«Barriers to Federal Home Mortgage Modification Efforts During the Financial Crisis Patricia A. McCoy August 2010 MF10-6 Paper originally presented at ...»
11 of loan workouts studied increased the borrowers’ principal by adding in overdue interest and fees. The average principal increase was a whopping $10,800.38 The Boston Fed study confirmed White’s findings. By the end of 2008, plans increasing principal remained the most common type of workout by far. Such capitalization plans accounted for 61.5% of all loan workouts in fourth quarter 2008. Interest rate reductions came in second (26.7%) and principal reductions remained rare (1.4%).39 (Figure 5).
The implications of these findings were profound. The wide prevalence of capitalization plans at year-end 2008 indicated that servicers were kicking the proverbial can down the road by temporarily delaying foreclosures, not preventing them. Increasing monthly payments for cashstrapped borrowers usually was a recipe for failure, in view of the long average stint of unemployment during this recession, coupled with negative equity. More would be needed to induce servicers to lower monthly payments.
2009 and HAMP HAMP sought to alter servicers’ fee calculus by handing them cash for making loan modifications. While this intention was worthy, HAMP’s progress remains modest. HAMP had some success in discouraging capitalizations and encouraging interest rate reductions. HAMP had great success in boosting the number of trial loan modifications. The number of HAMP borrowers who moved into permanent modifications, however, has been disappointingly low, although the numbers have started to improve.
By 2009, for approved workouts, interest rate reductions surpassed capitalizations that increased principal. By the third quarter of 2009, for example, fully 81.1% of workouts by national banks and thrifts involved reducing interest rates. Principal reductions increased slightly, while term extensions held steady.40 Fitch Ratings found similar trends for private-label securitized mortgages.41 Thus, HAMP may have spurred the trend toward more loan modifications with lower payments.
HAMP also produced large numbers of trial loan modifications. From HAMP’s inception through year-end 2009, over 1.1 million borrowers were offered trial loan 38 White (2009a), at 1114; White (2009b).
39 Adelino et al. (2009), at 11-12 & tbl. 3. According to the authors, Ocwen Loan Servicing, LLC and Litton Loan Servicing LP were the only servicers who granted principal reductions in nontrivial amounts. Id. at 12 n.17. See also Mason (2009), at 32.
40 Adelino et al. (2009), at 11-12 & tbl. 3; OCC and OTS (2009), at 23, 28.
41 Fitch Ratings (2009b), at 10.
12 modifications. But of the 900,000-some borrowers who accepted trial modifications, only 66,465 – less than 8% -- graduated into permanent modifications. Servicer performance was all over the lot: CitiMortgage had modified 47% of its eligible seriously delinquent loans as of December 31, 2009, while Wachovia had only modified 3%.42 The low rate of permanent modifications made Fitch Ratings conclude: “[T]he conversion from trial mod under HAMP to actual finalized modification status has been disappointing.”43 Reportedly, several issues stymied permanent modifications. Some borrowers failed to complete their paperwork; other times, servicers lost their files. Other borrowers who did complete their paperwork were ineligible for HAMP based on their verified income. The government also paid servicers to do trial modifications, leading some to ask whether servicers had adequate financial incentives to carry through with permanent modification plans. Some accused servicers of deliberately approving trial modifications, collecting incentive payments, and proceeding to foreclosure to maximize their fees.44 In November 2009, the Obama Administration turned up the heat, pressing servicers to grant more permanent modifications. In an effort to raise the costs of not making permanent modifications, the Treasury Department even threatened to shame servicers who dragged their feet and subject them to fines and other sanctions. In addition, the government announced it would withhold incentive payments until modifications became permanent. The following month, the number of permanent loan modifications more than doubled.45 In January 2010, the government streamlined borrowers’ documentation requirements to make it easier to convert temporary modifications to permanent ones.46 These measures, taken together, helped increase the permanent modification rate in just a few months. From HAMP’s inception through February 2010, 170,207 (15.5%) of all trial modifications resulted in permanent modifications. Another 91,843 permanent modifications 42 Department of the Treasury (2010d), at 3-5. See also OCC and OTS (2009), at 43-44 (among servicers that were national banks or thrifts, only 4.1% of seriously delinquent borrowers and borrowers in foreclosure received new loan modifications).
43 Fitch Ratings (2009), at 2.
44 Federal Housing Finance Agency (2010), at 13-14; Goodman (2009).
45 Department of the Treasury (2010d), at 3; Goodman (2009). At the end of December 2009, servicers had approved another 46,056 permanent loan modifications and were awaiting acceptance by borrowers. For borrowers who received and accepted permanent loan modifications under HAMP, the median monthly payment dropped by over $500. Their median housing debt-toincome ratio dropped from 45% to 31% and their total debt-to-income ratio dropped from 72.2% to 55.1%. Department of the Treasury (2010d), at 3-4.
46 Department of the Treasury (2010a).
13 were approved and awaiting acceptance in February 2010. Assuming all of those pending modifications were accepted by the borrowers, that would raise the graduation rate to 24%.47 Still, the federal government remained dissatisfied. In March 2010, it took even more aggressive steps to boost permanent loan modifications. One provision was designed to give breathing room to out-of-work borrowers. For the first time under HAMP, unemployed borrowers who qualified for HAMP could have their mortgages payments cut to 31% of gross income for three to six months while they looked for work.48 In a sister provision, the Administration offered first-time carrots to servicers to approve principal write-downs and extinguish junior liens.49 Going forward, for any underwater borrower owing more than 115% of the current value of his or her home, HAMP servicers had to calculate the borrower’s net present value using both the standard approach, plus an alternative approach containing incentives for writing down principal. If a principal write-down was needed to reduce the borrower’s monthly payment to 31% of income, the servicer could – but was not obliged to -- reduce principal. To induce principal write-downs, the federal government offered to pay 10 to 21 cents for each dollar of unpaid principal written down (depending on the loan-tovalue ratio).50 Another part of the March 2010 announcement was designed to assist relocation for delinquent borrowers who failed to qualify for loan modifications. To encourage more short sales, the government hiked subsidies to junior lien holders to release their liens to 6% of the outstanding loan balance. In addition, incentive payments to servicers to perform more short sales rose from $1000 to $1500. The government also planned to double relocation payments to borrowers who successfully completed short sales or deed-in-lieu transactions, up to $3000.51 Finally, the March 2010 initiative rolled out more measures to lower administrative barriers to HAMP modifications. The Treasury Department prohibited servicers participating in HAMP from pursuing foreclosure during loan modification negotiations and trial modification periods.
Servicers also had to start considering borrowers in bankruptcy for HAMP relief upon request.
47 Department of the Treasury (2010c).
48 To qualify, the loan in question had to be for the borrower’s owner-occupied principal residence, have a mortgage balance of less than $729,750, and be originated before 2009. In addition, the borrower had to prove financial hardship and have a monthly mortgage payment exceeding 31% of his or her income. After six months, if the borrower found work with lower pay or did not find work at all, he or she would respectively be considered for a permanent HAMP modification or a short sale combined with relocation assistance. Department of the Treasury (2010e).
49 Department of the Treasury (2010e).
50 If the servicer opted to reduce principal, it would initially treat the reduction as forbearance. Te encourage borrowers to remain current on their new, lower loan payments, servicers would then forgive the forborne amount in three equal steps over three years, so long as the homeowner remained current on the payments. Department of the Treasury (2010e).
51 Department of the Treasury (2010e).
14 HAMP was extended to homeowners with FHA loans. No foreclosure sale could go forward without written certification that the borrower was ineligible for HAMP. Finally, the government raised the incentive payments to servicers for making permanent loan modifications.52 Possible Obstacles to Optimal Levels of Loan Modifications Although HAMP increased loan modifications with lower payments, the level of permanent loan modifications remains lower than hoped for. Meanwhile, principal reductions, while growing, remain rare. This section asks, what are possible frictions that might impede principal write-downs and permanent loan modifications?
Servicer Compensation One of the underlying assumptions behind HAMP was that compensation structures make servicers resistant to loan modifications. When servicers reject modifications that would improve the net present value of loans relative to foreclosure, this presents a classic agency problem vis-àvis the investors who own those loans. HAMP aims to correct this incentive structure and the resulting agency problem by subsidizing servicers to approve more loan modifications.
HAMP’s diagnosis is at least partially correct, even if its solution falls short. In privatelabel RMBS, servicers have four main sources of revenue: fixed servicing fees, float, default fees, and income from residual interests in the loan pool. This compensation structure has many moving parts and creates incentives pointing in different directions. In the main, however, this structure tilts servicers of private-label RMBS away from principal and interest reductions toward foreclosures and capitalization of arrears.
Fixed Servicing Fees The single biggest component of servicer compensation is the fixed monthly servicing fee.
This fee is computed as a percentage of the outstanding principal balance of the loan pool. Typically, annual fees are 25 basis points (bps) for prime fixed-rate loans, 37.5 bps for prime ARMs, and 50 bps or more for subprime loans. The servicer takes this fee off the top of borrowers’ monthly payments before remitting principal and income to the trustee to pass on to investors.53
52 Department of the Treasury (2010e).53 Cordell et al. (2008), at 15.
15 This fixed-fee structure can cut in favor either of a loan workout or foreclosure. A workout that keeps a loan on the books allows the servicer to continue to collect the servicing fee on the loan. At the same time, the fixed-fee structure favors workouts that raise monthly payments over modifications that lower them. Capitalizing arrears and default fees are more attractive because they pump up the unpaid balance of the loan pool and thus the servicing fee. Principal writedowns are less attractive because they lower the unpaid balance of the loan pool.54 In other respects, the fixed-fee structure militates in favor of foreclosure. In private-label securitizations, investors do not pay servicers additional fees for completed loan workouts.
Instead, servicers must pay the high cost of overhead and labor for loss mitigation out of their fixed monthly servicing fees. The higher those costs, the more reluctant a servicer will be to pursue loss mitigation in lieu of foreclosure.55 HAMP seeks to alter this compensation structure by paying servicers for each completed loan modification. Similarly, Fannie Mae and Freddie Mac pay servicers bonuses for executing workouts. But the GSEs pay significantly more for short sales than for loan modifications, which skews incentives away from modifications for conforming loans.56 Bottom-line, the fixed-fee structure can promote workouts, but only those of a certain kind. This structure makes servicers averse to principal write-downs. Conversely, it causes servicers to favor capitalizations, while making them indifferent to interest rate reductions.
Float Servicers also collect “float,” consisting of the interest earned on mortgage payments that are held in escrow. Servicers collect this interest between the date when borrowers send in their payments at the beginning of the month and the 25th of the month, when the payments are usually passed through to investors.57 Float has little effect on incentives to do loan modifications.
Default Fees Servicers charge borrowers default fees, such as late fees and default management fees, for late loan payments. Under pooling and servicing agreements (PSAs) for securitized trusts, 54 Thompson (2009), at vi, 19-20.
55 Cordell et al. (2008), at 23.
56 Cordell et al. (2008), at 17, 20-22.
57 Tanta (2007).
16 servicers get to keep all or part of these default fees and can collect them out of the proceeds from foreclosure.
The ability to assess default fees creates incentives toward foreclosure and capitalizing arrears. Default fees are highly lucrative and servicers want to collect them as soon as possible.
In a loan workout, the way to do that is to condition an agreement on immediate upfront payment or to capitalize arrears, either of which will increase the borrower’s payments. But if the default fees continue to mount, at some point a workout will become impractical. At that point, servicers will generally initiate foreclosure to collect the default fees.58 Income from Residual Interests Some servicers of private-label RMBS own the residual tranche of the loan pool, also known as the “B piece.” Usually, the B piece only pays excess interest. In other words, the B tranche only pays out if the interest payments by the borrowers exceed the monthly interest payments due to the senior tranche holders. As the junior tranche, the B piece holds the first-loss position, meaning that it is the first to absorb losses from a delinquent loan.