rewriting contracts, wholesale data on voluntary mortgage modifications from 2007 and 2008...

7/31/2019 Rewriting Contracts, Wholesale Data on Voluntary Mortgage Modifications From 2007 and 2008 Remittance Reports

1/32

REWRITING CONTRACTS, WHOLESALE: DATA ONVOLUNTARY MORTGAGE MODIFICATIONS FROM 2007

AND 2008 REMITTANCE REPORTS

Alan M. White*

Abstract

The 2007 subprime mortgage crisis resulted in home foreclosures at

unprecedented levels, and calls for government action. The Bush

Administrations response relied primarily on exhorting the mortgageindustry to voluntarily modify the terms of existing mortgages to help

struggling homeowners reduce their debt burden and bring mortgage

debt in line with declining home values. Industry reports have tallied the

rapid increase in voluntarily negotiated workouts, without specifying

what the terms of those workouts have been.

To better understand the effectiveness of the voluntary mortgage

crisis resolution plan, this paper reports detailed empirical information

from a newly-created database. Loan-level information on the numberand type of negotiated mortgage modifications was compiled from

monthly servicer remittance reports from July 2007 through June 2008

for twenty-six mortgage loan pools.

The data show that while the number of modifications rose rapidly

during the crisis, mortgage modifications in the aggregate are not

reducing subprime mortgage debt. Mortgage modifications rarely if ever

reduced principal debt, and in many cases increased the debt. Nor are

modification agreements uniformly reducing payment burdens on

households. About half of all loan modifications resulted in a reduced

monthly payment, while many modifications actually increased the

monthly payment Finally there is tremendous variation in the extent if


2/32

monthly payment Finally there is tremendous variation in the extent if

2 [AUGUST2008]

will be worked out only over many years through painstakingly slow

repayment, foreclosure and disposition of properties.

I. INTRODUCTION

The 2007-08 credit crisis in the United States, precipitated bydefaults on subprime mortgages, resulted from a classic debtbubble, featuring massive borrowing on the basis of rapidlyinflating asset values, in this case residential real estate. Debtcrises can be resolved quickly or slowly. Crisis resolutioninvariably requires that the underlying asset values return tonormal levels and that debt which will never be paid off byborrowers be written off, with or without massive transfers of theunderlying assets.1 The revaluation of assets and write-down ofdebt can be accomplished in several ways, including governmentbailouts, currency inflation (which shifts losses from borrowers tolenders), legislation, asset seizures by lenders and subsequentresales, or by the slow workings of the market.2 The losses can be

borne by lenders, borrowers, taxpayers, or some combination ofthese.3 To take an example from the Great Depression, U.S. debtswere written down by 50% or more in 1935 through the legislativeexpedient of voiding gold clauses in contracts.4 Taxpayersabsorbed the losses from the savings and loan crisis of the 1980s.5Japans lost decade in the 1990s offers an example of the slowand agonizing approach to resolving a debt crisis; the central bankand government were unwilling to require banks to write down

their assets and instead allowed overvalued debts to remain on thebooks in the vain hope that they could eventually be repaid,


3/32

REWRITINGCONTRACTS,WHOLESALEWORKINGDRAFT 3

causing a huge drag on the national economy for more than adecade.6

Home mortgage debt in the United States mushroomed fromabout $4 trillion in 1998 to $10 trillion in 2007.7 During the samedecade, the median income remained virtually unchanged inconstant dollars,8 and the number of homeowners rose at arelatively modest pace.9 Inevitably, homeownership has becomeprogressively less affordable, and the ability of Americans toservice their growing mortgage debt has reached a breaking

point.10

The rapid growth of subprime mortgage lending was animportant contributor to the mortgage debt bubble.

In the case of the August 2007 collapse in the subprimemortgage market, U.S. policy makers have recognized the need forsome sort of intervention to readjust home values and mortgagedebt, but have to date been unwilling to use the tools of taxpayer-funded bailouts or legislated debt reduction. Instead, the Treasury

Department encouraged voluntary efforts by mortgage servicers,on behalf of the investors holding the inflated debt, to renegotiatethe terms of residential mortgages.11 Lenders and mortgageservicers, faced with rapidly escalating foreclosure rates, plunginghome prices, and mounting losses, were exhorted to renegotiatemortgage terms with borrowers in order to stave off even morewidespread defaults and foreclosures.

6 JAPAN'S LOST DECADE: POLICIES FORECONOMIC REVIVAL (Tim Callen &

Jonathan D. Ostry eds., 2003).7 FEDERAL RESERVE BOARD STATISTICAL RELEASE Z 1 t 8 (J 25 2008)


4/32

4 [AUGUST2008]

A year into the crisis, it is possible to begin evaluating thesuccess or failure of these voluntary efforts to resolve the debt

overhang by renegotiating contracts one at a time, albeit on anunprecedented scale. Data on voluntary mortgage modificationsare available from a number of sources, including the HOPE NOWad hoc coalition of mortgage servicers and counselors, theMortgage Bankers Association and the working group of statebanking and consumer credit regulators working on the foreclosurecrisis.

12These data provide some evidence as to the effectiveness

of the voluntary restructuring approach, which aims to avoid any

taxpayer contribution and to allocate losses between borrowers andlenders on a negotiated, and hopefully optimal basis. On the otherhand, the various available reports do not specify what kind ofmodifications are implemented, and the degree, if any, to whichmortgage debt is being reduced to a more sustainable level.

To supplement the national reports on mortgage modifications,this paper analyzes data derived from monthly remittance reports

by mortgage servicers to their investors, which provide rich detailon individual mortgage foreclosures and modifications. Theselected sample consisted of monthly reports from July 2007through June 2008 for twenty-six different subprime loan pools,and included data on 4,344 loan modifications. The data includethe loan balance, monthly payment and interest rate, before andafter modification. This paper will begin by reviewing somehistory of the voluntary plan to resolve the subprime mortgage

crisis and the previous reports on voluntary loan modifications. Iwill then present the new data from remittance reports, and inparticular on two key outcomes of loan modifications: whethert t l t d bt i b i d d d h th thl


5/32


II. LOSS MITIGATION AND LOAN MODIFICATIONSTHEVOLUNTARY APPROACH

While bankruptcy regimes can be a useful means to realigndebts and asset values, the U.S. Bankruptcy Code specificallyforbids bankruptcy judges from modifying most residentialmortgages by reducing the debt to the market value of theproperty.13 Efforts in Congress to amend the Bankruptcy Codeand permit judges to impose debt restructuring through principalreduction have thus far met stiff resistance from the banking

industry, and consequentially have been stymied.14

Without abankruptcy regime (or something comparable) as a coercive tool,homeowners have little choice but to attempt to negotiateconcessions, such as interest rate reductions or payment deferrals,individually with their servicers.

A significant and growing portion of the $10.5 trillion inmortgage debt owed by Americans by 200815 consisted of

subprime mortgages.16

By the end of 2007 mortgage debtexceeded total aggregate home equity (in other words Americanshad borrowed more than half the value of all homes in the country)for the first time.17 Although home prices began declining in 2007and dropped by 20% or more in some areas, housing affordabilityhad not improved much by the middle of 2008, because homeprices were still out of the reach of many Americans.18

13 13 U.S.C. 1322(b)(2); Nobelman v. Am. Sav. Bank, 508 U.S. 324 (1993);Adam J. Levitin & Joshua Goodman, The Effect of Bankruptcy Strip-Down on

Mortgage Markets (Georgetown Law and Economics Research Paper No.

1087816 (F b 6 2008)) htt // / b t t 1087816


6/32

6 [AUGUST2008]

The subprime credit crisis reached the breaking point in August2007, after several investment funds relying heavily on subprime

mortgage derivatives collapsed, securities affected by subprimedefaults were discovered in bank portfolios around the world,interbank lending suddenly froze, and the Federal Reserve andEuropean Central Bank had to inject billions of dollars and eurosinto the international financial system.19 As the subprime crisisunfolded, home values that had risen to unsustainable levels beganto decline.

20At the same time, increasing numbers of homeowners

defaulted on their mortgages and faced foreclosure.21 Investors

had assumed that in the event of defaults, securities backed by U.S.home mortgages would be safe, because the homes securing themortgages could be foreclosed and sold to recover any unpaidloans. In practice, however, subprime mortgage servicers rarelyrecover 100% of the debt in a foreclosure. After the 2007 crisis,the combined effect of high foreclosure rates and plummetinghome values meant that foreclosure recovery rates (usuallymeasured as loss severities22) progressively worsened. Bond ratingagencies have predicted loss severities on subprime foreclosures ashigh as 50%.23

In this environment, it makes economic sense for mortgageservicers, on behalf of lenders and investors, to seek alternatives to

19 Larry Elliott, Credit CrisisHow it All Began, GUARDIAN, Aug. 5 2008,

http://www.guardian.co.uk/business/2008/aug/05/northernrock.banking.20

Home prices in the United States more than doubled between 1997 and2007, before falling more than 15% from 2007 to 2008. CSI: Credit Crunch,

ECONOMIST, Oct. 18, 2007, http://www.economist.com/specialreports/display

story.cfm?story_id=9972489; Home Prices drop Record 15.8% in May, L.A.

TIMES t C 3 J l 29 2008 htt // ti l l ti /2008/j l/29/b i /fi


7/32


foreclosure by restructuring mortgage loans with borrowers, in anycase where the borrower can be expected to repay even 80% or

90% of the original debt.24

Wooden insistence on adherence to theoriginal contract terms may result in the servicer recovering farless than if the contract is modified. Borrowers who are unable topay subprime mortgages on their original terms may be able tomake reduced monthly payments. Payments may be reduced bydropping the interest rate or the loan balance, or both. Borrowerswhose mortgage debt exceeds their home value may have anincentive to default, but that incentive can be reduced if the

servicer agrees to write down the loan balance to the propertyvalue. Resolution of the mortgage debt crisis without trulymassive foreclosures thus depends on loan modifications thataccomplish two things: reducing principal debt and reducingmonthly payments. Based on these arguments, Administrationofficials and bank regulators called on mortgage servicers tonegotiate interest rate and principal reductions by modifyingmortgage contracts, as an alternative to foreclosure, in appropriate

cases.25

Securitization of mortgages has added layers of difficulty to thetask of loss mitigation for subprime mortgages.26 Servicers face

24 The Case for Loan Modification, 1 FDIC Q. 21 (2007), http://www2.fdic.

gov/qbp/2007sep/qbp.pdf; Patricia McCoy and Elizabeth Renuart, The Legal

Infrastructure of Subprime and Nontraditional Home Mortgages, 33-39 HarvardJoint Center for Housing Studies (2008) http://www.jchs.harvard.edu/

publications/finance/understanding_consumer_credit/papers/ucc08-

5_mccoy_renuart.pdf.25 B j i B k R d i P t bl M t F l (M h


8/32

8 [AUGUST2008]

constraints on their ability to renegotiate mortgages, and have littleeconomic incentive to incur the additional cost of loan

modifications.27

On October 10, 2007 Treasury Secretary Henry M. Paulson, Jr.announced the formation of a coalition of mortgage servicers andhousing counseling agencies, called HOPE NOW, to stimulate avoluntary effort to restructure mortgages, and ostensibly to respondto the subprime foreclosure crisis without mandatory debtrestructuring measures or a taxpayer-financed bailout.28 In

December 2007 the HOPE NOW coalition announced an initiativeto encourage mortgage servicers to freeze interest rates oncertain adjustable-rate mortgage (ARM) loans to preventforeclosures resulting from sudden payment increases.29 Apartfrom the rate freeze plan, HOPE NOW functioned primarily toexhort various industry participants to increase efforts to preventforeclosures, and to collect and report data on the success of thoseexhortations.

The payment increases targeted by the rate freeze plan were theresult of hybrid ARM structures where the initial interest rate waslower than the rate in effect for most of the loan life, the latterbeing calculated by adding an index rate to the stipulated margin.30Although the initial interest rates on subprime ARMs were notparticularly low, the loans were designed so that a paymentincrease due to the rate reset was inevitable after two to three

years. While much attention was focused on the payment resetissue, it has become apparent that the subprime foreclosure crisis

Engel & Patricia A McCoy Turning a Blind Eye: Wall Street Finance of


9/32


was not a result of payment resets, but instead reflected the factthat many subprime mortgage payments were unaffordable from

the date of origination.31

The HOPE NOW servicer coalition reports, issued beginningin February 2008, noted the significant increase in voluntary loanmodification agreements, as well as repayment plans, beginning inthe last quarter of 2007. HOPE NOW estimated that 140,000mortgages were modified in the fourth quarter of 2007, 170,000 inthe first quarter of 2008, and 220,000 in the second quarter of

2008.32

The HOPE NOW surveys did not include information onthe type of modifications being offered by its members.

A working group of state consumer protection and banksupervisor agencies issued a report in April 2008 that there were24,000 loan modifications closed and 139,000 in process in thefour-month period from October 2007 through January 2008.

33

The state regulators also noted that many payment plans and

proposed modifications were started but not completed. Althoughnot providing numbers, the report comments that freezing theinterest rate on variable-rate mortgages was the most commonmodification. The HOPE NOW and state regulator reports choseto emphasize different aspects of the same data. HOPE NOWpointed to the large increase in raw numbers of payment plans andmodifications, while also acknowledging the even larger increasein foreclosure starts and sales. The state regulators highlighted the

gap between total delinquencies and modifications, and the gapbetween workout efforts initiated and completed agreements,suggesting that many homeowners were seeking help but gettinglost in the shuffle 34 Although the state regulators recognized the


10/32

10 [AUGUST2008]

importance of determining what kind of modifications areimplemented, they have not provided statistical information on that

question to date.

The Mortgage Bankers Association (MBA) responded toindustry critics with its own survey of mortgage foreclosures,payment plans and modification agreements in January 2008.35The MBA reported that servicers encountered significant difficultycontacting borrowers in foreclosure, 23% of whom made noresponse to servicers attempts to contact them. The MBA survey

also found that 29% of borrowers in foreclosure had alreadydefaulted on prior repayment plans. The thrust of the MBA reportwas that the disproportion between loan modifications and the totalnumber of mortgage foreclosures could be explained in part bythese two factors.

While investors and rating agencies are understandablyskeptical about servicers and borrowers who repeatedly enter into

unrealistic payment or modification plans that are unlikely tosucceed,36 consumer advocates point out that many servicers haveoffered cookie-cutter repayment plans, plans that require increasedmonthly payments at a time when homeowners are cannot affordtheir current payments. If servicers would cut interest rates,principal balances and monthly payments, more loan modificationswould succeed and result in on-time repayment. Consumeradvocates have also criticized mortgage servicers for devoting

inadequate personnel and resources to the modification effort andshowing a lack of sensitivity to the plight of homeowners.37


11/32


There are at least three types of modifications.38 To deal withpayment arrears, a servicer can add the unpaid interest for the

months in arrears to the total loan balance, and then calculate anew (necessarily higher) monthly payment that will amortize theincreased balance over the remaining months of the mortgage life.This plan is sometimes called recasting arrears. Without amodification of the principal or interest rate, recasting plans neitherreduce mortgage debt nor diminish the payment stress faced by theborrower.

The rate-freeze modifications also do not diminish mortgagedebt or payment stress, except insofar as they prevent paymentsfrom increasing due to future rate adjustments. While a rate freezemay help prevent some loans from going into default, it will nothelp a borrower who had difficulty meeting the initial loanpayment. There is considerable evidence that rate resets on hybridARMs have not been the primary cause of the foreclosure crisis of2007-2008.39

The third type of loan modification addresses payment stressby reducing the interest rate in order to reduce the monthlypayment. An interest-only reduction deals with payment stress asa cause of default and foreclosure, but does not reduce thehomeowners overall mortgage debt, or deal with the negativeequity problem that results when home values decline below theamount owed.

A fourth type of loan modification would reduce principal debtand the interest rate, to deal with both payment stress and negativeequity issues. As we shall see, this fourth category of


12/32

12 [AUGUST2008]

payment burdens are being eased. The data reported in this papercan begin to answer these questions.

III. STUDY METHOD

Loan-level data on individual mortgage modifications andindividual foreclosures are available for one segment of themarket: subprime mortgages that were pooled and securitized.Trustees and their servicing companies report monthly to investorson the performance of the mortgage loans. These monthlyperformance reports, known as remittance reports, provide loan-by-loan details on defaults, foreclosures, losses on foreclosedhomes, and negotiated loan modifications. Selected portions of aremittance report are reproduced in Appendix One.

Mortgage servicers prepare monthly remittance reports for theinvestors who hold a stake in securitized mortgage loan pools.These monthly reports allow investors to see how the underlyingmortgages are performing, with detailed data about prepayments,defaults, foreclosures, losses and loan modifications. Theremittance reports are not filed with any public agency such as theSecurities and Exchange Commission, and their availability variesby servicer and trustee. In addition, different securitizationtrustees and servicers provide different data elements and detail intheir monthly reports. For this study, I used the reports madeavailable by CTS Link, the Wells Fargo trust service web site.40

These cover mortgage pools for which Wells Fargo serves astrustee; the pools are serviced by many of the leading mortgageservicing companies.

d l bl d f


13/32


collected for a twelve-month period, from July 2007 through June2008. The HOPE NOW initiative was announced in the early part

of this period, and national reports indicate a rapid growth in thenumber of mortgage modifications during this time frame.

Remittance reports were downloaded and entered into a databasefrom the following twenty-six loan pools:

Table 1

Pool name Originator Servicer

ABFC 2005-OPT1 Option One Option OneABFC 2006-OPT1 Option One Option One

ABFC 2006-OPT2 Option One Option One

ABFC 2006-OPT3 Option One Option One

Aames MIT 2005-4 Aames Mortgage Aames Funding

Aames MIT 2006-1 Aames Mortgage Aames Funding

ACE Securities 2006-CW1 Countrywide Countrywide

ACE Securities 2006FM-1 Fremont Fremont

First Franklin LT 2005-FF6 First Franklin First FranklinFFLT 2006-FF1 First Franklin First Franklin

FFLT 2006-FF11 First Franklin First Franklin

Fremont HLTrust 2005A Fremont Fremont

Fremont HLTrust 2006A Fremont Fremont

HSIASC WMC2006-1 WMC Wells Fargo

JPMAC2006-CW2 Countrywide Countrywide

Merrill Lynch MIT 2006-1 WMC Wells Fargo

Park Place 2005WHQ1 Argent/Ameriquest HomEqPark Place 2005 WHQ4 Argent HomEq

Renaissance HELT2006-1 Delta Ocwen

Renaissance HELT2005 1 Delta Ocwen


14/32

14 [AUGUST2008]

Countrywide and Wells Fargo were the number one and twooriginators, respectively, of mortgages in all categories in 2006.41

The sample includes seven of the top ten subprime originators in200642 and six of the top fifteen subprime servicers in 2007.43

This sample obviously has its limitations. Servicers ofsecuritized loans may have different incentives than lenders whoretain ownership of mortgage loans on their own balance sheets(so-called portfolio lenders.) The number and extent of voluntaryloan modifications in the sample may not be representative of loan

modifications by lenders who hold subprime mortgages in theirown portfolios. Nevertheless, given that subprime mortgagesaccount for more than half of all foreclosures,44 and that the vastmajority of subprime loans that led to the crisis were securitized,this sample provides important insights as to what the voluntarydebt resolution plan has yielded to date in the subprime market.

For each loan pool, the prospectus prepared for investors was

downloaded and retained. The prospectuses and supplementsprovide detailed descriptions of the pools of mortgage loans. Mostof the mortgage pools in the sample were dominated by hybridadjustable-rate subprime mortgages, although the Renaissancepools originated by Delta Funding Corporation were mostly fixed-rate loans. Many of the pools included significant percentages ofno-doc loans, i.e., mortgages approved without verification ordocumentation of the borrowers income. For example 52% of the

loans in the HSIASC 2006-WMC1 pool were stated income (noincome verification) loans. Loans tended to be geographicallyconcentrated in California and a few other high-growth states. Inother respects the mortgages in the study sample were typical of


15/32


IV. LOAN-LEVEL DATA ON MORTGAGE REFINANCINGS, DEFAULTSAND LOSSES DURING THE 2007-08 CRISIS

A. Growing Defaults, Foreclosures and REO

In July 2007 the average delinquency rate in the pools was19%, and 1.4% of all loans entered foreclosure that month (analready disturbing 16.8% annual rate of foreclosure starts.) ByJune 2008 the average delinquency rate had nearly doubled to34%, and foreclosure starts were at 2.3% per month, or 27%annually. In all pools, the number of loans entering the foreclosureand real-estate-owned (REO) categories far outpaced the numberof prepayments, so that the foreclosure and REO inventory steadilyincreased in every month. While some increase over time indefaults and foreclosures in a static mortgage pool is to beexpected, the magnitude of the increase in the study period isextraordinary, and is one indication that the voluntary resolutionplan is falling short.

B. Losses on Foreclosed Properties: Loss Severities Worsen

Mortgage servicers and investors measure the losses onindividual mortgage loans by comparing the dollar loss incurred tothe total outstanding mortgage debt. This ratio is known as lossseverity.45 Loss severities increased steadily in most pools fromJuly 2007 to June 2008. By June 2008 loss severities in individualpools ranged from 17% to 71%, with the average loss severityrunning at 38% (fig. 1).


16/32

16 [AUGUST2008]

Fig. 1

To take one example, an Illinois property in the First Franklin2006-F11 pool was originated in late 2006 with a $630,000original balance and a 100% loan-to-value or LTV ratio, and aninterest rate of 10.125%. The home was foreclosed and sold with atotal loss realized in June 2008 of $332,000. A California propertyin the same month in the same pool with an original mortgagebalance of $272,000, thought to represent an 80% LTV ratio, with7.375% interest, was sold at a loss of $185,000. In either case, hadthere been a viable homeowner with any reasonable income, it ishard to imagine that a modification reducing principal and interest

to an affordable level could not have produced a smaller ultimateloss for investors than wiping out all interest and 52% to 68% ofprincipal. Loss severities are a critical motivating factor forinvestors and servicers in deciding whether and how to renegotiate


17/32


they were in default, thus perhaps reducing the need for

modifications.

While it is clear that total residential mortgage debt in theUnited States reached a point that was not sustainable by themiddle of 2007, that debt total did not shrink much as the crisisbegan to unfold. From June 30, 2007 to March 31, 2008 totalresidential mortgage debt actually continued to increase, from$10.24 trillion to $10.61 trillion.46 The total mortgage debt in thesample loan pools shrank steadily during the study period, partly

from refinancing and sales (payment in full) and partly fromforeclosure liquidations. This shrinkage in outstanding balances instatic mortgage pools does not, however, equate to an overallreduction in mortgage debt. This is because until the middle of2008 many mortgages were prepaid by refinancing with a newmortgage, typically at a higher balance to pay the refinancingcosts.

Refinancing, while it solved the borrowers and servicersimmediate problem, did nothing to resolve the debt overhang,because the new mortgage will invariably bear a larger principal,especially when the borrower pays a prepayment penalty whenrefinancing within the first two or three years.47 Borrowers whorefinanced were simply kicking the can down the road,48increasing, not decreasing debt. The study sample shows that therefinancing outcome was still fairly common in the last six months

of 2007, but gradually disappeared during the first half of 2008. InJuly 2007 there were 314 loans liquidated (foreclosed propertiessold) compared with 2654 loans paid in full via refinancing or sale.By June 2008 however the totals were nearly equal: 948


18/32

18 [AUGUST2008]

Fig. 2

Surprisingly, refinancings continued long after the August2007 freeze-up of the credit markets, and even into early 2008.Nevertheless, by the final month of the study, June 2008, therefinancing (or voluntary sale) option was rapidly declining. Theoption of refinancing obviously is more attractive for the servicer,because its investor receives payment in full, and the costs ofnegotiating a mortgage modification are avoided, along with thelegal uncertainties. The apparent reluctance to engage in large-

scale modifications may in part be explained by the continuedviability of the refinancing option, which shifted the debt problemfrom one economic actor to another, rather than tackling it head-


19/32


loan pools studied were running as low as 12%,50 and nearly halfof all prepayments were foreclosure liquidations. These

prepayment speeds would mean that it would take ten to fifteenyears for the mortgages to be paid off (or foreclosed), suggesting avery drawn-out resolution process for the subprime mortgagecrisis.

The declining number of monthly refinancings has seriousimplications. If the trend as of June 2008, of equal numbers offoreclosure liquidations and loans paid in full, continues, the final

outcome for half of the remaining loans in the pools will be aforeclosure sale. It also means that the REO inventory willcontinue to accumulate and put downward pressure on homevalues.

IV. THE LOAN MODIFICATIONSTHENATURE AND SCALE OF THEEFFORT

A. Virtually No Principal Reductions, Some Payment Reductions

In the study sample of twenty-six loan pools over twelvemonths, there were a total of 4,342 loan modifications reported.The number of modifications increased from twenty-nine in July2007 to a high of 880 in April, declining somewhat to 582 in June,2008, the final month covered (see fig. 3).51 During the sametwelve months, there were 19,911 foreclosures started and 8,327properties foreclosed and taken into REO.


20/32

20 [AUGUST2008]

Fig. 3

In the aggregate, the loan modifications failed to reduce theoutstanding mortgage debt. The amount owed on the modifiedloans went from $912 million before modification to $933 millionafter modification. A few loans did have their principal balancereduced, but only 62 (1.4%) of the 4342 modifications reduced theprincipal balance by more than 1%, and only 40 (0.92%) reducedprincipal by more than 10%. Some of these large principalreductions may have resulted from litigation.

The most common forms of modifications involved either nochange in interest, with a modest increase in principal (recastingarrears) or a reduction in the interest rate and payment with

i i l i i h i i li h l i i


21/32


potentially for many years to come. Without principal reductions,the only way debt will realign with property values is through

eventual foreclosures or short sales, or through the slow process ofamortization of 30-year loans.

The news is slightly better regarding reduction in paymentstress. Monthly mortgage payments were reduced in 54% of themodifications. On the other hand, 23% of reported modificationsresulted in payment increases, likely a product of recasting arrears.The remaining 23% of modifications did not change the monthly

payment.

52

The share of modifications with payment reductionsincreased somewhat over time. (see fig. 4)

Fig. 4

The largest payment reduction involved a $730,000 mortgagein the Fremont HLT 2005A pool Witho t red cing the principal


22/32

22 [AUGUST2008]

reduction was 21%. The reports do not disclose whether thepayment changes are permanent or temporary.

Fig. 5

On the other hand, the average interest rate AFTERmodification on all modified loans was 7.54% (7.4% median).Given the near-complete absence of principal reductions and thestill above-market interest rates being paid,53 the mortgagemodifications of 2007/08 should not be regarded as bailouts for theaffected homeowners by any means.

The modifications that did not change either the interest rate orthe payment amounts are likely to have occurred when anadjustable interest rate was due to increase. Servicers may haveagreed to forego a scheduled rate and payment increase either for


23/32


the principal, interest, and payment amount before and aftermodification.

Thus rate freezes and recasting of arrears (which increasepayments and balances) accounted for about half the mortgagemodifications in the 2007 to 2008 period. While slightly morethan half of loan modifications involved an interest rate reductionand resulting payment reduction, fewer than 1% reduced the actualmortgage balance. Payment stress was thus reduced somewhat,and debt overhang not at all.

B. Modifications Compared to Foreclosures

To evaluate the overall scale of the voluntary modificationeffort, it is useful to compare the number of modifications of allkinds with the total number of delinquent loans, the number of newforeclosures started, and with the number lost to foreclosure.

Table 2

Modifications Delinquent

Loans

Foreclosures

Started

Liquidated

Loans

Monthly Average

per pool

14 951 64 24

Total all pools

July 2007

27 19,375 1,412 314

Total Dec 2007 467 25,170 1,783 511


24/32

24 [AUGUST2008]

C. Variability Among Servicers

Mortgage servicers were far from consistent in their approachto loan modifications. The number of modifications variedconsiderably among the different loan pools. The two Park Place2005 WHQ1 and WHQ 2 pools (Argent originator, HomEqservicer) had none and one, respectively, while the Ace Securities2006-FM1 pool (Fremont) had 701 modifications over twelvemonths. Even comparing modifications to the number ofliquidated foreclosure properties, or the number of delinquent

loans, the level of modification activity varied tremendously. Eachloan pool, of course, varies in size, and is composed of loans withdifferent characteristics, and the servicers may face differentconstraints on their ability to modify loans. Nevertheless, thevariation remains striking. To take one example, Option Onemodified 2% of its delinquent loans in one pool, 11% in an olderpool and 7.5% in a more recent pool. Fremont modified 38%,41.5% and 56% of its delinquent loans over the course of the

twelve months, compared with 0% to 10% in a number of pools.

Table 3


25/32


reduce monthly payments, while others limited their modificationsto recasting or rate freezes that did not reduce monthly payments.

Fig. 6

D. Adjustable Rate Loans and the Payment Shock Issue


26/32

26 [AUGUST2008]

defaults, foreclosures and workouts are driven by other factors,such as payments that were unaffordable at inception.55

Table 4

E. Discussion

The mortgage debt overhang continues to grow, and as of June2008, defaults, foreclosures and REO (foreclosed homes owned bymortgage servicers) continue to mount, adding to the glut of unsoldhomes and the drag on the economy. The voluntary mortgagemodification effort is providing some relief for many homeowners,but is not, broadly speaking, having an impact on the debt crisis.

The following salient points emerge from the data.

First the number of defaults foreclosures and REO properties


27/32


Third, there are no signs in the study data of the crisisbottoming out or reaching a turning point, through June 2008.

Instead, the disappearance of refinancing and the acceleration offoreclosures means that the number of homes in foreclosure andREO will continue to grow, and at present rates will not beliquidated for several more years. All things being equal, the 2005and 2006 loan pools in this study could take as many as ten yearsto process the delinquent loans into foreclosures, the foreclosuresinto REO, and to sell the REO, at present rates.

Fourth, the effort to increase loan modifications has had somesuccess, although the number of loans modified is still outstrippedby loans being foreclosed, even at very high loss levels.

Fifth, the modification effort is not solving the fundamentaldebt overhang problem, because principal balances are not beingreduced. The subprime crisis is unique in the sense that banks andinvestors have already written off hundreds of billions of dollars of

securities backed by subprime mortgages, but only a small fractionof those losses actually correspond to completed loan liquidations.More importantly, homeowners have not been relieved of thedevalued debt, either through completed foreclosure sales or loanconcessions. Many are still stuck in a sweat box struggling topay above-market interest rates on above-market mortgage loans.56

Sixth, only a bare majority of modifications involve anypayment relief for borrowers, while many modifications are simplyarrears capitalization arrangements that put borrowers in greaterpayment difficulty.

h h i i i di


28/32

28 [AUGUST2008]

announcements by Treasury and HOPE NOW, and some seemedreluctant to engage in loan modifications until late in the study

period, with a few modifications appearing first in March or April2008, months after the crisis began and the exhortations torenegotiate had been made. In some pools, 80% of modificationsreduced the borrowers monthly payment, while in others, 95% ormore of modifications increased payments or left them unchanged.

V. CONCLUSION

The need for mortgage restructuring was clearly growingduring the 2007-2008 period, as delinquencies and lossesprogressively worsened. Given the choice between permanentlywriting down principal debt and interest rates, or simplyrescheduling unpaid payments, it is understandable that mortgageservicers would avoid the former option and favor the latter.Thorny questions of servicer authority are avoided, borrowers and

their advocates are placated, and some foreclosures are avoided.On the other hand, if the twin objectives are to reduce theunsustainable levels of subprime mortgage debt, and to reduce thepayment burden on mortgage borrowers in or near default, thevoluntary plan is not working. Borrowers who remain indebted foramounts exceeding property values will retain the incentive to giveup their valiant struggles to repay their subprime loans, andforeclosures will be delayed rather than prevented.

The voluntary mortgage renegotiation plan, while it may besignificantly reducing hardship for individual homeowners

il b bi d l i f l l i d i li l


29/32


legislated debt reduction, or a similar solution, the subprime crisiswill drag on for years.57


30/32

Appendix One Sample Remittance Report, Selected Portions


31/32

*************

Prepayment Penalty Paid Amount 6,042.36 47,318.23 53,360.59

Prepayment Penalty Paid Count 2 13 15

Prepayment Penalty Waived Amount 0.00 0.00 0.00

Prepayment Penalty Waived Count 0 0 0

Ending Scheduled Balance 130,805,621.55 547,398,981.15 678,204,602.70

Actual Ending Collateral Balance 130,963,175.90 548,201,265.32 679,164,441.22

Ending Loan Count 837 3,074 3,911

Beginning Scheduled Balance 131,238,581.94 554,389,973.00 685,628,554.94

Beginning Loan Count 843 3,118 3,961

Loans Paid in Full 6 44 50

Weighted Average Remaining Term 328 346 342

Principal And Interest Constant 943,527.68 4,045,148.45 4,988,676.13

Weighted Average Net Rate 7.287792 7.679293 7.604355

Weighted Average Pass-Through Rate 7.277792 7.669293 7.594355

Collateral Description Mixed Fixed Fixed 15/30 & ARM Mixed Fixed & Arm

Weighted Average Coupon Rate 7.787792 8.179293 8.104355

J.P. MORGAN MORTGAGE ACQUISITION TRUST

Asset Backed Pass-Through Certificates

Distribution Date: 25-Mar-2008 J.P. MORGAN MORTGAGE ACQUISITION TRUST

Asset Backed Pass-Through Certificates

Series 2006-CW2

Collateral Statement

Group 1 2 Total


32/32

BANKRUPTCY FORECLOSURE REO TOTAL

No. of Actual No. of

1 148,504.30 1

30 Days 186 36,031,929.11 3 633,164.55 0 0.00 0 0.00 189 36,665,093.66

60 Days 125 25,320,755.03 1 110,695.18 2 198,378.47 0 0.00 128 25,629,828.68

90 Days 97 18,301,410.07 2 222,402.68 2 581,678.36 0 0.00 101 19,105,491.11

120 Days 72 14,344,109.65 5 659,150.94 15 2,445,749.05 1 88,354.73 93 17,537,364.37

150 Days 42 6,608,058.61 9 775,006.87 37 7,074,578.37 1 113,276.81 89 14,570,920.66

180+ Days 134 19,389,590.02 47 6,806,428.34 184 37,000,507.27 216 43,863,848.94 581 107,060,374.57

656 119,995,852.49 68 9,355,352.86 241 47,419,322.62 218 44,065,480.48 1,183 220,836,008.45

********

43

8,581,718.00

8,532,098.11

241

47,622,756.18

47,419,322.62

*******

0118656414 197,318.58 209,527.96 8.850% 8.850% 1,587.71 1,687.08

0119815469 223,078.19 240,599.75 7.870% 7.870% 1,539.11 1,539.11

0121763884 111,635.53 121,001.88 9.375% 9.375% 939.88 1,019.35

0133578094 91,669.79 101,053.97 10.000% 10.000% 816.14 896.35

0133736519 98,578.53 115,829.24 7.875% 7.875% 1,000.57 1,000.57

0133845828 105,597.39 118,343.46 8.300% 8.300% 808.53 906.79

0133983558 372,706.10 389,084.96 8.375% 8.375% 2,873.07 3,001.78

0134156077 108,046.63 122,737.13 10.700% 10.700% 1,013.40 1,013.40

0134662316 122,434.83 127,922.74 8.900% 8.900% 988.82 1,033.83

0135355571 304,167.35 313,082.65 6.500% 6.500% 1,962.57 2,021.47

Foreclosure Detail - All Mortgage Loans in Foreclosure during Current Period

Summary 12 Month Foreclosure History

New Foreclosure Loans

0 0.00 2 266,935.40

Loans in Foreclosure

118,431.10

Original Principal Balance

No. of

Loans

Actual

Balance

Current Foreclosure Total

DELINQUENT

Delinquency Status

Loans in Foreclosure

Original Principal Balance

Actual

Balance

Modifications

Beginning

Scheduled

Balance

Current

Scheduled

BalanceLoan

Number

Prior

Rate

Modified

Rate

Prior

Payment

Modified

Payment

No. of

Loans

Actual

Balance

No. of

Loans

Actual

Balance

Current Actual Balance

Current Actual Balance

rewriting contracts, wholesale data on voluntary mortgage modifications from 2007 and 2008...

Documents