-family, owneroccupied properties from the additional interest rate charged on two-, three-, or four-family properties, vacation homes, or relative to singlefamily, owner-occupied properties, we can isolate the amount of the additional interest rate in mortgage-origination pricing that covers the PMI exclusions of bankruptcy modification, special hazard, and fraud. To the extent that there is no difference, it indicates that the origination market, which bears the risk of bankruptcy modification, strip-down, and fraud, does not price for these risks. Table 1 shows major private mortgage insurers current premium adjustments for various property types above the premium for singlefamily principal residences. The seven companies listed issue substantially all of the PMI in the United States.92 The premiums vary from insurer to insurer, but a couple of points are notable. First, the additional PMI premium for investor properties is typically thirty-eight basis points. This is exactly the average additional premium mortgage originators charge for investor-property mortgages above single-family, principal-residence mortgages.93 Second, the additional PMI premium charged for vacation homes is fourteen basis points. I do not have national-average figures for the additional interest-rate premiums for vacation-home mortgages. Often
mortgages, and a lender can never be sure that what is an owner-occupied principal residence at the time a mortgage loan is made will be so in the future (and PMI coverage always excludes fraud); it is as easy as moving in a tenant the day before filing for bankruptcy. See Lipton & Rao, supra, at 5. Thus, lenders have been assuming the risk of strip-down all along and not relying on PMI. Prospectively, though, if stripdown risk grows, it is reasonable to expect markets to adjust, as lenders will demand modification coverage from private mortgage insurers or find equivalent coverage through swap and derivative products. 92. Johnstone, supra note 90, at 789. 93. Straightening Out the Mortgage Mess: How Can We Protect Home Cong. 3 (2007) [hereinafter Kittle Testimony] (statement of David G. Kittle, ChairmanElect, Mortgage Bankers Association), available at http://judiciary.house.gov/ media/pdfs/Kittle 071030.pdf.
Ownership and Provide Relief to Consumers in Financial Distress? Part II: Before the Subcomm. on Commercial and Admin. Law of the H. Comm. on the Judiciary, 110th
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
595
there was no premium for vacation homes in the rate-quote sampling, but when there was, it was for high-LTV properties and was around twelve to fourteen basis points.94 The near perfect fit between additional PMI premiums and additional mortgage interest-rate premiums for investor properties and vacation homes indicates that origination markets are indifferent to bankruptcy-modification risk because PMI does not cover bankruptcy-modification risk, yet is priced identically. Table 1: Current Additional PMI Premiums Above Single-Family, Principal Residence by Property Type in Basis Points 95
2FAMILY AIG United Guaranty CMG Genworth MGIC Radian Guaranty Co. Republic Mortgage Ins. Triad Guaranty96 0 0 0 0 0 0 0 3-4 FAMILY 0 32 38 0 38 38 0 VACATION HOME 14 14 14 14 14 14 1450 INVESTOR PROPERTY 38 32 38 38 38 38 3875
94. The explanation for this additional premium is that most second-home purchasers put down at least 20 percent of the purchase price so they are not required to have PMI coverage. See supra note 90. Therefore, the additional PMI premium for second homes likely reflects the smaller (and riskier) coverage pool of second-home buyers who do not put down at least 20 percent of the purchase price. 95. AIG UNITED GUAR., RATES EFFECTIVE FEB. 1, 2007 (2007), available at https://www.ugcorp.com/rates/Monthly.pdf; CMG MORTGAGE INS. CO., MONTHLY PREMIUM RATES, NATIONWIDE EFFECTIVE 12.15.08 (2008), available at http:// www.cmgmi.com/ca-19.aspx; GENWORTH FIN., NATIONAL RATE PLANS (July 14, 2008), available at http://mortgageinsurance.genworth.com/pdfs/Rates/National Rates.pdf; MGIC, NATIONAL RATE CARD: JANUARY 2008 (2008), available at http:// www.mgic.com/pdfs/71-6704_Natl_rates_jan08.pdf; RADIAN GUAR. INC., INDUSTRY RATES, EFFECTIVE DATE: FEBRUARY 1, 2008 (2008), available at http:// www.radian.biz/pdf/RAR167IndustryBPMI_1_11_08.pdf; REPUBLIC MORTGAGE INS. CO., PREMIUM RATES (NOV. 24, 2008), available at http://www.rmic.com/ ratesguides/premiumrates/ratecards/Documents/PremiumRates-3.08.pdf; TRIAD GUAR. INS. CO., MONTHLY RATES: DECEMBER 2007 (2007), available at http://www.tgic.com/pdf/TGRC.0116.1207RefundableMonthly.pdf. Since the quotes were collected, Triad Guaranty ceased underwriting new policies. Triad Guaranty to Go into Run-Off, INTL BUS. TIMES, June 19, 2008, available at http://www.ibtimes.com/articles/20080619/triad-guaranty-to-go-into-run-off.htm. 96. The range of rates for vacation homes and investor properties is dependent on credit scores.
LEVITIN - FINAL
4/23/2009 11:40 AM
596
Of course, debtors often surrender investor properties and vacation homes in bankruptcy; these properties are not essential to their reorganizations. Therefore, modification risk may not be particularly significant for these types of properties, so one would not expect to see price sensitivity. Debtors are more reluctant to surrender their own residences, however. Therefore the data on two-, three-, and fourfamily residences is of more relevance. PMI insurers do not distinguish between three- and four-family residences. Of the seven major private mortgage insurers, four charged a thirty-eight-basis-point premium for three- and four-family residences, but three charged no premium. I do not have an explanation for the differences between insurers on three- and four-family rate quotes.97 Mortgage-origination rates for three- and four-family properties, however, are often priced exactly like single-family properties, but are sometimes priced significantly higher than thirtyeight basis points. It is hard to infer a bankruptcy-modification-risk premium in mortgage-origination pricing from three- and four-family residence PMI rates. What is significant, however, is that none of the PMI insurers charged more for policies on two-family residences than for singlefamily residences. This matches with the lack of differentiation in origination pricing between single-family and two-family residences, despite the different bankruptcy-modification risks. The additional PMI premiums for the types of properties that can currently be modified in bankruptcy are essentially the same as the additional interest-rate premiums for these property types. Because most PMI coverage excludes bankruptcy modification, it indicates that mortgage lenders, who bear the cost of bankruptcy modification, do not price for the modification risk.98
97. It is not clear how competitive the PMI industry is on the consumer level. If PMI insurers have quasicaptive relationships with lenders or mortgage brokers, it would provide an explanation for the variation in rate quotes for three- and four-family properties. 98. Notable too, Radian Guaranty Inc., the sole insurer that I have been able to verify, does not exclude bankruptcy losses (including strip-down) from its coverage, and does not price two-family properties differently than single-family properties. See RADIAN GUAR. INC., supra note 91, at Condition 11, B.1(c) (2008), available at http://www.radian.biz/pdf/Master_Policy_2008.pdf. Securitized mortgage pools often have pool-level, stop-loss bond insurance to enhance their credit ratings. See KOTHARI, supra note 59, at 21130. The existence of pool-level insurance may weaken the conclusions that can be drawn from PMI pricing. Nevertheless, not all mortgages are securitized and there can also be a presecuritization exposure period for those that are.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
597
C. Secondary-Market-Pricing Variation by Property Type

The indifference of the market to bankruptcy-modification risk in mortgage pricing is also apparent from the delivery fees charged by Freddie Mac and Fannie Mae, GSEs that are the two largest purchasers of home mortgages on the secondary market.99 Freddie Mac and Fannie Mae charge a delivery fee, essentially a discount rate, on the mortgages they purchase from originators.100 The discount rate varies by the characteristics of the mortgage product, such as property type, LTV ratio, and the borrowers credit score.101 Notably, Freddie and Fannie have additional discount fees for investor properties and some multifamily residences, but not for vacation homes or for certain multi-family-residence configurations.102 The absence of a risk premium on all properties that can currently be modified in bankruptcy indicates that Freddie and Fannie are not pricing for bankruptcy-modification risk. This evidence conforms to the pricing in the mortgage-origination market. Given that a significant percentage of mortgage originations are sold into a secondary market103 and that GSEs are the largest players in the secondary market, Freddies and Fannies pricing shapes mortgage-origination pricing, so it is not surprising to see parallel pricing indifference. All current observational evidence indicates that the mortgage-lending market is indifferent to bankruptcy-modification risk.104 To be sure, the current mortgage-market data only show price insensitivity to bankruptcy-modification risk. The data do not tell us
99. David Reiss, Subprime Standardization: How Rating Agencies Allow Predatory Lending to Flourish in the Secondary Mortgage Market, 33 FLA. ST. U.L. REV. 985, 1055 (2006). 100. See, e.g., FREDDIE MAC, BULLETIN NO. 2009-5, at exh. 19 (Mar. 9, 2008), available at http://www.freddiemac.com/singlefamily/pdf/ex19.pdf. 101. Id. 102. FANNIE MAE, LOAN-LEVEL PRICE ADJUSTMENT (LLPA) MATRIX AND ADVERSE MARKET DELIVERY CHARGE (AMDC) INFORMATION 34 (Dec. 29, 2008), available at www.efanniemae.com/sf/refmaterials/llpa/pdf/llpamatrix.pdf; FREDDIE MAC, NOVEMBER 24, 2008 BULLETIN, at E19-9 to E19-10 (2008), available at http://www.freddiemac.com/singlefamily/pdf/ex19.pdf. 103. CREDIT SUISSE EQUITY RESEARCH, supra note 9, at 28 (estimating that 75 percent of outstanding first-lien residential mortgages are held by securitization trusts, and that two-thirds are in GSE MBS). Freddie and Fannie MBS alone comprise over 44 percent of residential first-lien mortgage debt outstanding. Id. 104. Likewise, there has been no problem securitizing mortgage debts that are modifiable, such as family-farm mortgages and vacation-home, multiunit, and investor properties. Indeed, the largest securitization market is in bankruptcy-modifiable, nonmortgage debts, such as credit cards and car loans. See Fed. Reserve Bank, Statistical Release, No. G.19, Consumer Credit (Jan. 8, 2009), available at http://www.federalreserve.gov/releases/g19/Current/g19.pdf.
LEVITIN - FINAL
4/23/2009 11:40 AM
598
whether mortgage markets reflect bankruptcy modification in other ways, such as by limiting credit availability or requiring higher LTV ratios rather than by raising the price. It is highly unlikely that contemporary mortgage markets would reflect variations in bankruptcymodification risk solely in ways other than price, but the available data do not permit us to rule out this possibility. Historical data, however, provide some insight.
D. Historical Impact of Permitting Strip-Down

In addition to the current data on mortgage markets, there is historical data on the impact of permitting strip-down on the mortgage market. After the effective date of the Bankruptcy Code in October, 1979, but prior to the Supreme Courts decision in Nobelman v. American Savings Bank105 in 1993, federal judicial districts varied as to whether they permitted strip-down of mortgages on debtors principal residences.106 This variation between districts in the timing and results of their decisions creates a natural experiment that permits the isolation of the effects of allowing strip-down on the home-mortgage market. In a companion article coauthored with Joshua Goodman,107 I undertook a detailed regression analysis of the impact of strip-down using data from the Federal Home Finance Boards Monthly Interest Rate Survey. We examined five outcome variables: LTV ratio, interest rates, loan size, loan volume, and bankruptcy-filing rates. For LTV ratios, our analysis found a statistically significant impact, especially for the highest interest rate (i.e., the most risky) loans. For these loans, permitting strip-down resulted in a 1.373 percent reduction in LTV ratios.108 The reduction increased to 2.794 percent when a six-month time lag from ruling dates was introduced.109 For the least risky loans, however, no statistically significant impact was observed.110 These results are exactly what one would expect lenders are less willing to take big risks on the (presumably) riskiest borrowers, who have to pay the highest interest rates.
105. 508 U.S. 324 (1993). 106. E.g., compare In re Nobelman, 968 F.2d 483 (5th Cir. 1992), with In re Bellamy, 962 F.2d 176, 180 (2d Cir. 1992), and In re Hougland, 886 F.2d 1182, 1184 (9th Cir. 1989). 107. See Joshua Goodman & Adam J. Levitin, Mortgage Market Cramdown Sensitivity (2008) (unpublished manuscript on file with author). 108. Id. 109. Id. 110. Id.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
599
When we substituted interest rates as the outcome variable, we observed a similar pattern. We found a statistically significant increase of twelve to sixteen basis points from permitting strip-down for the median borrower.111 For low-risk borrowers (those with interest rates at the lowest twentieth percentile), we detected no statistically significant impact on interest rates, but for high-risk borrowers (those at the eightieth percentile in terms of rates) there was a marginally statistically significant impact of thirty-two to sixty-two basis points.112 The statistically significant LTV-ratio and interest-rate outcomes validate our empirical model, which makes it possible for us to draw conclusions from our inability to reject the null hypothesis when using other outcome variables. We did not detect any statistically significant impact on the availability of mortgage credit or on the number of bankruptcy filings.113 Because of the statistically significant outcome for LTV ratios and interest rates, we believe that the lack of a statistically significant outcome for other outcome variables can be read to show that there is little or no effect on these outcome variables from permitting strip-down. Taken together, the historical data and current market-pricing data indicate that mortgage markets are largely indifferent to bankruptcymodification outcomes. The current market data suggest almost complete indifference, whereas the historical data (from a time when far fewer mortgages were securitized) show some sensitivity, particularly for higher-price and higher-LTV (i.e., riskier) borrowers. These findings indicate that permitting strip-down or other forms of modification for all mortgages would be unlikely to have anything more than a small impact on interest rates or on mortgage-credit availability. The impact, if any, would be primarily on marginal borrowers, which might be a good thing because, prospectively, it would help discourage the aggressive lending (such as nodocumentation and low-documentation loans, and high LTV ratios) and irresponsible borrowing (such as borrowing based on an assumption of refinancing before teaser rates expired) that is at the root of the current mortgage crisis.
111. 112. 113.
Id. Id. Id.
LEVITIN - FINAL
4/23/2009 11:40 AM
600

III. EXPLAINING MORTGAGE-MARKET INDIFFERENCE TO BANKRUPTCY STRIP-DOWN
Basic economic theory posits that lenders will charge more when faced with larger potential losses. Yet, scholars studies on the impact of variation in debt-collection laws, including bankruptcy, show that consumer-credit markets are sensitive to some changes, but not to others. Economist Karen Pence has shown that mortgage-credit availability is sensitive to whether nonjudicial foreclosure is available.114 Likewise, Mark Meador has found increased interest rates in states that do not permit deficiency judgments,115 and economist Claudia Wood has shown that states with defaulter-friendly laws have higher interest rates.116 Professors Reint Gropp, John Karl Scholz, and Michelle White have found that creditors charge more on car loans in states that exempt significant property from creditor attachment.117 Professors Jeremy Berkowitz and Rich Hynes have estimated that increases in state personal-property exemptions lead to an increase in mortgage rates.118 Economist Lawrence D. Jones has found that Canadian lenders increased down-payment requirements in response to increased costs imposed by Canadian deficiency laws.119 Economist Robert N. Collender has extrapolated an increase in farm-loan interest rates as the result of the enactment of Chapter 12 of the Bankruptcy Code, which offers fewer creditor protections than Chapter 11.120 Professor Lee J. Alston has shown that there were higher interest rates and a decreased supply of mortgage credit in states that prohibited
114. Karen M. Pence, Foreclosing on Opportunity: State Laws and Mortgage 88 REV. OF ECON. & STAT. 177, 180 (2006), available at www.mitpresjournal.org/doi/pdf/10.1162/rest.2006.88.1.177?cookieSet=1. Nonjudicial foreclosure is generally faster than judicial foreclosure. See infra text accompanying notes 13233. To the extent that speed is a proxy for costs for a lender, Pences article shows that mortgagees prefer faster (and therefore cheaper) lossmitigation procedures. 115. Mark Meador, The Effects of Mortgage Laws on Home Mortgage Rates, 34 J. ECON. & BUS. 143, 146 (1982) (estimating a 13.87 basis-point increase in interest rates on new homes as a result of antideficiency laws). 116. Claudia Wood, The Impact of Mortgage Foreclosure Laws on Secondary Market Loan Losses (1997) (unpublished PhD dissertation, Cornell University, on file with author). 117. Reint Gropp et al., Personal Bankruptcy and Credit Supply and Demand, 112 Q. J. ECON. 217, 245 (1997). 118. Jeremy Berkowitz & Richard Hynes, Bankruptcy Exemptions and the Market for Mortgage Loans, 42 J.L. & ECON. 809, 82526 (1999). 119. Lawrence D. Jones, Deficiency Judgments and the Exercise of the Default Option in Home Mortgage Loans, 36 J.L. & ECON. 115, 12627 (1993). 120. Robert N. Collender, An Estimate of the Efficiency Effects of Chapter 12 Bankruptcy, 53 AGRIC. FIN. REV. 65, 78 (1993).
Credit,
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
601
foreclosure during the Great Depression.121 Economist Dragos Ailoae has detected a greater drop in auto-loan rates in states with high exemption levels as a result of the 2005 Bankruptcy Abuse Prevention and Consumer Protection Act (BAPCPA) amendments to the Bankruptcy Code,122 which restricted strip-down for purchase money loans secured by automobiles used primarily for personal use and made within two-and-a-half years before the bankruptcy filing.123 And Professor Nadia Massoud and her coauthors have shown that creditcard late fees and overlimit fees correlate with aggregate loss levels to banks.124 In contrast, economist Frederick Link has found that the cost and availability of general unsecured credit in states with high exemption levels was statistically indistinguishable from those in states with low exemption levels.125 Professors Berkowitz and Hynes have estimated that an increase in state homestead exemptions, unlike personalproperty exemptions, correlates with a decrease in interest rates.126 Attorney Michael Simkovic has shown that credit-card interest rates did not decline following the 2005 amendments to the Bankruptcy Code that made it harder for debtors to file for bankruptcy and discharge credit-card debt in bankruptcy.127 And author Mark Kantrowitz has found that making private student loans nondischargeable in bankruptcy has had only a de minimis impact on their availability.128 This Articles examination of current and historical pricing shows that permitting strip-down and other modifications would likely have little or no effect on mortgage pricing, and only a small impact
121. Lee J. Alston, Farm Foreclosure Moratorium Legislation: A Lesson from the Past, 74 AM. ECON. REV. 445, 456 (1984). 122. Pub. L. No. 109-8, 119 Stat. 23 (2005). 123. Dragos Ailoae, The Auto Loan Market Post BAPCPA 1 (May 2008) (unpublished manuscript, on file with author) (noting a fifteen basis-point drop in highexemption states above low-exemption states). When strip-down is permitted, a secured lender is more likely to end up with an unsecured deficiency claim, which will be paid only after the debtors property exemptions are taken. 11 U.S.C. 522(c)(2) (2006); see also id. 1325(a) (limiting strip-down for automobile loans). 124. Nadia Massoud et al., The Cost of Being Late: The Case of Credit Card Penalty Fees (2007) (unpublished AFA 2007 Chicago Meetings Paper), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=890826. 125. Frederick Link, The Economics of Personal Bankruptcy 13950 (2004) (unpublished PhD dissertation, Mass. Inst. Tech.), available at http://dspace.mit.edu/handle/1721.1/29428. 126. Berkowitz & Hynes, supra note 118, at 826. 127. Michael Simkovic, The Effect of BAPCPA on Credit Card Industry Profits and Prices, 83 AM. BANKR. L.J. 22, 2223 (2009). 128. MARK KANTROWITZ, IMPACT OF THE BANKRUPTCY EXCEPTION FOR PRIVATE STUDENT LOANS ON PRIVATE STUDENT LOAN AVAILABILITY 5 (2007), available at http://www.finaid.org/educators/20070814pslFICOdistribution.pdf.
LEVITIN - FINAL
4/23/2009 11:40 AM
602
(historically) on the size of down payments required for the highest-risk borrowers. To the extent that mortgage markets are sensitive to bankruptcy-modification risk, it is, however, reflected in pricing and credit availability to the most marginal borrowers. So why are mortgage markets not more sensitive to bankruptcymodification risk? The key to understanding current mortgage markets apparent indifference to bankruptcy-modification risk (and historic mortgage markets relative indifference) is that bankruptcy modification of mortgages does not exist in a vacuum. The relevant trade-off for a lender is not between no loss whatsoever and losses due to bankruptcy modification, but between losses due to modification in bankruptcy and losses due to foreclosure. Basic price theory of demand economics says that the mortgage market will respond to this trade-off by pricing against the outcome resulting in smaller losses. So which loss will be smaller? This is a question that can be addressed both by statute and by empirical evidence. As a statutory matter, one would expect bankruptcy-modification losses to generally be less than foreclosure-sale losses. By definition, a lender should do as well in bankruptcy as in foreclosure. Bankruptcy law provides that a secured lender must receive at least what the lender would receive in foreclosure, namely the value of the collateral.129 At worst, then, bankruptcy only imposes a de minimis time delay on the lender, which may itself be helpful, depending on the housing market. The adequate protection provisions of the Bankruptcy Code ensure that the mortgagee is protected against declines in the houses value.130 Thus, as a function of statute, bankruptcy modification will result in a lender receiving at least as much as in foreclosure, and often more. Of course, how mortgagees would fare in bankruptcy would depend heavily on property valuations if their collateral properties were underwater and subject to strip-down. So how does bankruptcy compare with foreclosure in terms of valuations? Unfortunately, it is impossible to do property-by-property comparisons, but we can
129. 11 U.S.C. 1325(a)(5) (2006). 130. Id. 361. Bankruptcy modification also allows the lender to avoid the expenses of foreclosure, particularly the expenses of maintaining the property as real estate owned by the lender and undertaking a private sale after the foreclosure, often involving a broker. Bankruptcy modification also provides lenders with insurance against interest-rate risk; if a lender forecloses, the lender will have to redeploy the capital it has recovered, just as if there were a prepayment If interest rates have fallen, the lender will be worse off, as it will have traded a higher-interest-rate investment for a lower-interest-rate one. Even for property types where modification is permitted, bankruptcy protects the lender against such a down side; the lender is guaranteed to get the risk-free rate plus an appropriate risk-premium. See Till v. SCS Credit Corp., 541 U.S. 465, 47980 (2004).
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
603
compare the average foreclosure valuation with the average bankruptcy valuation. While the pools of properties are different, there is no reason to think that the differences in the pools would be determinative of the valuations. The following Sections undertake an empirical analysis of whether foreclosure losses would outweigh bankruptcy-modification losses due to strip-down. First, the Article considers what sort of losses could be expected in foreclosure, then it turns to the losses that would hypothetically be incurred in strip-down.
A. The Baseline for Loss Comparison: Foreclosure Sales

The baseline for examining bankruptcy-modification losses is not zero losses, but the losses lenders incur in foreclosure.131 Foreclosure is widely recognized as an expensive process. Lenders incur legal costs in foreclosures, do not receive interest on defaulted properties (time-value loss), and often are forced to sell the property at a significant loss. Yet there is very little hard data on loss severities in foreclosure. The costs of foreclosure undoubtedly vary by lender and state, and have likely fluctuated over time. State foreclosure procedures can vary significantly. Some states require judicial foreclosurethe sale is supervised by the court and run by a sheriff or other public official.132 Other states permit nonjudicial foreclosure, whereby the foreclosing lender advertises and conducts the sale.133 Some states permit both procedures, and in any case, judicial and nonjudicial foreclosure procedures are far from uniform among states. In general, though, the availability of nonjudicial foreclosure is an important factor in cost, both because the greater speed of nonjudicial foreclosure reduces timevalue losses from nonperforming loans, and because of the lower transaction costs of foreclosing outside of the court system.134 Despite the paucity of hard, granular foreclosure data, most estimates put lender loss severities between 30 and 60 percent of the outstanding loan value,135 although it is not clear whether these numbers account only for losses of real cash outlays by lenders or also include notional losses like junk fees that are unrelated to costs and which simply pad outstanding balances due.
131. Cf. Berkowitz & Hynes, supra note 118, at 809 ([I]t is default and not necessarily bankruptcy that creates losses for creditors.). 132. See Pence, supra note 114, at 177. 133. Id. 134. Id. at 17778. 135. See id. at 177 (listing estimates).
LEVITIN - FINAL
4/23/2009 11:40 AM
604
One may get some sense of foreclosure-sale loss severities, however, from sheriffs foreclosure-sale data from Monmouth County, New Jersey.136 Monmouth County should not be understood as being in any way representative or typical, only that it is the most detailed publicly available data set on foreclosure-sale outcomes I have been able to obtain. New Jersey does not permit nonjudicial foreclosure,137 so the data on 845 completed foreclosure sales between July 2005 and December 2008 capture all completed foreclosure sales in Monmouth County during this period.138 The average foreclosure judgment was approximately $366,984.11.139 The average foreclosure sale netted only around $103,600.54, meaning there is an average 72 percent loss at the sale, not counting expenses, which are paid before the foreclosing lender is paid.140 Part of the reason for this is that there are often few bidders at foreclosure sales. In 50 percent of the sales in Monmouth County, there were no bids beyond the foreclosing plaintiffs opening $100 bid.141 Likewise, sheriffs sales data from sixteen Indiana counties show a similar pattern.142 From January 2006 to August 2008, there were 2,554 completed foreclosure sales in these counties, 266 of which were to third-party purchasers.143 When third-party purchasers purchased the
136. Sheriffguadagno.com, Monmouth County Foreclosures, http:// www.sheriffguadagno.com/Foreclosures.html (last visited Jan. 30, 2009). 137. N.J. STAT. ANN. 2A:50-19 (West 2009) (requiring the sale to be conducted by a sheriff). 138. Sheriffguadagno.com, supra note 136. I selected Monmouth County because of the high level of detail and easy accessibility of its sheriffs sale data via the Internet. 139. See id. 140. See id. 141. Id. Monmouth Countys requirement that the foreclosing creditor begin the auction with a $100 bid, see Sheriffguadagno.com, supra note 136, might inflate the overall loss number, as there is no reason for the foreclosing creditor to place a credit bid for the full amount of its claim unless another party shows up at the auction and starts placing bids. Otherwise, the sensible move for a foreclosing creditor is to bid the required $100 and maximize the size of the deficiency judgment. Unfortunately, the number of third-party auction winners in Monmouth County is so small as to frustrate more detailed analysis. The foreclosing plaintiff won the auction 66 percent of the time; only 6 percent of the winning bids were identified as third parties. Id. The remaining 28 percent were not identified in the data. Id. 142. Indiana data comes from SRI, Inc., which administers the sheriffs sales for Blackford, Boone, Crawford, Fayette, Kosciusko, Marshall, Monroe, Orange, Parke, Scott, Steuben, Tipton, Union, Vigo, Warrick, and Washington counties. See SRI-sheriffsale.com, SRI, Inc. Sheriffs Sale System, http://sri-sheriffsale.com/ SaleResults.aspx (last visited Jan. 31, 2009). 143. Id.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
605
foreclosed property in Indiana, the foreclosing lender recovered less than 67 percent of its claim from the sale.144 There are a few factors involved in the thin bidding at foreclosure sales.145 First, they are not well advertised compared with private sales.146 Second, the defaulted homeowner is still in possession of the property and inspection of the property is not possible.147 A foreclosuresale purchaser buys as is without an inspection or even knowing what the interior floor plan is.148 Third, defaulted properties are often not well maintained, which further pushes down foreclosure-sale prices.149 Fourth, any bidder at a foreclosure sale will likely have to bid over the outstanding mortgage amount in order to win because the foreclosing lender will place a credit bid for the outstanding amount of the mortgage.150 And fifth, the homeowner can, in some states, redeem the property after the foreclosure sale by simply paying the foreclosure-sale price.151 This means that sale results are not always final. As a result, the foreclosing plaintiff (i.e., the lender) was the purchaser at the foreclosure sale at least two-thirds of the time in Monmouth County (and possibly as high as 94 percent of the time), and no less than 88 percent of the time in the sixteen Indiana counties surveyed.152
144. Id. The mean and median sale prices were 32.99 percent and 34.36 percent of the claim, respectively, but the lenders recovery must be reduced for the costs of the sale. Id. This contrasts with 7.10 percent mean and 7.27 percent median pre-sale-expense losses when the foreclosing creditor won the foreclosure sale. Id. Indiana does not require a minimum bid, but a foreclosing lender may announce a minimum bid that signals its intent to credit bid up to a certain level. Because Indiana permits deficiency judgments, Arnold v. Melvin R. Hall, Inc., 496 N.E.2d 63, 64 (1986), a foreclosing lender might not want to credit bid for the full amount of the debt. Nevertheless, given the difficulties in collecting deficiency judgments, foreclosing creditors frequently credit bid the full amount of the debt in order to maximize their chances of winning the auction. A foreclosing lender is not able to inspect the property before sale; stepping foot on the property would constitute trespass. Therefore, lenders want to win the auction to be able to inspect the property and then sell it privately. By credit bidding to win the auction the lender is able to effectively purchase access to the property and information without risking new funds. It appears that lenders value this information more than they do a deficiency judgment against an impecunious borrower. Frequent credit bidding would explain the low level of losses when the foreclosing lender purchases the property. 145. See generally LYNN LOPUCKI & ELIZABETH WARREN, SECURED CREDIT, A SYSTEMS APPROACH 5568 (5th ed. 2006). 146. Id. at 6263. 147. Id. at 6364. 148. Id. 149. Id. at 67. 150. Id. at 7071. 151. Id. at 68. 152. See SRI-sheriffsale.com, supra note 142.
LEVITIN - FINAL
4/23/2009 11:40 AM
606
When the foreclosing plaintiff ends up owning a foreclosed property, it will usually be resold in a private sale. The reselling foreclosing lender is likely to get a depressed price because he is trying to sell an unoccupied house. The resale will mitigate the lenders loss, so a 72 percent loss rate, like in Monmouth County, might be too high for lenders net average losses, but it also does not account for lost time value; maintenance and improvement costs; sale costs, like advertising and brokers fees; and, for lenders with capital requirements (like banks and insurance companies), reduced lending capacity. These additional costs can be significant. In 2002, one analyst estimated the direct costs of foreclosure (excluding resale loss) at $58,759.153 All in all, Fitch Ratings is predicting loss severities of 58 percent for subprime loans originated in 2006, and 64 percent for those originated in 2007.154 While 2006 and 2007 subprime loans are not the entirety of the current foreclosure crisis, they are an important piece of it, and loss severities of 58 to 64 percent speak to the particular problems in the current foreclosure-sale market, as a glut of properties depresses sale prices both for foreclosure sales and for regular home sales. Depressed housing prices make negative equity more common, especially for the high-LTV-ratio mortgages that became common in recent years. And negative equity, combined with a mortgage that is a stretch for a homeowner, can lead to a strategic default and walk away, particularly if the mortgage is nonrecourse or the homeowner believes (rightly) that the lender will not pursue his or her other nonexempt assets.155 Strategic negative-equity defaults then fuel more foreclosures, and the process becomes a downward cycle in home prices and foreclosures.
B. Projecting Lender Losses in Bankruptcy

It is possible to predict lender losses from bankruptcy modification by examining actual bankruptcy filings to see what lender losses would be like if strip-down, the most drastic form of modification, were allowed. Strip-down losses can be projected from two proprietary databases. First, there is the 2001 Consumer Bankruptcy Project (CBP) Database. The 2001 CBP database is an extensive multidistrict database
153. 154.
See CRAIG FOCARDI, SERVICING DEFAULT MANAGEMENT: AN OVERVIEW

FOR
OF THE PROCESS AND UNDERLYING TECHNOLOGY
12 (2002). FITCH RATINGS, REVISED LOSS EXPECTATIONS SUBPRIME VINTAGE COLLATERAL 2 (2008) (on file with author). 155. See infra text accompanying notes 25465.
2006
AND
2007
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
607
collected during the 2001 minirecession.156 Among the data included in the 2001 CBP database is information from court records about property values and mortgage claims on debtors principal residences. Second, to account for the current severely depressed market conditions, I created an original dataset from Chapter 13 bankruptcy filings in two of the counties with the highest foreclosure rates in the nation, Riverside and San Bernardino counties in California, during the final four months of 2007. 1. 2001 CONSUMER BANKRUPTCY PROJECT DATABASE An analysis of the mortgage debts in the 2001 CBP database provides a sense of both how mortgage lenders would fare if modification were permitted and the likelihood that a mortgage would be modified in bankruptcy. Using the 2001 CBP data, one can estimate the impact on lenders of allowing bankruptcy strip-down on all mortgages. Strip-down is only one type of possible modification, but it is the most significant type because it affects the treatment of the principal of the mortgage claim as well as the interest. The Bankruptcy Code provides very different protections for secured and unsecured claims in Chapter 13. Secured claims are entitled to receive at least the value of their claims under a plan unless the debtor surrenders the property or the lender consents to alternative treatment.157 Unsecured Chapter 13 claims are entitled to receive only as much as they would have received in a Chapter 7 liquidation, which is usually nothing.158 Strip-down bifurcates a mortgage lenders bankruptcy claim into a secured claim for the value of the collateral and an unsecured claim for the deficiency.159 Because the unsecured claim is frequently of negligible value, strip-down typically has a larger effect on a mortgage lender than other types of modification, such as extending the term of the loan, changing its amortization schedule, or changing its interest rates.160 Indeed, the requirement that plans pay at least the present value of a secured-claim lender severely limits non-strip-down modifications, and the Supreme Court has set a floor for modified interest rates of
156. The 2001 CBP has data from the Central District of California, Northern District of Illinois, the Eastern District of Pennsylvania, the Middle District of Tennessee, and the Northern District of Texas. For a detailed description of the 2001 CBP and its methodology, see ELIZABETH WARREN & AMELIA WARREN TYAGI, THE TWO-INCOME TRAP 18182 (2003). 157. 11 U.S.C. 1325(a)(5) (2006). 158. Id. 1325(a)(4). 159. Id. 506(a). 160. See supra text accompanying notes 3942.
LEVITIN - FINAL
4/23/2009 11:40 AM
608
secured creditors in Chapter 13 of the prime rate, subject to various risk adjustments.161 This means that by examining historical data on potential strip-downs, we are examining the worst-case scenario for lenders. Only a small percentage of mortgages ever end up in bankruptcy. There is no data on the exact percentage, but if we use foreclosure rates as a guideline, it seems safe to estimate that less than 1 percent of all first-lien mortgages end up in bankruptcy. From 1993 to the third quarter of 2008, foreclosure rates have averaged 1.23 percent of all outstanding mortgages, with a low of 0.86 percent and a high of 2.97 percent.162 Many mortgage-delinquent homeowners never file for bankruptcy, although some do file before foreclosure proceedings commence. Thus, it seems reasonable to assume that, at least currently, a smaller percentage of mortgages end up in bankruptcy than end up in foreclosure. Using Monmouth County as our guide, 15 percent of homeowners whose houses were scheduled for sale filed for bankruptcy, whereas 52 percent of the scheduled sales were completed.163 The remainder of the cases involved settlements between the servicer or lender and the homeowner, including reinstatement of the mortgage or sales cancelled without indication why.164 If Monmouth County is in any way typical, it seems that even among seriously financially distressed homeowners, only a minority file for bankruptcy. Of the mortgages that end up in bankruptcy, many do not end up in Chapter 13. Extrapolating from the 2001 CBP database, 75 percent of the relatively small number of mortgages that end up in bankruptcy
161. Till v. SCS Credit Corp., 541 U.S. 465, 479 (2004). There is reason to believe that the prime rate would not be the relevant interest-rate benchmark for mortgages. Till dealt with a subprime auto loan with a 21 percent contract rate of interest. Id. at 470. As the prime rate has frequently been above the rate of 30-yearfixed mortgages, using the prime rate as a floor could result in an inequitable windfall for creditors. See Fed. Reserve Bank, Statistical Release, No. H.15, Selected Interest Rates (2009), available at http://federalreserve.gov/releases/h15/data.htm. Arguably for a mortgage loan, the appropriate baseline would be either the average 30-year-fixedmortgage rate or the 10-year-Treasury-bond rate. 162. MORTGAGE BANKERS ASSN, NATIONAL DELINQUENCY SURVEY (Feb. 2008), available at http://www.mbaa.org/ResearchandForecasts/ProductsandSurveys/ NationalDelinquencySurvey.htm. This average includes the recent historically high foreclosure rates. From 1993 to 2006, the average foreclosure rate was 1.12 percent, with a high of 1.51 percent. Id. 163. See Sherifguadagno.com, supra note 136. 164. See id. The Indiana data does not indicate the cause of cancelled foreclosure sales. New Jersey has an equity of redemption, but homeowners redeemed their homes in less than 2 percent of completed foreclose sales in Monmouth County.
See id.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
609
will end up in Chapter 13. In contrast, in the 2007 CBP, a more extensive multidistrict survey, 47 percent of debtors with mortgages filed for Chapter 13.165 I am cautious about extrapolating from these figures, however, because of the significant variation between them and the significant variation by district in the percentage of nonbusiness bankruptcy filings that are in Chapter 13.166 Within the limited universe of mortgages that end up in Chapter 13, the 2001 CBP is more instructive. The 2001 CBP database has information on 1,096 mortgages scheduled by debtors on Schedule D in their bankruptcy petitions in Chapter 13 cases.167 As Table 2 shows, of these claims, 29 percent were undersecured, around 4 percent were fully secured, and approximately 67 percent were oversecured. Table 3 shows that the undersecured claims were undersecured by an average of $13,172.23, but the median undersecured Chapter 13 claim was undersecured by only $8,000. On average, undersecured claims were undersecured by 21 percent of the total claim amount. I term this percentage the Security Ratio. A claim that is undersecured by 21 percent has a security ratio of -21 percent. In contrast, the average oversecured claim was oversecured by an equity cushion of $27,603.39, or 37 percent of the total claim amount, so it has a Security Ratio of 37 percent. Table 2: Mortgage Bankruptcy Chapter168
Undersecured Fully Secured Oversecured
Claim
Amount
to
Property
Value
Total 27% 4% 69%
by
Chapter 7 22% 4% 74%
Chapter 13 29% 4% 67%
165. 2007 Consumer Bankruptcy Project Database. The 2007 CBP Database is a proprietary database unavailable to the public. For a description of the 2007 CBP, see Robert M. Lawless et al., Did Bankruptcy Reform Fail? An Empirical Study of Consumer Debtors, 82 AM. BANKR. L.J. 349, 35356 (2008). 166. See, e.g., Gordon Bermant et al., Thoughts on the Local Legal Culture: The Case of Consumer Chapter Choice, 21 AM. BANKR. INST. J. 24, 24 (2002); Chrystin Ondersma, Are Debtors Rational Actors? An Experiment, 13 LEWIS & CLARK L. REV. 279, 295303 (2009); Teresa Sullivan et al., The Persistence of Local Legal Culture: Twenty Years of Evidence from the Federal Bankruptcy Courts, 17 HARV. J.L. & PUB. POLY 801, 828 (1994). 167. The amounts scheduled by debtors do not necessarily match up with those listed in creditors claims. See Katherine M. Porter, Misbehavior and Mistake in Bankruptcy Mortgage Claims, 87 TEX. L. REV. 121, 163 (2008) (Among all loans, the median claim exceeded its corresponding scheduled debt by $1,366. The average difference between a claim and its scheduled debt was $3,533.). 168. See supra note 156.
LEVITIN - FINAL
4/23/2009 11:40 AM
610
Table 3: Mortgages Claims in Chapter 13 (2001 CBP)169

Undersecured NUMBER 316 MARKET VALUE Mean $79,990.13 Std. Error $3,459.30 Median $72,000.00 CLAIM AMOUNT Mean $62,103.71 Std. Error $2,974.66 730 $112,681.40 $3,131.94 $91,750.00 $74,262.98 $2,112.03 50 $64,780.78 $4,267.22 $60,000.00 $59,722.70 $4,418.32 1,096 $102,818.86 $2,466.04 $87,000.00 $70,093.87 $1,668.42 Oversecured Fully Secured All
Median $48,126.00 $65,000.00 $57,000.00 $61,403.50 MARKET VALUE MINUS (CLAIM AMOUNT PLUS AMOUNT OF SENIOR LIENS) Mean -$13,172.23 $27,603.39 $0.00 $14,587.64 Std. Error $876.30 $1,606.16 $0.00 $1,233.87 Median -$8,000.00 SECURITY RATIO -21.21% $15,000.00 37.17% $0.00 0.00% $6,209.50 20.81%
Looking at all Chapter 13 mortgage claims, the average claim was oversecured by $14,587.64, which was around 21 percent of the average claim amount. Chart 3 shows the distribution of the amount by which mortgages in the 2001 CBP database were over- or undersecured. The impression from these figures is that most mortgages in bankruptcy are over- or undersecured by a relatively small amount, and that mortgages are far more frequently oversecured than undersecured.
169.
2001 Consumer Bankruptcy Project Database.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
611
$400,000 $350,000 $300,000 $250,000 $200,000 $150,000 $100,000 $50,000 $0 -$50,000 -$100,000 0 200 400 600 800 1000 1200
Chart 3: Chapter 13 Mortgages Ranked by Amount Oversecured or Undersecured (2001 CBP)170 The 2001 CBP data indicate, then, that under an unlimited stripdown regime, in normal market conditions only a limited subset of the already highly limited universe of mortgages that end up in Chapter 13 bankruptcy would be subject to strip-down, and in those cases the average lender losses would be limited to $13,172.23, with a disproportionate share borne by second-lien mortgages. These are losses a lender would incur in a foreclosure in addition to the costs of the foreclosure process and the likely lower price the lender would get at a foreclosure sale. It is not surprising, then, that mortgage markets are indifferent to strip-down riskthe scope and magnitude of the potential loss is small and often less than that incurred in a foreclosure. 2. 2007 RIVERSIDE-SAN BERNARDINO DATABASE The 2001 CBP provides a guide for traditional mortgage-market conditions, but I was concerned about its predictive value for the current market, even though it was collected during a recession. Accordingly, I created an original data set of Chapter 13 filings that reflect current distressed real-estate market conditions. I recorded the market values and the mortgage-claim values for all real estate
170.
Id.
LEVITIN - FINAL
4/23/2009 11:40 AM
612
scheduled in Chapter 13 cases filed in the Riverside office of the Bankruptcy Court for the Southern District of California in the last four months of 2007. I selected Riverside because it represents a worst-case scenario for lenders, and because it can be isolated as a discrete community. Riverside-San Bernardino, located to the east-southeast of Los Angeles in the Inland Empire, had the third highest foreclosure rate of any metropolitan area in the country at the time,171 as well as for at least a year prior to the bankruptcy filings in the dataset.172 This means housing prices should have already been severely depressed in Riverside-San Bernardino, so the property values in my sample are likely to be lower than they might otherwise be. As of the third quarter of 2007, one out of every forty-three homes in Riverside-San Bernardino was in foreclosure.173 I was able to isolate Riverside-San Bernardino as a discrete community in bankruptcy filings because the bankruptcy court for the federal judicial district in which Riverside-San Bernardino is located has one of its five offices in Riverside. All of the cases filed in the Riverside office for this period were for debtors who listed their principal place of residence as either in Riverside or San Bernardino counties. Between September 1, 2007 and December 31, 2007, 984 Chapter 13 petitions were filed in the Riverside office. Of these, 720 were unique cases for which mortgage debt was scheduled and for which there was information on both the property value and mortgage-claim amount. I excluded timeshares, land-only mortgages, and liens for property taxes and homeowners association dues from my definition of mortgage debt. The 2007 Riverside data include principal homes, vacation homes, and rental properties. These cases yielded a sample of 1,094 mortgages, almost the precise size as the 2001 CBP database. The 2007 Riverside database includes 720 first-lien mortgages, 347 second-lien mortgages, 25 third-lien mortgages, and 2 fourth-lien mortgages. As Table 4 suggests, 48 percent of the properties in the dataset had more than one mortgage, and 3.5 percent had more than two mortgages.
171. Press Release, RealtyTrac, Stockton, Detroit, Riverside-San Bernardino Post Top Metro Foreclosure Rates in Q3 (Nov. 14, 2007), available at http://www.realtytrac.com/ContentManagement/pressrelease.aspx?ChannelID=9&Item ID=3609. 172. Press Release, RealtyTrac, Riverside-San Bernardino Foreclosures Increase 52 Percent in August (Sept. 24, 2006) (on file with author) (noting that 1 in 437 houses was in foreclosure). 173. See Press Release, RealtyTrac, supra note 171.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
613
Table 4: Mortgages in Chapter 13 (2007 Riverside)174 Cases Filed Cases with Mortgage Debt Scheduled Total Number of Mortgages Scheduled Number of 1st Lien Mortgages Number of 2nd Lien Mortgages Number of 3rd Lien Mortgages
984 720 1,094 720 347 25
The 2007 Riverside data track the 2001 CBP data closely. In Riverside, 21 percent of all mortgages in Chapter 13 were undersecured. In the 2001 CBP Chapter 13 sample, it was higher around 29 percent. This tells us that even in the hardest hit areas of the country, most mortgages that end up in bankruptcy are still not undersecured (at least according to debtors schedules). The likely explanation for the lower percentage of undersecured mortgages in Riverside is that, even with the current problems in the Riverside realestate market, many of the loans are still benefiting from the appreciation of the mortgage bubble. As Table 5 shows, the average undersecured mortgage in Riverside was undersecured by $47,800.45, or 3.6 times the average undersecured mortgage in Chapter 13 in the 2001 CBP database. This discrepancy must be discounted for the overall higher home values and loan amounts in California. In Riverside, the average property value was more than $425,000, over four times higher than the average Chapter 13 home value in the 2001 CBP of $105,266.05.175 Because of the influence on average property value on the amount over- or undersecured, the Security Ratio is a better metric for evaluating the impact on lenders.
174. See Pacer.psc.uscourts.gov, PACER Service Center, http:// www.pacer.psc.uscourts.gov (last visited Mar. 26, 2009) (displaying all Schedules A and D in Chapter 13 bankruptcy filings in the United States Bankruptcy Court, Central District of California: Riverside Division, between September 1, 2007 and December 31, 2007). 175. My initial sampling from other districts indicates that there is a strong correlation between average home prices and average amount undersecured. Id.
LEVITIN - FINAL
4/23/2009 11:40 AM
614
Table 5: Mortgages in Chapter 13 (2007 Riverside)176

Chapter 13RiversideAll Mortgages Fully Secured Undersecured Oversecured NUMBER 231 MARKET VALUE Average Std. Error Median CLAIM AMOUNT Average Std. Error Median $388,289.00 $8,668.56 $380,000.00 $206,365.00 $11,504.93 $114,200.00 $438,629.90 $6,079.97 $415,000.00 $257,111.60 $5,730.17 $264,160.00 $344,806.70 $29,096.23 $304,000.00 $264,735.20 $34,202.04 $280,000.00 $426,199.40 $5,102.01 $400,567.80 $246,542.70 $5,112.08 $244,316.00 842 21 1,094
All
MARKET VALUE MINUS (CLAIM AMOUNT PLUS AMOUNT OF SENIOR LIENS) -$47,800.45 $88,012.37 $0.00 $57,645.81 Average $3,031.18 $3,239.39 $0.00 $3,078.67 Std. Error -$39,316.42 $67,853.74 $0.00 $45,500.00 Median SECURITY RATIO -23.16% 34.23% 0.00% 23.38%
Table 6 and Chart 4 present comparisons of the 2001 CBP data and the 2007 Riverside data. The data are very similar. The Security Ratio for all mortgages in the 2001 CBP was 21 percent, compared to 23 percent for Riverside 2007. For undersecured claims, the Security Ratio was -21 percent for the 2001 CBP, compared to -23 percent for Riverside in 2007, while for the oversecured claims it was 37 percent for the 2001 CBP, compared to 34 percent for Riverside 2007. Thus, even in the worst-case market scenario for lenders, the loss as a percentage of claim amount from permitting strip-down is about the same as in the 2001 CBP.
176.
See Pacer.psc.uscourts.gov, supra note 174.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Table 6: Comparison 2007 Riverside)177
2001 CBP Total
Foreclosure Crisis
of
2001 CBP 1st Lien
615
(2001 CBP and
Security
2001 CBP 2nd Lien
Ratios
2007 Riverside Total
2007 Riverside 1st Lien
2007 Riverside 2nd Lien
Undersecured Mortgages Oversecured Mortgages All Mortgages

$500,000 $400,000 $300,000 $200,000 $100,000 $0 -$100,000 -$200,000 -$300,000 0
-21% 37% 21%
-10% 34% 22%
-84% 116% 11%
-23% 34% 23%
-11% 30% 23%
-57% 80% 22%
200
400 CBP 2001
600
800
1000
1200
Riverside 2007
Chart 4: Chapter 13 Mortgages Ranked by Amount Oversecured or Undersecured (2001 CBP and 2007 Riverside)178 Riverside in 2007 represents a worst-case scenario for lenders. The large number of foreclosures in the community for at least a year prior to the bankruptcy filings in the dataset pushed down already declining housing prices, and lenders had unusually large exposure on individual loans because the average loan size in California is larger than
177. Id.; 2001 Consumer Bankruptcy Project. 178. See Pacer.psc.uscourts.gov, supra note 174; 2001 Consumer Bankruptcy Project. This chart omits extreme outlier data points for scaling purposes.
LEVITIN - FINAL
4/23/2009 11:40 AM
616
anywhere else in the country.179 Most real-estate markets in the country are unlikely to ever be this bad. Accordingly, Riverside represents an outer limit of lender losses in an unlimited strip-down regime. Yet, strikingly, the 2007 Riverside data show little qualitative difference from the 2001 CBP data. Even in far worse market conditions, lenders Security Ratios are virtually unchanged, and a disproportionate amount of potential strip-down losses lies with the second-lien debt, much of which were piggyback mortgages in which lenders chose to forego PMI coverage.180 Piggyback lenders would thus fare better in bankruptcy under current law than in foreclosure. Taken as a whole, the 2001 CBP data and the 2007 Riverside data indicate that only a very small subset of all mortgagesonly undersecured mortgages in Chapter 13 casescould potentially be stripped down if section 1322(b)(2) were amended to allow modification of all mortgages. Therefore, even if one makes the overly conservative assumption that there would be no recoveries on the undersecured portion of the claim, lenders losses on undersecured mortgages if unlimited strip-down were allowed would be limited to around 20 percent of their claims. These are losses a lender would almost assuredly incur in a foreclosure situation and are far less than the 40 to 50 percent of loan value lenders are estimated to typically lose in foreclosure.181 Chart 5 illustrates the deadweight loss of foreclosure relative to bankruptcy. It superimposes the deficiencies from Monmouth County, New Jersey sheriffs sales and a random sampling of Indiana counties sheriffs sales182 with the amount by which mortgages in the 2001 CBP or 2007 Riverside data are undersecured. The juxtaposition is striking and shows that lenders would incur greater losses on a greater percentage of mortgages in foreclosure than in bankruptcy modification. The graph shows the deadweight loss of foreclosure relative to a modificationthe space between the CBP lines and the foreclosure lines is that deadweight loss.
179. See INSIDE MORTGAGE FIN., MORTGAGE STATISTICAL ANNUAL (2008). 180. See Jack Guttentag, Piggyback Loans vs. Insurance, WASH. POST, Mar. 15, 2008, at G2. 181. See supra text accompanying notes 910. 182. A random sampling of the Indiana data was used in order to have datasets of roughly comparable size for Chart 5.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
617
$500,000 Collateral Value Minus Claim Amount
$250,000
Riverside 2007 CBP 2001
$0
Indiana Counties Foreclosure Sales
Monmouth County, NJ Foreclosure Sales
-$250,000
-$500,000 0 200 400 600 800 1000 1200 CBP 2001 Mounmouth County, NJ Foreclosures Riverside 2007 Indiana Counties Foreclosures
Chart 5: Comparison of Foreclosure Sales with Hypothetical Bankruptcy Strip-downs183 The 2007 Riverside data confirm what the 2001 CBP data indicated: that mortgage markets are indifferent to strip-down risk because it is small in magnitude and likelihood, and may represent lesser losses than lenders would incur in foreclosure. Even in an unlimited strip-down regime (the most drastic type of modification) in the worst possible real-estate markets, mortgage lenders would not be exposed to substantial losses. Strip-down is a risk of very small probability and magnitude. Strip-down losses, relative to the size of the mortgage market, are just too inconsequential for lenders, and are not specifically figured into pricing models. Thus, the key to explaining the mortgage markets relative insensitivity to bankruptcy-modification risk lies in mortgage-market sensitivity to foreclosure costs. The markets indifference to bankruptcy-modification risk is because losses due to modification (including strip-down) would generally be smaller than those incurred
183. See Pacer.psc.uscourts.gov, supra note 174; 2001 Consumer Bankruptcy Project; see also Sheriffguadagno.com, supra note 136; SRI-sheriffsale.com, supra note 142. The Indiana counties sheriffs sale data plotted represent a random sampling designed to result in the same number of observations as the CBP datasets. The graph omits extreme outlier data points for scaling purposes.
LEVITIN - FINAL
4/23/2009 11:40 AM
618
in foreclosure.184 There is no reason for the market to price against bankruptcy modification if bankruptcy modification would result in smaller losses than foreclosure.185 Instead, modification (be it voluntary or in bankruptcy) represents the best realistic outcome for a defaulted loan. Moreover, bankruptcy-modification risk is small in likelihood and magnitude relative to all the other risk factors that determine mortgage interest rates above the cost of funds.186 IV. POLICY IMPLICATIONS
A. Voluntary Versus Involuntary Modification of Mortgages

As Part III shows, lender losses from bankruptcy modification would be less than from foreclosure. Accordingly, there is no reason for mortgage markets to price or ration credit against bankruptcymodification risk. Of course, if modification were permitted, more people might file for bankruptcy and modify their mortgages. But, lenders calculations of losses between foreclosure and bankruptcy modification do not depend on the number of mortgagors filing for
184. It is possible that the apparent market indifference is actually the result of cross-subsidization between mortgages that can be modified in bankruptcy and those that cannot. I do not believe that this is likely, however. First, it would require crosssubsidization to occur in several different segments of the marketoriginations, secondary market, and insurance. Insurers in particular are unlikely to have crosssubsidized price structures, as they are not subject to the same political pressure as the GSEs. Second, competition in these markets is a major force against crosssubsidization. If lender A has a cross-subsidized pricing structure that inflates the prices of mortgage loans on single-family, owner-occupied houses in order to hold down the prices of mortgage loans on multifamily and vacation homes, lender B will gladly come along and offer lower prices on single-family, owner-occupied, home-mortgage loans because they are a much larger segment of the market. And third, if there is crosssubsidization, we must ask why it would not also occur with rental properties. Accordingly, although cross-subsidization cannot entirely be ruled out, I believe it is not the likely explanation. 185. To the extent that bankruptcy judges valuations would sometimes be lower, it will be offset by higher returns on modified loans for creditors in some cases. As long as losses in bankruptcy are no greater than those in foreclosure, there should not be any effect on mortgage credit from allowing bankruptcy modification. 186. See supra text accompanying notes 16265. Another explanation might be possible, namely that even though foreclosure may be a worse outcome for lenders for any particular mortgage, it benefits lenders portfolios overall by creating a general deterrence against borrowers both entering into overly burdensome mortgages and not keeping their financial affairs in order after they have a mortgage. There is no empirical evidence to support such an explanation, however, and it is not clear that the deterrent effect would be less costly than more diligent initial underwriting. Moreover, the fact that lenders rarely pursue deficiency claims on mortgages, even when permitted, cuts against a deterrent function to foreclosure instead of workouts.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
619
bankruptcy. The calculations are on the individual loans, not on a pool; the general principle that foreclosure results in larger losses than modification does not depend on volume. Indeed, lenders should actually prefer more bankruptcy filings to more foreclosures. The trade-off for lenders, though, is not the binary bankruptcy versus foreclosure, but a tertiary one of bankruptcy modification versus consensual workout versus foreclosure. Modifications can be voluntary as well as involuntary. Bankruptcy is merely a forced workout, which limits lender control of the negotiation. It also imposes significant costs, such as attorneys fees. As such, one would expect lenders to prefer voluntary modification in loan workouts to involuntary ones in bankruptcy. Likewise, one would expect homeowners to prefer consensual workouts, as they incur costs in bankruptcy, too. In a Chapter 13 bankruptcy, the homeowner must live on a strict, court-supervised budget for the next three or five years.187 The homeowner will also incur a severely negative mark on her credit report, which will increase her future costs of credit. She will incur a privacy cost, as her finances will become a matter of public record. And she will incur a variety of transaction costs, including filing fees, attorneys fees, paperwork, and court-appearance time. Given that both lenders and borrowers would seem to prefer voluntary workouts, why have there not been more consensual modifications of distressed and defaulted mortgages? If workouts are a better option for lenders and borrowers, and voluntary workouts are less expensive than bankruptcy, why is not this what the market has moved to? To be sure, under strong U.S. Treasury Department (Treasury) jawboning, the HOPE Now Alliance, an organization of various mortgage-industry actors, including major mortgage servicers, has attempted to coordinate voluntary efforts to modify distressed mortgages.188 But a great many mortgages are still not being modified.189 Many of the workouts have been repayment plans that simply give the borrower ninety days to become current (monthly payments are the same going forward).190 A repayment plan is only a good solution to a defaulted mortgage when the default was caused by an isolated reduction in income or unexpected expense for the homeowner. In those cases, the homeowner
187. 11 U.S.C. 1325(b)(1)(4) (2006). 188. See Lynnley Browning, Distressed Owners Are Frustrated by Aid Group, N.Y. TIMES, Apr. 2, 2008, at C1. 189. See infra Chart 7. 190. Id.
LEVITIN - FINAL
4/23/2009 11:40 AM
620
just needs some time to readjust her finances. But if the problem is that the loan is generally unaffordable for the homeowner, such as because of an interest-rate reset, then a repayment plan is affirmatively unsuitable and likely only to delay the inevitable foreclosure. Not surprisingly, nearly 30 percent of foreclosure sales in the third quarter of 2007 involved failed repayment plans.191 Thus, although Chart 6, below, shows a robust quarterly ratio of total workouts (modifications and repayment plans) to foreclosure sales completed, and a nearly one-to-one ratio of workouts to foreclosure sales started, Chart 7, below, shows that repayment plans have predominated, and foreclosure starts still outstrip modifications by almost two-to-one. If one were to compare modifications or even all workouts with the total universe of distressed mortgages (many defaulted mortgages do not go into foreclosure for a while, especially given the backlog of foreclosure cases in the current market), the ratio would be far more discouraging; indeed, the ratios for the fourth quarter of 2008 are improved in part because of delayed foreclosures due to state legislation192
3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 2007Q1 2007Q2 2007Q3 2007Q4 2008Q1 2008Q2 2008Q3 2008Q4 Workouts to Foreclosure Sales Completed Workouts to Foreclosures Started
193
Chart 6: Total Workouts to Foreclosures
191. JAY BRINKMANN, AN EXAMINATION OF MORTGAGE FORECLOSURES, MODIFICATIONS, REPAYMENT PLANS, AND OTHER LOSS MITIGATION ACTIVITIES IN THE THIRD QUARTER OF 2007, at 10 (2008), available at http://www.mortgagebankers.org/ files/News/InternalResource/59454_LoanModificationsSurvey.pdf. 192. See infra text accompanying note 228. 193. See HOPE NOW, supra note 1
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
621
3.0 2.5 2.0 1.5 1.0 0.5 0.0 2007Q1 2007Q2 2007Q3 2007Q4 2008Q1 2008Q2 2008Q3 2008Q4 Repayment Plans to Foreclosures Started Repayment Plans to Foreclosure Sales Completed Modifications to Foreclosures Started Modifications to Foreclosures Sales Completed
Chart 7: Workouts to Foreclosures by Type194 Absolute numbers of modifications or repayment plans, however, present an incomplete picture of voluntary workout efforts. Many repayment plans and even many of the modifications, have resulted in higher monthly payments (such as by reamortizing arrearages into principal).195 Indeed, a survey by the Office of Comptroller of the Currency and Office of Thrift Supervision found that in 2008 workouts resulted in an increase in monthly payments 31 percent of the time and no change 27 percent of the time.196 While a quarterly breakdown, shown in Chart 9, below, shows a marked increase in the percentage of workouts in which monthly payments were reduced by more than 10 percent, it remains that nearly two-thirds of workouts result in little or no reduction in monthly payments in percentage terms.197 Not surprisingly, as Chart 10 shows, redefault rates were significantly
Id. See Alan White, Rewriting Contracts, Wholesale: Data on Voluntary Mortgage Modifications from 2007 and 2008 Remittance Reports, 36 FORDHAM URBAN L.J. (forthcoming 2009), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_
194. 195. id=1259538. 196. OFFICE OF THE COMPTROLLER OF THE CURRENCY & OFFICE OF THRIFT SUPERVISION MORTGAGE METRICS REPORT: FOURTH QUARTER 2008, at 7 (2009) [hereinafter OCC & OTS MORTGAGE METRICS REPORT], available at http://www.occ.treas.gov/ftp/release/2009-37a.pdf; see also infra Chart 8. 197. Id.
Workouts per Foreclosure
LEVITIN - FINAL
4/23/2009 11:40 AM
622
lower for workouts that resulted in greater reductions in monthly payments.198
P a ym e nts Incre a se d 31%
P a ym e nts D e cre a se d by M ore tha n 10%
P a ym e nts D e cre a se d by 1 0 % or Le ss 13% P a ym e nts U ncha nge d 27%
Chart 8: Impact on Monthly Mortgage Payments of Workouts in 2008199
198. 199.
Id. at 3537. Id. at 7.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
623
40% 35% 30% 25% 20% 15% 10% 5% 0% Q1 Q2 Q3 Q4 P a ym e nts D e cre a se d by M ore tha n 1 0 % P a ym e nts U ncha nge d P a ym e nts D e cre a se d by 1 0 % or Le ss P a ym e nts Incre a se d
Chart 9: Impact on Monthly Mortgage Payments of Workouts in 2008 by Quarter 200

60% 50% 40% 30% 20% 10% 0% 3 4 5 6 7 8 9 Months After Modification Decreased by More than 10% Unchanged Decreased by 10% or Less Increased
Chart 10: 60+ Day Serious Delinquency Rate on Loans Modified in the First Quart of 2008201
200. 201.
Id. Id. at 35.
LEVITIN - FINAL
4/23/2009 11:40 AM
624
There is a multicausal explanation for the paucity and inadequacy of voluntary workouts,202 including the psychology of financial distress that makes it hard for borrowers and servicers to get in contact,203 limited staffing and organization problems at servicers,204 contractual restraints on servicers ability to modify the terms of loans,205 servicers fear of lawsuits initiated by out-of-the-money MBS holders in response to modifications (tranche warfare),206 and perhaps most importantly, incentive misalignment that creates a principal-agent conflict between MBS holders and servicers.207 Foreclosure is often much more profitable for servicers than modification, so servicers have a strong financial incentive to foreclose on defaulted mortgages, even if a workout would be better for the MBS holders.208
202. 203.
the Judiciary, 110th Cong. 23 (Jan. 22, 2008) (written testimony of Adam J. Levitin) [hereinafter Levitin Testimony ], available at http://judiciary.house.gov/
hearings/pdf/Levitin090122.pdf. When people get deeply in debt, they frequently stop answering phone calls and opening mail because they do not want to face dunning calls and collection letters. Id. Mortgagors are also often mistrustful of servicers workout overtures because they follow months of dunning. CONG. OVERSIGHT PANEL, supra note 30, at 3738. 204. See Morgenson, supra note 9. 205. See CREDIT SUISSE FIXED INCOME RESEARCH, THE DAY AFTER TOMORROW: PAYMENT SHOCK AND LOAN MODIFICATIONS 46 (2007), available at http://www.credit-suisse.com/researchandanalytics (finding prohibitions or restrictions on modification in fourteen out of thirty-one securitization deals surveyed). Sometimes servicers are forbidden from engaging in any sort of modification, other times they are limited to modifying only a small percentage of loans in a pool, other times they are forbidden from writing down principal, and other times they are required to purchase any modified loans at par. See Gelpern & Levitin, supra note 29. 206. See generally Eggert, supra note 28. Redefault rates have been higher for modifications of securitized loans than loans held directly by lenders. OCC & OTS MORTGAGE METRICS REPORT, supra note 196, at 23. 207. Until the 2007 tax year, the federal income-tax system created a strong disincentive for homeowners to engage in voluntary workouts, as any debt voluntarily forgiven by the lender would be imputed to the homeowner as taxable income. 26 U.S.C. 61(a)(12) (2006) (including income from discharge of indebtedness in gross income, except as otherwise provided); id. 108(a)(1)(A) (excluding discharge of debt in a bankruptcy case from the definition of gross income ); see also United States v. Kirby Lumber Co., 284 U.S. 1, 3 (1931); Stephen B. Cohen, Mortgage Double Whammy: First You Default; Second, Youre Taxed, 117 TAX NOTES 169 (Oct. 8, 2007) (arguing that tax case-law doctrine might have provided a solution to the voluntary-workout disincentive). The Mortgage Forgiveness Debt Relief Act of 2007 has removed this disincentive for homeowners with less than $2 million in principalresidence mortgage debt, but only for debt discharge that occurs before 2010, after which the Internal Revenue Code will again create a disincentive for all voluntary workouts. Pub. L. No. 110-142, 121 Stat. 1803, 180304 (codified as amended in scattered sections of 26 U.S.C.). 208. See Levitin Testimony, supra note 203, at 1314.
See Eggert, supra note 28, at 757. See Hearing on H.R. 200 and H.R. 225: Hearing Before H. Comm. on
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
625
In particular, the servicer/MBS-holder dynamic appears responsible for the political economy of bankruptcy-modification legislation. Servicers have been the political voice of the mortgage industry, but servicers interests often diverge from those of MBS holders.209 Thus, voluntary mortgage modification is often hindered by contractual and agency impediments. Notably, when agency and contractual impediments to modification cease to exist, as with loans assumed by the FDIC as received for failed insured banks, voluntary modifications do take place. Thus, when the FDIC took over the failed IndyMac Federal Savings Bank, it promptly proposed a plan to modify tens of thousands of portfolio and securitized mortgages in order to boost the value of the failed thrift.210 Loan modifications rather than foreclosures increase the value of loan assets. Therefore, when a single portfolio lender holds the assets unimpeded by agency or contractual obligations, it is likely to engage in the modifications. Permitting modification of mortgages in Chapter 13 would provide a solution to contractual and agency problems that parallels Chapter 11s solution to the contractual and statutory impediments in bond debt workouts. The Trust Indenture Act of 1939 provides that the terms of bond indenturethe legal document that sets the terms of a bond issuancerelating to defaults may not change, absent consent of a majority of bondholders, and that no bondholder may be forced to accept a modification of its financial rights under the indenture.211 Moreover, bond indentures typically prohibit modification of any of the bonds terms absent consent of all or a supermajority of the bondholders. Because of such high consent thresholds, voluntary consensual modification of the bonds is near impossible.212 Instead, if a
209. Id. 210. Joe Adler, FDIC Offers an IndyMac Loan Mod Plan, AM. BANKER, Aug. 21, 2008, at 1. 211. 15 U.S.C. 77ppp(a)(1)(B) (providing that, unless specified in the indenture, holders of at least a simple majority of the principal amount of a bond issue must consent to waivers of past defaults); id. 77ppp(b) (providing individual bondholders with a right to refuse modification of the payment of principal and interest on their bond as due). 212. See Mark J. Roe, The Voting Prohibition in Bond Workouts, 97 YALE L.J. 232, 278 (1987). Bond workouts typically involve an exchange offer, in which the old bonds are exchanged for new bonds with a lower interest rate (or principal amount), but individual bondholders may refuse to participate in the exchange offer. 15 U.S.C. 77ppp(b). This creates a potential holdout problem, as the bondholders who do not participate in the exchange benefit from the debtors increased ability to service their bonds because of the decreased debt-service demands from those bondholders who participated in the exchange. This situation creates an incentive for bondholders to try to be the holdout and unfairly benefit from those who will participate in the exchange,
LEVITIN - FINAL
4/23/2009 11:40 AM
626
company wishes to restructure its public debt, it must use the involuntary-modification procedures of Chapter 11. But for Chapter 11, companies with public debt would not always be able to reorganize if they were not able to meet their bond-coupon obligations.213 Voluntary bond workouts thus present parallel problems to voluntary home-mortgage workouts. Accordingly, a Chapter 13 mortgage modification parallel to Chapter 11s provisions permitting bond-indenture modification is necessary. The existence of a bankruptcy-modification option would shift the dynamics of voluntary workouts. To the extent that borrowers have a bankruptcy option, it puts pressure on servicers and lenders to make a deal outside of bankruptcy.214 Permitting bankruptcy modification of mortgages would make voluntary modification of mortgages more likely, and would also make it more likely that Chapter 13 plans would succeed, as debtors who wished to retain their homes would have lower payment burdens to meet.215
B. Bankruptcy Modification Compared with Other Policy Responses

Bankruptcy modification is not the only possible response to the foreclosure crisis. There are a number of other possible solutions, which can be divided into six categories: (1) self-correction within the market; (2) voluntary industry efforts to modify mortgages; (3) mandatory procedural steps to encourage voluntary workouts; (4) foreclosure and/or rate-increase deferrals and moratoria; (5) state or federal takings of mortgages, followed by unilateral modification; and (6) federal refinancing or insurance of mortgages. Compared with
and the collective-action problem the holdout engenders can frustrate voluntary workouts. See Roe, supra, at 236. 213. See Roe, supra note 212, at 236 (explaining that public debt creates a holdup problem). Companies without public debt, mass tort liabilities, or complicated collective-bargaining or pension/benefit liabilities are often able to restructure their debts voluntarily outside of Chapter 11. When Chapter 11 is used, it is typically to take advantage of bankruptcy-court blessings of asset sales or to engage in a liquidation absent a trustee (and with favorable provisions for the debtors attorneys fees). 214. Indeed, even the threat of bankruptcy-modification legislation can shift the dynamics of voluntary workouts. The HOPE Now Alliance was formed in part to diffuse political pressure for changing bankruptcy law to permit modification on all property types. Hopenow.com, About Us, http://www.hopenow.com/hopenow_ aboutus.html (last visited Feb. 11, 2009). 215. See Michelle White & Ning Zhu, Saving Your Home in Chapter 13 Bankruptcy (Natl Bureau of Econ. Research Working Paper No. W14179, 2008), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1165507 (modeling lower Chapter 13 default rates if strip-down were permitted).
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
627
bankruptcy modification, all of these proposed solutions have serious drawbacks. 1. LAISSEZ-FAIRE MARKET SELF-CORRECTION A complete laissez-faire approach of letting the market correct itself fails to address two core problems. First, the market may take significant time to correct itself, and second, the correction, featuring high levels of foreclosures, will have major externalities that will hurt the economy overall. Moreover, the residential mortgage market features rigid and hard-to-amend contracts that prevent the market from arriving at what would be the efficient outcome.216 In contrast, bankruptcy modification would be immediately effective as soon as a homeowner filed for bankruptcy, and by helping families keep their homes, it would mitigate the externalities of foreclosure. 2. COORDINATED VOLUNTARY-WORKOUT EFFORTS Voluntary modification of mortgages would itself be a form of market correction. Indeed, voluntary modification of distressed mortgages would be the ideal solution that would occur in the perfectly efficient economy that functioned without transaction costs, contractual limitations, and agency problems. But in the modern mortgage industry, these factors pose considerable obstacles to voluntary modification. Accordingly, industry efforts (with the support of the Treasury and the Department of Housing and Urban Development) to encourage voluntary modification have been of limited effectiveness.217 Indeed, it is hard to see how coordination among lenders would add significantly to voluntary modifications other than by creating an industry standard that would provide some comfort for servicers who want to follow best practices. The failure (or limited success) of these coordinated efforts, however, argues for the need for an involuntary modification procedure. 3. PROCEDURAL REQUIREMENTS TO ENCOURAGE CONSENSUAL
WORKOUTS
Likewise, proposals to encourage voluntary modification via procedural requirements, such as requiring lenders to engage in goodfaith workout discussions with borrowers before proceeding with
216. 217.
See generally Gelpern & Levitin, supra note 29. CONG. OVERSIGHT PANEL, supra note 30, at 3.
LEVITIN - FINAL
4/23/2009 11:40 AM
628
foreclosure, are likely to fall short.218 Proposed legislation that would require such discussions does not require any substantive offers from lendersit is purely procedural. The main impact of such legislation would be to slow down the foreclosure process. This delay could be helpful to the extent it gives the market the necessary time to correct, or to the extent it facilitates meaningful workout discussions that would not have otherwise occurred. But delaying foreclosures could also delay the markets correction because it will take longer for the foreclosed properties to come on the market. Additionally, such procedural requirements would increase the costs of foreclosure both through transaction costs and delay, and this, unlike the costs of bankruptcy, might well impact mortgage-market credit availability and cost, although one would expect the impact to be de minimis and possibly offset by any successful resulting workouts.219 4. FORECLOSURE AND RATE-INCREASE MORATORIA AND OTHER
LIMITATIONS ON THE FORECLOSURE PROCESS
A variety of foreclosure deferral and moratorium plans have been proposed. Such deferrals and moratoria were a common reaction to the last great foreclosure crisis, that of the Great Depression.220 The deferral plans proposed currently would defer foreclosure and/or rate increases for a limited window if the borrower undertook certain steps and continued making certain minimum payments.221 The Sheriff of
218. See, e.g., Foreclosure Prevention and Sound Mortgage Servicing Act of 2008, H.R. 5679, 110th Cong. 6A(d) (2008). 219. Philadelphias Residential Mortgage Foreclosure Diversion Pilot Program, a program in which the city brokers loan modifications between homeowners and lenders, appears to be successful. See Jon Hurdle, Success Seen in Program to Save Homes, N.Y. TIMES, Sept. 28, 2008, at A22. 220. D.P.K., Comment, Constitutional LawMortgage Foreclosure Moratorium Statutes, 32 MICH. L. REV. 71, 71 (1933) (noting that, in 1933, twenty-one states enacted legislation that functioned as foreclosure moratoria). Depression-era foreclosure-moratorium statutes seem to have either extended the period of redemption postforeclosure, prohibited foreclosures unless the sale price was at some minimum percentage of property appraisal, or granted state courts the power to stay foreclosures.
Id.
221. See, e.g., Home Retention and Economic Stabilization Act of 2008, H.R. 6076, 110th Cong. 128A(a)(2) (2008) (providing for deferral of foreclosure up to 270 days if, inter alia, minimum payments were made); MD. CODE ANN., REAL PROP. 3104.1, 7-105.1 (LexisNexis 2008) (requiring post-default delay and a specific form of service for foreclosure actions); Minnesota Subprime Foreclosure Deferment Act of 2008, H.F. 3612, 2008 Leg., 85th Sess. (Minn. 2008), available at https://www.revisor.leg.state.mn.us/bin/getbill.php?number=HF3612&session=ls85& version=list&session_number=08session_year0 (providing for foreclosure deferral up to one year if, inter alia, minimum payments were made).
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
629
Cook County, Illinois (where Chicago is located) has refused to enforce foreclosure actions.222 Likewise, the Sheriff of Philadelphia County, Pennsylvania, has instituted his own foreclosure moratorium, refusing to conduct foreclosure auctions, but even though this action has the blessing of the Philadelphia City Council (to which the Sheriff does not answer), it appears to be a civil-disobedience measure rather than a legislative response.223 Foreclosure deferrals or moratoria would delay foreclosures temporarily, which might facilitate workouts. Delay would extend the window of time in which homeowners could work with servicers and lenders. Delay might also facilitate workouts by imposing time-value costs on lenders and on servicers. Lenders have to wait longer for a foreclosure realization and, unless the property sells for more than the principal balance due, the lender will not be able to recover the interest that accrued between the time the foreclosure was commenced and completed. If state law allows, the lender will have a hard-to-collect, unsecured deficiency claim. For servicers, delay imposes costs too because servicers must advance delinquent payments to lenders (or the securitization trust) out of pocket. These advances are reimbursed off the top of foreclosure-sale proceeds, but the reimbursement does not include time value, which can be considerable if a foreclosure takes eighteen to twenty-four months. Nonetheless, it is questionable whether merely imposing costs through delay would solve the underlying problem. If anything, it might result in some economically efficient foreclosures from not happening. To the extent that delay from a de facto or de jure foreclosure moratorium is positive, though, it would function like bankruptcy currentlythe automatic stay stops foreclosure proceedings, but unless the homeowner can cure and reinstate the mortgage, the stay will be lifted. Likewise, rate-freeze moratoria, such as the one proposed by thenSenator Hillary Clinton,224 also only delay problems, not solve them, and they do not even delay foreclosures caused by negative equity or existing rate resets. Both types of moratoria arguably impinge on lenders property rights, raising serious constitutional issues of
222. Ofelia Casillas & Azam Ahmed, Sheriff: I will stop enforcing evictions, CHI. TRIB., Oct. 9, 2008, at C1. 223. Jeff Blumenthal, Moratorium on Sheriffs Foreclosure Sales Draws Debate, PHILA. BUS. J., Apr. 4, 2008, available at http://philadelphia.bizjournals.com /philadelphia/stories/2008/04/07/story10.html. 224. Press Release, Senator Hillary Clinton, Details on Senator Clintons Plan to Protect American Homeowners (Mar. 24, 2008), available at http://2008central.net/2008/03/24/clinton-press-release-clinton-calls-for-bold-action-tohalt-housing-crisis.
LEVITIN - FINAL
4/23/2009 11:40 AM
630
takings225 and state-law impairment of existing contract obligations226 that do not exist when lenders property rights are curtailed under Congresss bankruptcy power.227 Moreover, foreclosure moratoria are an overly broad solution; there are some mortgagors that cannot be helped and foreclosures should happen in these cases. Moratoria might also reduce mortgage-credit availability in states that adopt such a solution, at least for a while. Short of an outright foreclosure moratorium, however, there are several steps that states could take that would alleviate some of the foreclosure crisis, and some have started to do so.228 States are certainly free to make foreclosure more onerous for lenders. This could be done by putting up lots of procedural barriers to foreclosure, such as mandatory good-faith negotiations with borrowers, personal-service requirements, longer notice periods, redemption rights, pre-saleappraisal requirements, bans on deficiency judgments, minimum foreclosure-sale bids based on appraisal value, and requirements that servicers show their net-present-value (NPV) calculation for foreclosure versus modification to the court as a condition of foreclosure. Alternatively, because section 1322(b)(2) looks to state law to determine whether a security interest is a security interest solely in real property, states could clarify or amend their law to narrow the definition of real property such that mortgages that cover rents, profits, extracts, and fixtures would not be treated as covering solely real property. Such a change, however, could have adverse affects on other areas of state law. The initial question when considering foreclosure or rate-increase moratoria or any other change in state foreclosure law is whether the social-welfare gains of preventing foreclosures that should not occur (because modifications would be feasible and more efficient) outweigh the social-welfare costs of some necessary foreclosures not happening.
225. U.S. CONST. amend. V (applying to the states via section 1 of the Fourteenth Amendment). 226. U.S. CONST. art. I, 10, cl. 1. But see Home Bldg. & Loan Assn v. Blaisdell, 290 U.S. 398, 437 (1934) (upholding a Depression-era Minnesota foreclosure moratorium in the face of a contracts-clause challenge, and noting that the economic conditions of the Depression may justify the exercise of its continuing and dominant protective power notwithstanding interference with contracts). Blaisdell has been heavily criticized. See, e.g., Richard A. Epstein, Toward a Revitalization of the Contract Clause, 51 U. CHI. L. REV. 703, 737 (1984). 227. U.S. CONST. art I, 8, cl. 4. 228. See, e.g., CAL. CIV. CODE 2923.5.6 (West 2008) (imposing a delay and a NPV-maximization requirement); MD. CODE ANN., REAL PROP. 7-105.1 (LexisNexis 2008); MASS. GEN. LAWS ch. 244, 35A(a) (2008) (imposing a ninetyday-pre-foreclosure cure period).
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
631
It would seem that this balancing tilts in favor of foreclosure moratoria, but if we then ask whether it is the best possible solution, the answer is clearly no, because foreclosure and rate-freeze moratoria are clumsy and overbroad market interventions. 5. GOVERNMENT MODIFICATION FOLLOWING EMINENT-DOMAIN TAKINGS Involuntary modification could occur through methods other than bankruptcy. Either the states or the federal government could seize and modify, refinance, or guarantee troubled mortgages.229 Potentially, the states or the federal government could use their eminent-domain powers to seize mortgages from lenders in exchange for fair compensation.230 State or federal governments could then unilaterally modify the mortgages (likely contracting out this work). Eminent-domain seizures are not a quick process, and time is of the essence in dealing with foreclosures. There is a right to a jury trial in federal eminent-domain actions on the issue of just compensation.231 Further, state or federal seizure of mortgages would be prohibitively costly, place the public fisc at risk, and likely result in significant litigation over the valuation of mortgages. It would also create moralhazard problems by providing an unfair bailout for homeowners (and depending on the compensation paid, lenders). 6. GOVERNMENT REFINANCING, GUARANTY, OR
INSURANCE OF MORTGAGES
Federal or state-government refinancing or insurance of mortgages is also problematic. Any governmental refinancing or insurance would put the public fisc at risk and potentially result in an unfair bailout of either lenders or borrowers, creating moral hazard. To date there have been three major federal attempts to stanch the foreclosure crisis: FHAs FHASecure program, the HOPE for Homeowners Program, and the Homeowner Affordability and Stability Program (HASP).232
229. Howell E. Jackson, Build a Better Bailout, CHRISTIAN SCI. MONITOR, Sept. 25, 2008, at Op. 9; Lauren E. Willis, Stabilize Home Mortgage Borrowers, and the Financial System Will Follow (Loyola-L.A. Legal Studies Paper No. 2008-28, 2008), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1273268. 230. See, e.g., Kelo v. City of New London, 545 U.S. 469, 489 (2005). 231. United States v. Reynolds, 397 U.S. 14, 19 (1970); FED. R. CIV. P. 71.1(h)(1). Congress may create a special tribunal in lieu of a jury to determine just compensation. Id. 232. There is also a pair of programs proposed by FDIC that have not been adopted, but which merit consideration. Some features of these programs have been
LEVITIN - FINAL
4/23/2009 11:40 AM
632
incorporated into HASP. While the FDIC proposals are by no means the only existing proposals, they present detailed, comprehensive proposals for government refinancing and/or guarantee of mortgages and merit discussion. One of the FDIC proposals calls for direct federal refinancing of distressed mortgages, or Home Ownership Preservation Loans. The FDIC plan calls for Congress to authorize the Treasury Department to refinance up to 20 percent of the outstanding principal on eligible unaffordable mortgages in exchange for the lenders agreement to restructure the mortgages into fully-amortized, fixed-rated loans for the balance of the original loan term. FDIC.gov, FDIC Home Ownership Preservation Loans: Questions and Answers, http://www.fdic.gov/consumers/loans/hop/qa.html (last visited Jan. 31, 2009) (The program could be limited to mortgages for owner-occupied residences that are unaffordabledefined by front-end [debt-to-income ratios] exceeding 40 percent at origination. In addition the loan could be required to be below the FHA conforming loan limit. Finally, the loans eligible for the program could be limited to those originated between January 1, 2003 and June 30, 2007.). Additionally, the plan caps interest rates at the Freddie Mac 30-year-fixed rate and imposes a superpriority lien on the property for the Treasury Department, meaning that the Treasury would get paid first in any refinancing or foreclosure sale before any existing mortgagees. FDIC.gov, FDIC Home Ownership Preservation Loans, http:// www.fdic.gov/consumers/ loans/hop (last visited Jan. 31, 2009). Homeowners would be required to pay off the restructured mortgage, including the Treasury-refinanced component, but the payments on the Treasury-refinanced component would be delayed for five years and then amortized over the rest of the mortgages tenor. See FDIC.gov, FDIC Home Ownership Preservation Loans: Questions and Answers, supra. Lenders would have to pay the first five years of interest (the governments cost of carry) on the Treasuryrefinanced component upon closing. Id. There are many good qualities to the FDIC plan. It avoids the second-mortgageeholdup problem by simply not addressing the issue; because there is no refinancing, the second mortgagee remains subordinated and has no holdup power. Second mortgages are not a concern for the FDIC plan because it does not aim to address negative-equity problems. The FDIC plan would also avoid the HOPE for Homeowners programs lemon problem, and, assuming that properties maintain at least 25 percent or so of their value (which seems reasonable even in the worst-case housing market) and that the governments superpriority lien is properly perfected, it ensures that the government will eventually be paid back in full. (The Treasury loan would be 20 percent of the outstanding principal. Id. As there is no mention of a LTV requirement for refinancing, the Treasury loan could be for 20 percent of 110 percent or 125 percent or the propertys current value. That is, the Treasury LTV ratio could be closer to 2225 percent in extreme cases.) It would be much simpler administratively, and available much quicker than HOPE for Homeowners. But the FDIC plan suffers from some of the same critical flaws of HOPE for Homeowners and HASP program. It would still leave an unresolved moral-hazard issue for borrowers, who would find themselves bailed out by the government. Likewise, it would create moral hazard for lenders, as they would be placed in a more favorable mortgage at no cost. And more crucially, the FDIC Homeownership Preservation Loan plan also fails to address the contractual and agency problems that exist because of securitization. The FDIC plan would place the onus on lendersthat is servicersfor applying for Treasury refinancing and undertaking the restructuring. These servicers would receive no benefit from the refinancing and would potentially incur both the large cost of fronting the five years of interest payments and the costs of restructuring the mortgage. The Homeownership Preservation Loan program lacks the servicer incentive payments
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
633
featured in HASP. Moreover, the FDIC plan could create additional problems for securitized mortgages because it would alter the securitization trusts cash flows without buying out the mortgages in whole. Securitization trusts have timely-payment obligations to investors, which require a certain schedule of cash inflows that the FDIC plan would interrupt. Critically, as proposed and unlike HASP, the FDIC program would only help those borrowers who received loans that were unaffordable when made. It would not help homeowners whose incomes declined since taking out the mortgages, since it would look to affordability at time of origination. Nor would it address the sizeable problem created by negative equity. The FDICs proposed government refinancing program would undoubtedly help many homeowners, but it would fail to resolve many distressed-mortgage situations or to create long-term housing market stability. The FDIC has also proposed a loan-modification program that is designed to promote systematic modifications of delinquent owner-occupied-property mortgages by offering incentives to servicers and investors. FDIC.gov, FDIC Loss Sharing Proposal to Promote Affordable Loan Modifications, http://www.fdic.gov/consumers/ loans/loanmod/index.html (last visited Jan. 31, 2009). The core of the FDIC program is an eight-year government guarantee of up to 50 percent of losses on redefaults of qualifying modified loans, with the guarantee being phased out as the LTV ratio rises from 100 percent to 150 percent. Id. The guarantee would kick in after six months of successful payments on modified loans. Id. In order to qualify for the government guarantee, the FDIC proposal would require servicers to undertake a systematic review of their entire servicing portfolios, and subject each loan to a standardized NPV calculation to determine if it is suitable for modification based on a 31 percent front-end (housing debt only) debt-to-income (DTI) ratio. Id. Servicers would be required to modify all qualifying loans in order to qualify for the government redefault guarantee. Id. The FDIC also proposes paying servicers a $1,000 bounty per modification. Id. There are two major questions about the FDIC plan. First is whether it will induce mass modifications. Its all-or-nothing approach could easily backfire if the incentives are not properly calibrated overall. There are many reasons to think that they are not. While this FDIC plan recognizes servicers incentives as an issue, it only proposes paying a modification bounty that is a rough proxy for servicers modification costs. Modification costs are but one piece of servicer incentives. There are also issues like the cost of making advances on nonperforming loans to the securitization trust and litigation risks, which the FDIC plan does not address. The FDIC plan would require servicers to breach their servicing contract in many cases. Absent a litigation safe harbor, servicers will be reluctant to do so. Likewise, the plan is intended to shift servicers NPV calculation through the guarantee against redefaults. But this fails to account for PMI cancellations impact on NPV. Loan modification can often result in the termination of PMI. PMI is typically first-loss-position insurance, capped at 30 percent of the mortgage. See Johnstone, supra note 90, at 786. Loss of 30 percent first-loss-position insurance is not compensated for by the maximum 50 percent pari-passu government guarantee that would only kick in after six months of successful payments. To wit, assuming a 45 percent loss in foreclosure, investors would do better by foreclosing and taking the PMI coverage (net loss of 15 percent after the 30 percent PMI first-loss position) than if the loan redefaulted, even after ten months of payments and the government guarantee kicked in (net loss of 22.5 percenthalf of a 45 percent lossminus some payments made). Thus, the FDIC guarantee might not tilt NPV calculations toward modification. Second, even if the FDIC plan were successful at producing mass modifications, it is far from clear that these would be sustainable. The plan does not require principal
LEVITIN - FINAL
4/23/2009 11:40 AM
634
a. FHASecure and HOPE for Homeowners Act

FHASecure and HOPE for Homeowners were the federal governments initial attempts at foreclosure prevention. These programs failed abysmally. The FHASecure program refinanced homeowners with non-FHA ARMs (including those with negative equity) into FHA fixed-rate mortgages.233 It was predicted to help 240,000 homeowners,234 but had only helped a few thousand delinquent borrowers before being shut down at the end of 2008.235 Likewise, the HOPE for Homeowners program, established by Congress in July 2008, authorizes the FHA to insure certain refinanced distressed mortgages.236 It was predicted to help 400,000 homeowners, but as of mid-December 2008, it had attracted only 312 applications,237 and had not actually refinanced any mortgages238 in part because of its reliance on private-market cooperation, which means that it cannot get
write-downs, so it cannot address the negative-equity problem. Nor does it attempt to deal with second liens, which contribute to the negative-equity problem. And its reliance on a front-end DTI target, rather than an examination of the homeowners entire finances, raises redefault risk. Finally, even if the FDIC plan would produce mass sustainable modifications, it would do so at taxpayer expense. The FDIC Wholesale Modification Plan creates moral-hazard problems for homeowners and lenders and fails to convincingly address restrictive PSAs, servicer incentives, negative equity, and second liens. Moreover, without exacting affordability standards, the FDIC plan cannot guarantee a good return for taxpayers. While the Plan is also more promising than HOPE for Homeowners, it suffers from many of the same problems. 233. Portal.hud.gov, FHASecure Fact SheetRefinance Options, http://portal.hud.gov/portal/page?_pageid=73,1824154&_dad=portal&_schema=POR TAL (last visited Mar. 26, 2009) (FHASecure is a refinancing option that gives homeowners with non-FHA adjustable rate mortgages (ARMs), current or delinquent and regardless of reset status, the ability to refinance into a FHA-insured mortgage.). 234. See, e.g., Press Release, U.S. Dept of Hous. & Urban Dev., Bush Administration to Help Nearly One-Quarter of a Million Homeowners Refinance (Aug. 31, 2007), available at http://www.hud.gov/news/release.cfm?content=pr07-123.cfm; Press Release, U.S. Dept of Hous. & Urban Dev., FHA Helps 400,000 Families Find Mortgage Relief (Oct. 24, 2008), available at http://www.hud.gov/news/ release.cfm?content=pr08-167.cfm. 235. Michael Corkery, Mortgage Cram-Downs Loom as Foreclosures Mount, WALL ST. J., Dec. 31, 2008, at C1. 236. HOPE for Homeowners Act of 2008, 1402(a), (e)(2)(B), Pub. L. No. 110-289, 122 Stat. 2654, 2801. 237. Dina ElBoghdady, HUD Chief Calls Aid on Mortgages a Failure, WASH. POST, Dec. 17, 2008, at A1. 238. All Things Considered: Despite Program, No Hope for Homeowners (NPR broadcast, Dec. 17, 2008), available at http://www.npr.org/templates/story/ story.php?storyId=98409330.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
635
around restrictions in servicing agreements or servicers misaligned incentives or second liens.239 HOPE for Homeowners presents an adverse-selection problem that creates exposure of the public fisc. The program requires lenders to write down loans to a 85.5 percent LTV ratio based on a new, independent appraisal.240 To the extent that lenders are willing and able to do the write-down necessary for the FHA refinancing, they will only do so for loans that they think are worth less than eighty-five cents on the dollar (or actually much lower, given that the original face value outstanding might be more than 100 percent of the current property value).241 Lenders will retain loans with a higher expected recovery rate. This means there is an adverse-selection problem for the FHA refinancing. Lenders will only sell the FHA their worst lemons, so the FHA will be overpaying for bum loans. Lenders contributions to an FHA loss-reserve fund, and a special tax on Fannie Mae and Freddie Mac, are supposed to protect against FHA losses,242 but this depends on untested assumptions about the number of loans that will be refinanced and the potential losses on those loans. While the legislation specifically directs the FHA to adopt regulations to prevent adverse selection,243 it is not at all clear how this
239. Adam J. Levitin, Flaws in the FHA Housing Bill, WALL ST. J., July 11, 2008, at A15. 240. The HOPE for Homeowners Act of 2008 requires a maximum 90 percent LTV ratio for FHA refinancing. 1402(e)(2)(B), 122 Stat. at 2801. This means that if the lender is perfectly secured, the lender will have to write down the principal by 10 percent. If the lender is undersecured, the lender will have to write down the principal by a greater amount. Additionally, all lenders are required to pay insurance premiums on the mortgage of 3 percent of the principal initially, and 1.5 percent of the principal remaining on an annual basis. Id. 1402(i)(2). The LTV requirement has been reduced to 96.5 percent in some cases, where debt-to-income ratios are close to traditional underwriting guidelines. Press Release, U.S. Dept of Hous. & Urban Dev., Bush Administration Announces Flexibility for Hope for Homeowners Program (Nov. 19, 2008), available at http://www.hud.gov/news/release.cfm?content=pr08-178.cfm. 241. Kate Berry, Refi-Program Previewers Raise Issues, AM. BANKER, Aug. 19, 2008 (If a bank owns the loan outright, its fine because theyre basically swapping a semi-nonperforming asset for one that is now insured by the government.) (quoting Dan Cutaia, president of Fairway Independent Mortgage Corp); id. (A Wall Street firm may say that a loan is worth only 80 cents on the dollar if they try to sell it, but if FHA can refinance it at 90 cents, this is a viable way to do it.) (quoting Ronald Faris, president of Ocwen Financial Corp., a leading mortgage servicer). 242. The Hope for Homeowners Act of 2008, 1402(a), (i), 122 Stat. at 2801. 243. Id. 1402(h), 122 Stat. 2654, 2804 (Standards To Protect Against Adverse Selection-(1) IN GENERAL- The Board shall, by rule or order, establish standards and policies to require the underwriter of the insured loan to provide such representations and warranties as the Board considers necessary or appropriate to enforce compliance with all underwriting and appraisal standards of the HOPE for Homeowners Program.).
LEVITIN - FINAL
4/23/2009 11:40 AM
636
could be accomplished. Indeed, if there were an obvious solution, it should have been in the legislation itself. Finally, there are moral-hazard problems with a program that bails out homeowners and lenders. These problems are mitigated somewhat by the requirements of a lender write-down and sharing of appreciation in the propertys value, but any government program that puts lenders or borrowers in a better position than they would be without government intervention raises a moral-hazard flag. Ultimately, government insurance of refinanced mortgages is unlikely to help more than a minority of distressed homeowners, and poses potential moral-hazard problems and very serious adverseselection problems.
b. Homeowner Affordability and Stability Program

HASP, announced in general terms on February 18, 2008, with subsequent detailed guidelines announced on March 4, 2008, has two main foreclosure prevention components. First, it authorizes lowinterest-rate refinancings of mortgages held or guaranteed by Fannie Mae and Freddie Mac, even if there was negative equity of up to a 105% LTV ratio. Second (and not exclusively), HASP creates a loanmodification program aimed at achieving affordable monthly payments. Only owner-occupied principal residences are eligible for HASP (among other conditions),244 but the Treasury has made additional Troubled Asset Relief Program (TARP) funds contingent upon servicers participation in the program.245 Under the HASP modification program, servicers would be required to modify loans only for those which the NPV of the modified loan, using standardized assumptions, would be greater than the net present foreclosure value. HASP provides that if a servicer reduces monthly payments of principal, interest, taxes, and insurance (PITI) to 38 percent of gross monthly income, then the federal government will match further payment reductions dollar for dollar until PITI is 31
244. See U.S. DEPT OF THE TREASURY, HOMEOWNER AFFORDABILITY AND STABILITY PLAN FACT SHEET 2 (2009), available at http://www.treasury.gov/initiatives/ eesa/homeowner-affordability-plan/FactSheet.pdf; Press Release, U.S. Dept of the Treasury, Home Affordable Modification Program Guidelines 2 (Mar. 4, 2009), available at http://www.ustreas.gov/press/releases/reports/modification_program_ guidelines.pdf. 245. U.S. DEPT OF THE TREASURY, HOMEOWNER STABILITY AND AFFORDABILITY PLAN: EXECUTIVE SUMMARY (Feb. 18, 2008), available at http://www.ustreas.gov/press/releases/tg33.htm.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
637
percent of gross monthly income.246 Servicers are permitted to reduce payments by reducing interest rates (to as low as 2 percent), forbearing on principal, or forgiving principal; servicers are not required to engage in forgiveness or forbearance.247 The reduced payments are not for the life of the loan, but only for five years, after which point payments may be increased up to the GSE conforming rate in place at the time. To encourage participation in the program, the government is offering a variety of financial incentives for servicers and lenders. First, servicers will receive an up-front fee of $1,000 for each modification.248 Second, servicers will receive a pay for success fee of up to $1,000 per year in monthly installments for up to three years as long as the borrower stays current on the loan and monthly payments are reduced by at least 6 percent.249 Third, if servicers modify at-risk loans before they default, they will receive an incentive payment of $500 and the mortgage holder will receive a payment of $1,500.250 Fourth, in order to encourage homeowners to stay current, HASP will also provide payments of up to $1,000 per year for five years to be applied directly to principal for homeowners who do not redefault.251 This payment also benefits lenders, as it is a contingent guarantee of a portion of loan principal. Finally, the Treasury is offering a partial guarantee of home prices for modified loans in exchange for insurance payments by the servicer.252 HASP requires three months of successful modified payments before the modification is eligible for the various incentive payments.253 HASP is undoubtedly an improvement over FHASecure and HOPE for Homeowners, and addresses many of these programs deficiencies, but whether HASP will succeed is uncertain. To the extent that servicers participate in HASP, it should have some success in ensuring affordability of modified loans. Whether the incentives to servicers are adequate to offset the financial advantages to pursuing foreclosures is uncertain. Moreover, while the Treasury has conditioned TARP funds on HASP participation, not all servicers are TARP eligible or looking to receive TARP funds. Even if the incentives work, they are objectionable
246. U.S. DEPT OF THE TREASURY, HOME AFFORDABLE MODIFICATION PROGRAM GUIDELINES 1 (Mar. 4, 2009), available at http://www.ustreas.gov/ press/releases/reports/modification_program_guidelines.pdf. 247. Id. at 89. 248. Id. at 1, 11. 249. Id. at 1, 1113. 250. Id. at 1. 251. Id. at 1, 1113. 252. Id. at 14. 253. Id. at 910.
LEVITIN - FINAL
4/23/2009 11:40 AM
638
from a policy perspective as they are essentially bounties or bribes for servicers to do what they are already supposed to domaximize value for MBS certificate holders. Even if servicer participation in HASP is robust, HASP modifications are unlikely to return long-term stability to the housing market because HASP modifications do not address negative equity, which the Congressional Oversight Panel has identified as the single best predictor of mortgage defaults.254 Because of loans originated at high loan-to-value ratios and the falling real-estate market, many homeowners have found themselves with negative equity in their homes.255 Nearly one in five homeowners currently have negative equity in their homes, and if market declines continue, the number can be expected to rise.256 Traditionally, negative equity has not alone led to foreclosures. In regional housing busts, in the past, homeowners with negative equity typically remained in their homes, as long as the mortgage payments remained affordable.257 The likely explanation for this is that relocation can be difficult and expensive, and in spite of the increasing mobility of the American family,258 the home retains a strong emotional center for many.259 Therefore, homeowners who do not have an immediate need to move are unlikely to do so, even with negative equity. In this housing crisis, however, negative equity is deeper than it has ever been in past regional housing busts and the economy overall is worse.260 Data assembled by the Congressional Oversight Panels survey of federal banking and housing regulators, however, indicates that negative equity is a central problem in the current housing crisis.261 Although the Oversight Panel noted that there were significant limitations on its data, it found that negative equity is the single best
available at www.facorelogic.com/newsroom/marketstudies/negative-equity-report.jsp.
(stating that over 7.5 million (or 18 percent of all) mortgages were in a negative-equity position as of September 30, 2008); see also J.W. Elphinstone, Homeowner Equity Hits All-Time Low, Fed Says, N.Y.SUN.COM, Mar. 6, 2008, http://www.nysun.com/ business/homeowner-equity-hits-all-time-low-fed-says/72453. 257. Christopher L. Foote et al., Negative Equity and Foreclosure: Theory and Evidence (Federal Reserve Bank of Boston, Public Policy Discussion Papers No. 08-3, 2008), available at www.bos.frb.org/economic/ppdp/2008/ppdp0803.pdf. 258. U.S. CENSUS BUREAU, GEOGRAPHICAL MOBILITY: 2002 TO 2003, at 2 (Mar. 2004), available at www.census.gov/prod/2004pubs/p20-549.pdf (noting an increasing occurrence of long-distance moves). 259. See generally Radin, supra note 11; Stern, supra note 11. 260. See CONG. OVERSIGHT PANEL, supra note 30, at 26. 261. Id. at 26.
254. 255. 256.
See CONG. OVERSIGHT PANEL, supra note 30, at 2628. David Streitfeld, Coming Up Short, N.Y. TIMES, Sept. 19, 2008, at C1.
FIRST AM. CORELOGIC, NEGATIVE EQUITY DATA REPORT (Sept. 30, 2008),
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
639
indicator that a property is likely to enter foreclosure.262 As the Oversight Panel noted: Given the depth of negative equity and the strained state of many consumers finances generally, it is not surprising that negative equity is a leading indicator of the likelihood of default. When there is only a small level of negative equity and prospects for a recovery of the housing market in the shortterm, a homeowner might reasonably be willing to continue to pay through the negative equity period. Given the slim prospects of the housing market recovering to 20052007 price levels in the near future, some homeowners might begin to question whether they will ever have positive equity in their homes.263 A homeowner with negative equity might hang on to the property and continue making payments. But if mortgage payments are a strain on the homeowners finances or there is an external impetus to move, then the homeowner might decide to abandon the house rather than pay an often nonrecourse mortgage that is for more than the house is worth.264 Families need to relocate all the time for reasons like new jobs, an increase or decrease in household size, such as from the birth of children or divorce or death, or because a household member needs assisted living or a handicapped-accessible house. If a family with negative equity wishes to sell their house, they are faced with a balloon payment that they are unlikely to be able to afford. Walking away is an attractive option for homeowners with negative equity.265 For a homeowner with negative equity, it might take several years, if not a decade, to regain positive equity in the property. In the meantime, these homeowners are functionally renting the homes, in that they are paying money for occupancy rights, but not building up positive
Id. at 2628. Id. at 27. See, e.g., Vikas Bajaj, Mortgage Holders Find It Hard to Walk Away from Their Home, N.Y. TIMES, May 10, 2008, at C1. Even if the mortgage is recourse, lenders frequently do not pursue deficiency judgments after foreclosure. Id. 265. See, e.g., David Streitfeld, Ruins of an American Dream, N.Y. TIMES,
262. 263. 264. Aug. 24, 2008 (describing a family that abandoned a house and $3,400 monthly mortgage payment to foreclosure and began renting a brand new house two miles away for $1,200 per month). The credit-report hit from a foreclosure can generally last no longer than seven years. 15 U.S.C. 1681c(a)(2) (2006). This means that if a homeowner abandoned a negative-equity property to foreclosure, the homeowner would be able to go back in the home-buying market in seven years without a diminished credit score. If the homeowner would be functionally renting for at least that long by remaining in a negative-equity property, then the credit-score hit would likely be outweighed by cheaper rental opportunities if the homeowner abandoned the property.
LEVITIN - FINAL
4/23/2009 11:40 AM
640
equity. If these homeowners can find a better rental deal elsewhere, walking away from the home can be a sensible solution. At the very least, homeowners with negative equity have reduced incentives to take care of their properties; why fix up what is effectively a rental unit? HASP also fails to address contractual restrictions in pooling and servicing agreements; indeed, it explicitly defers to them.266 It also does little to address servicer-capacity, organization, and outreach problems. HASP only obliquely deals with outreach problems by providing funding for nonprofit counseling organizations,267 but it is unclear whether this alone is sufficient. Nor does HASP present a clear method of dealing with junior mortgages, noting only that additional incentives will be provided to extinguish junior liens on homes with first-lien loans that are modified under the program.268 And by subsidizing both borrowers and lenders, HASP creates a moral hazard problem. There is a question of whether an individualized approach to foreclosure mitigation makes sense. Programs like HASP that look to NPV calculations on an individual loan level capture loan-level efficiency for modifications, but they fail to account for the systemic externalities of a certain volume of foreclosures. If the foreclosure crisis is one of a certain volume of foreclosures creating a downward cycle of falling real estate prices and more foreclosures, then an individualized approach to foreclosure mitigation makes little sense; instead, the proper approach would be limiting foreclosures in general, including foreclosures that might be efficient in isolation. Finally, HASP has a definite price tag: $275 billion, consisting of $200 billion in additional federal backstopping of Freddie Mac and Fannie Mae, and $75 billion for modifications and supporting programs.269 HASP will likely be more effective than previous federal programs, but it comes at a steep price. 7. THE ADVANTAGES OF BANKRUPTCY MODIFICATION Compared with all of these options, amending the Bankruptcy Code to permit modification of all mortgages in Chapter 13 would provide a far superior solution. Permitting mortgage modification in Chapter 13 would provide an immediate solution to much of the current home-foreclosure crisis that would not affect the public fisc, create moral hazard, or impose costs on future borrowers.
266. 267. 268. 269.
See U.S. DEPT OF THE TREASURY, supra note 246, at 2. See U.S. DEPT OF THE TEASURY, supra note 244, at 5. See U.S. DEPT OF THE TREASURY, supra note 246, at 15. See U.S. DEPT OF THE TREASURY, supra note 245.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
641
Bankruptcy modification is the only solution other than eminent domain that can deal neatly and cleanly with impediments to loan modification or refinancing, such as restrictions on modifications in servicing contracts, servicer litigation-risk aversion, misaligned servicer incentives, practical obstacles to servicers, and junior mortgages. Bankruptcy removes servicers entirely from the loan-modification process and eliminates junior mortgagees holdup ability. Bankruptcy modification would also provide a solution for both of the distinct mortgage crisesnegative equity and payment shockwhile not creating a windfall for speculators. While Federal Reserve Bank Chairman Ben Bernanke has called for lenders to voluntarily write down principal on negative-equity mortgages,270 this type of voluntary modification has generally not occurred. Bankruptcy modification would help negative-equity homeowners by eliminating their negativeequity position in strip-down, which would reduce their incentives to abandon the properties. Other homeowners are unable to afford their mortgages because of a rate reset due to the expiration of a teaser rate, the resetting of an ARM, or the reamortization of a pay-option ARM.271 Many of these homeowners have negative equity or at least an underwater second mortgage, both of which make refinancing impossible.272 For these homeowners, bankruptcy modification also offers a solution through the lowering of the monthly payment to an affordable level. Modification of mortgages in Chapter 13 bankruptcies would not create a moral-hazard problem for lenders or borrowers. For starters, permitting bankruptcy modification of mortgages in order to prevent inefficient foreclosures fits into a well-recognized exception to moral hazard, that for contagion fires.273 It would create a moral hazard for the fire department to rescue people from fires caused by smoking in bed, yet we rescue in-bed smokers without hesitation, in part because fires can spread and harm third parties, like neighbors.274 Foreclosures function like fires, and a rash of foreclosures can destroy property
270. Ben S. Bernanke, Chairman, Fed. Reserve Bank, Reducing Preventable Mortgage Foreclosures, Speech at the Independent Community Bankers of America Annual Convention (Mar. 4, 2008), available at http://www.federalreserve.gov/ newsevents/speech/bernanke20080304a.htm. 271. See CONG. OVERSIGHT PANEL, supra note 30, at 1823. 272. Nick Ravo, The Trials of Ownership in a Recession, N.Y. TIMES, Aug. 8, 1993, at I10; see also Holden Lewis, Some Lenders Blocking Refinances if Second Mortgage Isnt Paid Off, SEATTLE POST-INTELLIGENCER, Feb. 27, 2009, available at http://seattlepi.nwsource.com/money/401802_real28.html. 273. Larry Summers, Beware Moral Hazard Fundamentalists, FIN. TIMES (London), Sept. 23, 2007. 274. See id.
LEVITIN - FINAL
4/23/2009 11:40 AM
642
values throughout a neighborhood. Moral-hazard concerns are inapplicable to bankruptcy modification of mortgages given the immense third-party costs of foreclosures. Irrespective, the Bankruptcy Code already has powerful antidotes to moral-hazard risk. Lenders would still incur losses in bankruptcy, even if they would generally be smaller than in foreclosure. Lender losses in bankruptcy will check lender moral hazard. While permitting modification of mortgages in Chapter 13 would likely result in more bankruptcy filings, this should not be interpreted as abusive or improper behavior. There is no reason to believe that most filers would be anything other than homeowners trying to save their homes, rather than strategic borrowers who can afford their current mortgage but just want a better deal. Debtors are unlikely to receive a windfall from Chapter 13 modification. Strip-down would only result in the debtor having zero equity in the property, not positive equity. Given the large transaction costs to a sale, debtors are unlikely to sell their properties for anything beyond a de minimis profit absent a remarkable recovery of the housing market.275 And debtors are not guaranteed any particular modification;
275. It is also important to note that bankruptcy modification reducing loan principal does not produce a windfall for a debtor, even if the property later appreciates. The debtor cannot benefit from the appreciation during the course of the plan. If the mortgage appreciates in the three to five years of a plan, the debtor can only benefit upon a sale or disposition of the house. If the debtor sells the house at an appreciated value during the term of the plan, the debtors income from the sale will be available to satisfy unsecured claims, including any unsecured mortgage claim that results from bifurcation under section 506(a). 11 U.S.C. 1325(b) (2006) (requiring debtors to commit all disposable income to unsecured creditors); id. 1329 (permitting modification of a plan to account for increases in debtors income). Thus, there is no windfall possible for the debtor in the short term. If the property appreciates in the long term, that appreciation would belong to the debtor, but the debtor has a better claim to it than the mortgagee. Seen from a perspective of the original loan, letting the debtor keep future appreciation looks like a windfall. But this is the wrong perspective. The original loan was unable to perform, and insisting on its terms would have resulted in foreclosure. When a property is sold in foreclosure, the foreclosing creditor does not receive the future appreciation on the property; that belongs to the foreclosure-sale purchaser. Giving the creditor more in bankruptcy than the creditor would have received in a foreclosure is a windfall to the creditor, not the debtor. The creditor has already been rewarded in bankruptcy by getting a loan modification that will provide at least the value the creditor would have received in foreclosure. If the creditor were able to claw back future appreciation, the bankruptcy modification would be equivalent to a temporary loan modification, and temporary modifications are less likely to succeed than life-of-the-loan modifications. In the case of securitized loans, permitting an appreciation claw-back would also reward precisely the parties whose irresponsible behavior created the foreclosure crisis. Securitization trusts are often short-lived entities. When the outstanding principal balance reaches a certain threshold, often 10 percent, the servicer will exercise a clean-up call and purchase out the remaining balance from the trust; it is not
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
643
the modification must be part of a plan, which imposes various other requirements on debtors. Chapter 13 bankruptcy also protects against moral hazard from borrowers. Chapter 13 is not a drive-by process. In order to receive a discharge in Chapter 13, a debtor must bear her finances to the public, incur a variety of monetary and temporal transaction costs, and live on a court-supervised, means-tested budget for three or five years.276 Having to get the court and the Chapter 13 trustee to sign off on the reasonableness of daily expenses creates a powerful disincentive against filing for bankruptcy unless the filing is absolutely necessary. Moreover, Chapter 13 insists on full repayment of certain debts, including allowed secured claims, domestic support obligations, and tax liabilities.277 A below-median-income debtor who does not repay creditors in full can only receive a Chapter 13 discharge once every six years; an above-median-income debtor who does not repay creditors in full can only receive a Chapter 13 discharge once every ten years.278 This means that the minimum time between repeat Chapter 13 filings is longer than the time a foreclosure stays on a credit report.
economical for the servicer to service small balances. KOTHARI, supra note 59, at 43. Most trusts reach this clean-up-call threshold in their first seven years, as loans are refinanced out of the trust or default. See Phoa, supra note 42, at 363. Thus, the trust that owned the mortgage at the time of bankruptcy may well not exist to receive the shared appreciation. Instead, the clawed-back appreciation would accrue to the party who held the residual rights in the mortgagesoften the servicer/originator. This is particularly troubling because, in many cases, principal reductions are necessary because the original lender condoned or even encouraged inflated property appraisals in order to make larger loans that it could then securitize for more money. See Homeowners Sue KB Home and Countrywide over Appraisals, REUTERS.COM, Feb. 7, 2008, http://www.reuters.com/article/domesticNews/idUSN07417124200 80208. Thus, rather than being a windfall to debtors, an appreciation claw-back would reward the very entities that fueled the mortgage bubble through irresponsible lending. Finally, it is important to emphasize that appreciation claw-backs do not exist for any other sort of lien-stripping in bankruptcy. Likewise, unsecured creditors do not get to claim future income or assets after the debtor is discharged. Even if the debtor wins the lottery the next day, the core bankruptcy policy of the fresh start emphasizes that prepetition creditors have no claim on postdischarge assets. 11 U.S.C. 727(b), 1328(a). 276. 11 U.S.C. 1325(b). 277. Id. 1322(a), 1325(a)(5). 278. See id. 1328(f)(2). Section 1328 prohibits a Chapter 13 discharge if a Chapter 13 discharge was granted within the two preceding years, but for debtors who do not repay creditors in full, a Chapter 13 plan must last at least three or five years, depending on whether the debtor is below or above the applicable states median income. Id. 1325(b)(1), (4). Thus, it is the length of the plan, not the time between discharges, that controls for debtors who have to repay less than 100 percent of their debts.
LEVITIN - FINAL
4/23/2009 11:40 AM
644
Historically, concerns about anything more than isolated serial and strategic filings are greatly overstated and unsupported by empirical evidence. The concern over moral hazard in bankruptcy appears to be more an economists fantasy than an empirically grounded reflection of real Americans behavior. Most bankruptcy filers are seriously financially strapped and bankruptcy is a last-resort option.279 They are confused, ashamed, and unhappy. Americans do not behave as strategically with bankruptcy as economists assume people act. All things being equal, economists seem far more likely to file for bankruptcy than actual consumers. While there are undoubtedly some debtors who abuse bankruptcy, there is no evidence suggesting that abusive debtors are anything other than a small minority. And for those cases, there are numerous safeguards built into the system to discourage strategic use of the bankruptcy system, such as the good-faith catch-all for Chapter 13 filings and Chapter 13 plans.280 If those safeguards were insufficient, Congress could legislate eligibility requirements keyed to affordability or a foreclosure filing. Permitting modification of mortgages in Chapter 13 bankruptcy would be unlikely to result in wealthy or spendthrift debtors receiving unmerited relief. Traditionally, wealthy debtors rarely file for bankruptcy. The mean income of Chapter 13 bankruptcy filers in 2007 was $35,688,281 and less than 10 percent of all debtors earn over $60,000.282 Indeed, wealthy debtors with significant secured debt are not eligible for Chapter 13. To file for Chapter 13, an individual must have less than $336,900 in noncontingent, liquidated, unsecured debts and less than $1,010,650 in noncontingent, liquidated, secured debts.283 This means that a homeowner with a million-dollar mortgage cannot avail himself of Chapter 13. Instead, if that homeowner wishes to keep his mansion, he must file for Chapter 11 bankruptcy, where creditor controls are much stronger because creditors must vote to approve a plan.284 And, if a high-income debtor has low enough secured and unsecured debt levels to be eligible for Chapter 13, Chapter 13 would be of limited benefit as the debtor would be required to pay all of his disposable incomeas determined by a statutory testto unsecured
279. Robert M. Lawless et al., Did Bankruptcy Reform Fail? An Empirical Study of Consumer Debtors, 82 AM. BANKR. L.J. 349, 376, 381 (2008) (suggesting that those who file for bankruptcy are seriously in debt and wait until it is absolutely necessary to file before doing so). 280. 11 U.S.C. 1325(a)(3), (7). 281. Lawless, supra note 279, at 361. 282. Id. at 360 fig.2. 283. 11 U.S.C. 109(e) (2006 & Supp. 2008). 284. 11 U.S.C. 1129 (2006); see also id. 1111(b).
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
645
creditors,285 including any unsecured mortgage claim resulting from claim bifurcation under section 506. Speculators, too, are unlikely to benefit from bankruptcy modification. The secured-debt limit will keep many out of Chapter 13. The parts of the country where there has been the most real-estate speculation are also the parts of the country with the highest home prices. In California, where the average loan amount is $331,926,286 three of these mortgages plus a $15,000 car loan would make a debtor ineligible for Chapter 13. Thus, a speculator with a fairly average car, a mortgage on his own home, and two investment properties would not be eligible for Chapter 13 bankruptcy. Even if the speculator is eligible for Chapter 13, he is unlikely to be able to retain his investment properties, much less modify the mortgages thereon. A mortgage-loan modification in bankruptcy can occur only as part of a plan.287 The automatic stay would likely be lifted on an investment property (or second home) before a plan could be confirmed.288 The automatic stay must be lifted either if the property is underwater and not necessary for an effective reorganization,289 or for cause, including lack of adequate protection.290 Strip-down is only useful for underwater properties, and unless the debtors business is being a small-time landlord, the property is not necessary for an effective reorganization. The areas that have been hardest hit by the decline in housing prices are areas where there had been price run-ups fueled by speculation.291 These are the parts of the country where investor properties are most likely to be underwater, and where the mortgagee would most likely be able to have the stay lifted.
285. Id. 1325(b)(1)(B). 286. See Mortgagebankers.org, Stop the Bankruptcy Cram Down Resource Center, http://www.mortgagebankers.org/StopTheCramDown (last visited Jan. 27, 2009). 287. 11 U.S.C. 1322(b). 288. Id. 362(d). The Bankruptcy Code provides that the automatic stay shall be lifted for cause, including either lack of adequate protection of a secured creditors interest in the propertythat is payments to compensate the secured creditor for depreciation in its collateral during the bankruptcyor if the debtor does not have equity in the property and the property is not necessary for an effective reorganization. Id. Thus, debtors with positive equity who could not handle mortgage payments prepetition would be unlikely to be able to make the adequate-protection payments necessary to prevent the lifting of the stay. Id. 362(d)(1). Debtors with negative equity would find the stay lifted because investment properties and second homes are not essential to their reorganizations. Id. 362(d)(2). 289. Id. 362(d)(2). 290. Id. 362(d)(1). 291. See CONG. OVERSIGHT PANEL, supra note 30, at 2123.
LEVITIN - FINAL
4/23/2009 11:40 AM
646
Even if the debtor has equity in the property or needs it for an effective reorganization, the debtor will have to provide adequateprotection payments unless there is a significant equity cushion.292 Adequate protection is supposed to protect the creditor against depreciation in the collateral, but is often equivalent to rent or mortgage payments. Speculators typically either cannot or will not make these payments, which are the buy-in to modifying the mortgage. As a result, the stay will be lifted, and the mortgagee is free to foreclose. If the speculator were able to avoid the lifting of the automatic stay, the loan could only be modified as part of a repayment plan proposed by the debtor, which would have to be confirmed by the court.293 Plan confirmation might not be possible because of a goodfaith294 or a disposable-income objection.295 Creditors could argue that it is not good faith for a debtor to keep an investment (and keep building up equity in the investment) when they are not getting paid in full. Likewise, unsecured creditors could argue that the debtor is not paying all disposable income to them if they are instead paying the investmentproperty mortgagee. The Bankruptcy Code is replete with provisions to protect against abuse by small-time real-estate speculators, and it is extremely unlikely that a speculator would be able to take advantage of bankruptcy modification. Even if the speculator were able to avoid the lifting of the automatic stay, plan confirmation might not be possible because of an objection that a plan that allows a debtor to retain an investment property and continue building up equity in the investment by making mortgage payments either lacks good faith or demonstrates that the debtor is not paying unsecured creditors all disposable income.296 Accordingly, speculators and homeowners intent on keeping their second homes are unlikely to file for bankruptcy to seek mortgage modification in the first place. Permitting bankruptcy modification of
292. 11 U.S.C. 362(d). 293. Id. 1321, 1325 (2006 & Supp. 2008); 11 U.S.C. 1322(b) (2006). 294. Id. 1325(a)(3), (7). 295. Id. 1325(b). 296. Id. 1325(a)(7) (describing the good-faith requirement); id. 1325(b)(1)(B) (describing the disposable-income requirement). Most courts, however, permit debtors to continue making contributions to 401(k) and 403(b) plans. See, e.g., In re Mati, 390 B.R. 11 (Bankr. D. Mass. 2008). On the other hand, this can be differentiated on tax-benefit grounds; whereas there is a tax-benefit from the 401(k) or 403(b) contribution that would otherwise be lost, there is no tax deduction available for an investment-property-mortgage interest payment (although there is for second homes). Chapter 13 does seem to recognize tax-benefits as an issue in at least some situations; small-business owners are able to deduct business expenses from their disposable income as part of the disposable income test. Id. 1325(b)(2)(B).
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
647
primary-home mortgages thus steers a true course between extending the right sort of relief and not extending it too broadly. Permitting modification of all mortgages in bankruptcy would not have prevented the economic situation leading to the foreclosure crisis, although it would have encouraged more prudent and sustainable underwriting and limited some of the systemic risk involved.297 Bankruptcy modification is not a magic-bullet solution, but it is a quick, fair, efficient, and administrable response that would be far more effective at stabilizing the housing market and preventing the deadweight social and economic losses of foreclosure than any other proposed solution.298 CONCLUSION A wide range of empirical data shows that permitting bankruptcy modification of all mortgages would have little or no impact on mortgage credit cost or availability. Because lenders face smaller losses from bankruptcy modification than from foreclosure, the market is unlikely to price against bankruptcy modification. This finding neuters the hitherto untested policy assumption underlying the special protection for mortgages in Chapter 13, namely that permitting modification would result in mortgage credit constriction. In light of this market neutrality, permitting modification of all home mortgages in bankruptcy stands out as the best of all possible solutions proposed to the mortgage crisis. Unlike any other proposed response, bankruptcy modification offers immediate relief, solves the market problems created by securitization, addresses both problems of
297. Adam J. Levitin, Helping Homeowners: Modification of Mortgages in Bankruptcy, 3 HARV. L. & POLY REV. ONLINE 1, 79 (2009), available at http://www.hlpronline.com/Levitin_HLPR_011909.pdf. 298. Arguably, states might hold the key to bankruptcy relief. The Bankruptcy Code prohibits the modification of a claim secured only by a security interest in real property that is the debtors principal residence. 11 U.S.C. 1322(b)(2). As several courts of appeals have held, this means that if the security interest includes anything other than the debtors principal residence, then the mortgage can be modified. See supra note 41 (listing examples). Bankruptcy law looks to state law to determine property rights. Butner v. United States, 440 U.S. 48, 49, 5657 (1979). So, if the security interest covers a rental unit on the property or connected farmland or fixtures, then modification is possible. States could redefine property rights to clarify that a mortgage secured by a debtors principal residence is not secured solely by the debtors principal residence if it includes a security interest in land (not just the house), fixtures, rents, profits, etc. Thus, states might be able to make it possible for a large number of homeowners to file for Chapter 13 bankruptcy and modify the mortgage on their homes simply by fine-tuning state property law.
LEVITIN - FINAL
4/23/2009 11:40 AM
648
payment-reset shock and negative equity, screens out speculators, spreads burdens between borrowers and lenders, and avoids both the costs and moral hazard of a government bailout. As the foreclosure crisis deepens, bankruptcy modification presents the best and least invasive method of stabilizing the housing market. Beyond the immediate policy questions of whether to allow modification of mortgages in bankruptcy and how best to resolve the foreclosure crisis, however, this Articles findings also speak to the fundamental policy debate in consumer-bankruptcy lawthe balance between providing debtors with a fresh start and limiting losses to creditors. Bankruptcy policy is concerned with limiting losses to creditors both because of the impact on the creditors themselves and because of the systemic impact on the credit market. The basic economic (as opposed to moral) assumption of creditors rights has been that creditors compensate for bankruptcy losses by raising rates and restricting lending. In a traditional world of relational portfolio lending, this argument makes sense theoretically and accords with the available evidence. But the bulk of consumer finance no longer operates in this fashion. Consumer finance has become a highly impersonalized, nonrelational business; it is no longer a parallel to Main Street finance. And, as this Article shows, at least in the mortgage industry, limiting creditor losses in bankruptcy does not translate into lower credit costs or greater credit availability. This Article is part of a growing body of research299 that suggests the need for a new theory of consumer finance: because of diversification among millions of borrowers, risk-spreading through securitization and insurance, and fee-based profit models, the scope of the bankruptcy discharge has very little impact on the price or availability of credit except at the margins. If this theory is correct, then we must both update our thinking about the effect of bankruptcy law on consumer finance and rethink consumer-bankruptcy policy from the ground up, with an eye to expanding the scope of the discharge. As consumer finance becomes more complex, it is time to update the model of the effect of bankruptcy law on consumer finance. The theory and modeling of consumer finance can serve as a meaningful policy guide, but to do so, it must account for the actual structure of the evolving consumer-finance industry.
299. See, e.g., KANTROWITZ, supra note 128; Link, supra note 125; Ronald J. Mann, Bankruptcy Reform and the Sweatbox of Credit Card Debt, 2007 U. ILL. L. REV. 375 (2007); Simkovic, supra note 127.
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Foreclosure Crisis
POSTSCRIPT
649
At the time this Article went to press, Congress was considering legislation to permit modification of single-family, principal-residence mortgages in Chapter 13 bankruptcy. On March 5, 2009, the Helping Families Save Their Homes Act of 2009 passed the House of Representatives;300 a companion bill, the Helping Families Save Their Homes in Bankruptcy Act of 2009301 is still in committee in the Senate. The Obama administration has endorsed bankruptcy modification with some qualifications,302 but it is uncertain whether the legislation will gain the requisite support to pass the Senate absent compromises that would seriously reduce its effectiveness for dealing with the foreclosure crisis; in 2008 similar legislation passed the House, and was reported out of the Senate Judiciary Committee, but never came to a floor vote
300. H.R. 1106, 111th Cong (2009). 301. S. 61, 111th Cong. (2009). 302. The Homeowner Affordability and Stability Program calls for modification of mortgages in bankruptcy as a backstop to voluntary modifications; the Obama administration has limited its call for bankruptcy modification to mortgages within the Fannie Mae/Freddie Mac conforming loan limit, so that millionaire homes dont clog the bankruptcy courts. See U.S. DEPT OF THE TREASURY, supra note 244, at 6. It is unclear if this proposed limitation is for loans that were conforming at the time of origination or at the time of the bankruptcy filing; the former would effectively bar many Californians and New Yorkers from bankruptcy relief given the high percentage of jumbos in those states. Despite the administrations statement that the proposed limitation is meant to prevent bankruptcy abuse by millionaires, a more likely explanation is a concern about bankruptcy-loss allocations in jumbo-mortgage securitizations that could further threaten the solvency of financial institutions. Many jumbo-mortgage securitizations have a pari passu, rather than senior/subordinate loss allocation for excess bankruptcy losses, that is losses due to bankruptcy that are over and above a de minimis amount for the entire trust. Maurna Desmond & Daniel Fisher, The Mortgage Investors Protection Act of 2009, FORBES.COM, Mar. 19, 2009, http://www.forbes.com/ 2009/03/17/congress-foreclosures-bankruptcy-cramdowns-business-wall-streetcramdowns.html. A pari passu loss allocation would impose losses even on the senior most MBS in a jumbo securitization, which would cause the senior bonds to lose their AAA rating. Id. Financial institutions holding the AAA bonds would have to carry the bonds as assets at a lower value and/or sell them immediately, at a reduced value. Id. In either case, it would result in losses to already strained financial institutions. Id. The bankruptcy-modification legislation that passed the House of Representatives (H.R. 1106, the Helping Families Save Their Homes Act of 2009) has a provision specifically designed to address this concern. Helping Families Save Their Homes in Bankruptcy Act of 2009, H.R. 1106, 111th Cong. The provision provides that contractual lossallocation provisions for excess bankruptcy losses apply, as a matter of public policy, only to bankruptcy losses of the type that could have occurred before the passage of the bill. Id. 124.
LEVITIN - FINAL
4/23/2009 11:40 AM
650
because of the inability to get cloture and avoid a filibuster.303 Hopefully, the changes advocated by this Article will have become law by the time these words are read. Bankruptcy modification alone will not solve the entirety of the foreclosure crisis, much less the entire economic downturn, but it is a critical element of stabilizing the housing market and the economy.
303. This legislation was the Helping Families Save Their Homes in Bankruptcy Act of 2008, S. 2136, 110th Cong. (2008).
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
Resolving the Foreclosure Crisis
651
APPENDIX A: MORTGAGE-ORIGINATION RATE QUOTES (SELECTED) Table A1: IndyMac Rate Quotes on Jan. 27, 2008 (20 Percent Down, Conforming)
Single Family Primary Residence State of Property Lender Credit Score Conforming? Property Value Loan Amount Loan Term Interest Rate Points APR CA IndyMac 560 Y $400,000.00 $320,000.00 30-Fixed 7.250% 0.858% 7.414% 2-Family Primary Residence CA IndyMac 560 Y $400,000.00 $320,000.00 30-Fixed 7.250% 0.858% 7.414% 3-Family Primary Residence CA IndyMac 560 Y $400,000.00 $320,000.00 30-Fixed 7.250% 0.858% 7.414% 4-Family Primary Residence CA IndyMac 560 Y $400,000.00 $320,000.00 30-Fixed 7.250% 0.808% 7.383% Vacation Home or Second Home CA IndyMac 560 Y $400,000.00 $320,000.00 30-Fixed 7.250% 0.858% 7.414% Investor or Rental Property CA IndyMac 560 Y $400,000.00 $320,000.00 30-Fixed 8.250% 0.711% 8.408%
Table A2: IndyMac Rate Quotes on Jan. 27, 2008 (20 Percent Down, Conforming)
Single Family Primary Residence State of Property Lender Credit Score Conforming? Property Value Loan Amount Loan Term Interest Rate Points APR CA IndyMac 660 Y $400,000.00 $320,000.00 30-Fixed 5.750% 0.868% 5.900% 2-Family Primary Residence CA IndyMac 660 Y $400,000.00 $320,000.00 30-Fixed 5.750% 0.868% 5.900% 3-Family Primary Residence CA IndyMac 660 Y $400,000.00 $320,000.00 30-Fixed 5.750% 0.868% 5.900% 4-Family Primary Residence CA IndyMac 660 Y $400,000.00 $320,000.00 30-Fixed 5.750% 0.868% 5.900% Vacation Home or Second Home CA IndyMac 660 Y $400,000.00 $320,000.00 30-Fixed 5.750% 0.868% 5.900% Investor or Rental Property CA IndyMac 660 Y $400,000.00 $320,000.00 30-Fixed 6.750% 1.045% 6.928%
LEVITIN - FINAL
4/23/2009 11:40 AM
652
Single Family Primary Residence State of Property Lender Credit Score Conforming? Property Value Loan Amount Loan Term Interest Rate Points APR CA IndyMac 760 Y $400,000.00 $320,000.00 30-Fixed 5.375% 1.545% 5.585% Vacation Home or Second Home CA IndyMac 760 Y $400,000.00 $320,000.00 30-Fixed 5.375% 1.545% 5.585%
2-Family Primary Residence CA IndyMac 760 Y $400,000.00 $320,000.00 30-Fixed 5.375% 1.545% 5.585%
Investor or Rental Property CA IndyMac 760 Y $400,000.00 $320,000.00 30-Fixed 6.500% 0.771% 6.648%
Single Family Primary Residence State of Property Lender Credit Score Conforming? Property Value Loan Amount Loan Term Interest Rate Points APR CA IndyMac 660 C $356,000.00 $320,000.00 30-Fixed 5.750% 0.868% 6.227% 2-Family Primary Residence CA IndyMac 660 C $356,000.00 $320,000.00 30-Fixed 5.750% 0.868% 6.227% 3-Family Primary Residence CA IndyMac 660 C $356,000.00 $320,000.00 30-Fixed 7.625% 1.250% 8.223% 4-Family Primary Residence CA IndyMac 660 C $356,000.00 $320,000.00 30-Fixed 7.625% 1.250% 8.223% Vacation Home or Second Home CA IndyMac 660 C $356,000.00 $320,000.00 30-Fixed 5.750% 0.868% 6.317% Investor or Rental Property CA IndyMac 660 C $356,000.00 $320,000.00 30-Fixed 6.875% 1.301% 8.241%
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
653
Single Family Primary Residence State of Property Lender Credit Score Conforming? Property Value Loan Amount Loan Term Interest Rate Points APR CA IndyMac 760 C $356,000.00 $320,000.00 30-Fixed 5.375% 1.545% 5.898% 2-Family Primary Residence CA IndyMac 760 C $356,000.00 $320,000.00 30-Fixed 5.375% 1.545% 5.898% 3-Family Primary Residence CA IndyMac 760 C $356,000.00 $320,000.00 30-Fixed 7.500% 1.250% 8.386% 4-Family Primary Residence CA IndyMac 760 C $356,000.00 $320,000.00 30-Fixed 7.500% 1.250% 8.386% Vacation Home or Second Home CA IndyMac 760 C $356,000.00 $320,000.00 30-Fixed 5.375% 1.545% 5.985% Investor or Rental Property CA IndyMac 760 C $356,000.00 $320,000.00 30-Fixed 6.750% 0.745% 7.532%
LEVITIN - FINAL
4/23/2009 11:40 AM
654
APPENDIX B: EVALUATING THE MORTGAGE BANKERS ASSOCIATIONS MODIFICATION-IMPACT CLAIM The Mortgage Bankers Association (MBA) has claimed that permitting modification of mortgages in bankruptcy will result in an effective 200 basis-point increase in interest rates on single-family, owner-occupied properties (principal residences).304 The MBA figure is derived from a comparison of the current interest-rate spread between mortgages on single-family principal residences and on investor properties.305 It includes not only the current additional interest-rate premium for investor properties of 37.5 basis points, but also amortizes the higher down payments and points generally required on investor properties in order to achieve the 200 basis-point figure.306 More recent MBA press releases have claimed only an increase of 150 basis points, without explaining the 50 basis-point decline from the 200 basis-point figure featured in Congressional testimony.307 The MBA figure is based on an assumption that the entire spread between principal-residence and investor-property mortgage interest rates is due to lack of modification protection on investor properties.308 Current mortgage interest-rate spreads among different property types disprove the MBAs claim. The MBAs calculation is based on looking selectively at the effective interest-rate spread between investment properties and single-family principal residences.309 But
304. See Kittle Testimony, supra note 93. Notably, in response to a request from U.S. Representative Brad Miller (D-N.C.), for clarification in later communications with members of Congress, the MBA changed its explanation of the 150-basis-point-increased-cost-of-mortgages claim, arguing (without providing any evidence or methodology for the derivation of its numbers) that seventy to eighty-five basis points would be due to higher default-incidence rates, twenty to twenty-five basis points would be due to higher loss-severity rates, ten basis points would be due to the administrative costs imposed by bankruptcy, and fifty to sixty basis points would be due to market uncertainty and increased political risk. Letter from Stephen A. OConnor, Senior Vice President of Govt Affairs, Mortgage Bankers Assn, to Representative Brad Miller (Apr. 18, 2008) (on file with author). 305. See Letter from Stephen A. OConnor, supra note 304. 306. Id. 307. Press Release, Mortgage Bankers Association, MBAs Stop the Cram Down Resource Center Puts a Price Tag on Bankruptcy Reform (Jan. 15, 2008), available at http://www.mortgagebankers.org/NewsandMedia/PressCenter/593 43.htm. 308. Id. 309. David Kittle, the president-elect of the MBA, claimed that prior to the enactment of the Bankruptcy Code there was no difference in interest rates for singlefamily, owner-occupied principal residences and investor properties. Growing
Mortgage Foreclosure Crisis: Identifying Solutions and Dispelling Myths: Hearing on H.R. 3609 Before the Subcomm. on Commercial and Admin. Law of the H. Comm. on the Judiciary, 110th Cong. (2007) (statement of David G. Kittle, Mortgage Bankers Association), available at http://judiciary.house.gov/hearings/pdf/Kittle080129.pdf. The
LEVITIN - FINAL
4/23/2009 11:40 AM
2009:565
655
mortgages on investor properties are not the only type of property that can currently be modified in bankruptcy; mortgages on vacation homes and multifamily residences in which the owner occupies one unit can also be modified.310 As noted above, conforming mortgages on vacation homes and multifamily properties are currently priced the same as single-family principal residences.311 Only investor-property mortgages are priced higher.312 This pattern is confirmed by PMI rates and Fannie and Freddie delivery fees. This means higher interest rates on investor properties must be attributed to nonbankruptcy risk factors entailed in lending against an investor property. The MBA figure is thus the result of a cherry-picked comparison.313 Likewise, if the historical experiment on strip-down provides a reasonable basis for extrapolating to the current market, and we believe it provides general parameters, then statistically there is a 0 percent chance that the MBAs 150-basis-point claim is correct. All empirical and market observational data indicate that the MBAs claim of an effective 150- to 200-basis-point increase from allowing stripdown is groundless. The empirical evidence indicates that there is unlikely to be anything more than a de minimis effect on interest rates as a result of permitting bankruptcy modification.
MBA has produced no data or other source to support this assertion, including in response to inquiries from major media outlets, and I know of no data source on interest rates that both goes back to 1978 with rates broken down by property type. Indeed, the idea that investor properties and owner-occupied properties would ever be priced the same, even if there were no bankruptcy system whatsoever, ignores the significant default risk entailed in lending against investor properties caused by various tenancy risks. 310 See supra App. A; see also supra Part II.A.2. 311. See supra App. A; see also supra Part II.A.2. 312. See supra App. A; see also supra Part II.A.2. 313. Additionally, the MBAs amortization of the higher down payments typically required on investor properties is debatable. Lenders bear no risk on down payments, unlike on interest payments. Down payments receive different tax treatment than interest payments for borrowers, and down payments create equity in a house, unlike interest. By amortizing down paymentsturning them into interest dollar-fordollar adjusted for present valuethe MBA is equating two very different types of payments that should not be treated as dollar-for-dollar equivalents. Regardless, even if the MBA were correct that higher down payments and/or points will be required, and that it will be harder to make high-LTV loans, this is not necessarily a bad thing, as it might compel more prudent lending practices and would inherently protect lenders from ending up with undersecured loans that could be stripped down by creating an instant equity cushion.

Resolving The Foreclosure Crisis-Modification of Mortgages in Bankruptcy

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Resolving The Foreclosure Crisis-Modification of Mortgages in Bankruptcy

Uploaded by

Copyright:

Available Formats

LEVITIN - FINAL

RESOLVING THE FORECLOSURE CRISIS: MODIFICATION OF MORTGAGES IN BANKRUPTCY

Electronic copy available at: http://ssrn.com/abstract=1071931

WISCONSIN LAW REVIEW

Electronic copy available at: http://ssrn.com/abstract=1071931

Electronic copy available at: http://ssrn.com/abstract=1071931

WISCONSIN LAW REVIEW

3.5% 3.0% 2.5% 2.0% 1.5% 1.0% 0.5% 0.0%

WISCONSIN LAW REVIEW

WISCONSIN LAW REVIEW

WISCONSIN LAW REVIEW

WISCONSIN LAW REVIEW

WISCONSIN LAW REVIEW

A. Treatment of Mortgages in Bankruptcy

WISCONSIN LAW REVIEW

WISCONSIN LAW REVIEW

B. Structure of the Modern Mortgage Market

WISCONSIN LAW REVIEW

WISCONSIN LAW REVIEW

A. Mortgage-Interest-Rate Variation by Property Type

WISCONSIN LAW REVIEW

WISCONSIN LAW REVIEW

WISCONSIN LAW REVIEW

B. Private-Mortgage-Insurance Rate Premiums

WISCONSIN LAW REVIEW

WISCONSIN LAW REVIEW

C. Secondary-Market-Pricing Variation by Property Type

WISCONSIN LAW REVIEW

D. Historical Impact of Permitting Strip-Down

111. 112. 113.

Id. Id. Id.

WISCONSIN LAW REVIEW

WISCONSIN LAW REVIEW

A. The Baseline for Loss Comparison: Foreclosure Sales

WISCONSIN LAW REVIEW

WISCONSIN LAW REVIEW

B. Projecting Lender Losses in Bankruptcy

See CRAIG FOCARDI, SERVICING DEFAULT MANAGEMENT: AN OVERVIEW

OF THE PROCESS AND UNDERLYING TECHNOLOGY

WISCONSIN LAW REVIEW

Chapter 7 22% 4% 74%

Chapter 13 29% 4% 67%

WISCONSIN LAW REVIEW

Table 3: Mortgages Claims in Chapter 13 (2001 CBP)169

2001 Consumer Bankruptcy Project Database.

WISCONSIN LAW REVIEW

984 720 1,094 720 347 25

WISCONSIN LAW REVIEW

Table 5: Mortgages in Chapter 13 (2007 Riverside)176

See Pacer.psc.uscourts.gov, supra note 174.

2007 Riverside Total

2007 Riverside 1st Lien

2007 Riverside 2nd Lien

Undersecured Mortgages Oversecured Mortgages All Mortgages

-21% 37% 21%

-10% 34% 22%

-84% 116% 11%

-23% 34% 23%

-11% 30% 23%

-57% 80% 22%

400 CBP 2001

WISCONSIN LAW REVIEW