Rebooting the Private MBS Market

A significant revival of the private-label residential mortgage-backed securities (RMBS) market may be at least two years away. While there are many barriers and detours along the way, there may also be an occasional fast lane that could lead to new RMBS issues.

Mortgage Banking
October 2010

By Robert Stowe England

At a June conference on the future of mortgage finance sponsored by Moody’s Investors Service at its corporate headquarters in New York, the audience was polled with this question: When do you see the private-label residential mortgage-backed securities (RMBS) market making a comeback? 

An overwhelming 74 percent of the 150 attendees at a panel on RMBS said the market will come back no sooner than 2012, mirroring the views of the panel. A distinct minority—6 percent—said it would come back in 2011. However, a skeptical 20 percent voted in the poll that a comeback will happen “when hell freezes over,” according to a source who attended the event. 

The number of people who identify with that hell-freezing-over option may have grown since June, given the policy and market issues that have already emerged in the wake of unintended consequences from the passage of the 2,319-page Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010. The new law creates a number of barriers to the resurgence of RMBS, many of which were anticipated by the financial markets prior to passage. 

“The general consensus for the future is that there is doubt that [private-label RMBS] will come back and that it will come back in any way like it was before,” says Guy Cecala, publisher of Inside Mortgage Finance. “The basic problem is that the market has been dead for two years, and it’s really showing no signs of coming back whatsoever,” he adds. 

The cautiously hopeful see the prospects for RMBS as a glass with a splash of water in it rather than a glass virtually empty. 

“The market is very, very slowly coming back to life. It’s a very, very slow process,” says David Spector, chief investment officer at Calabasas, California–based Private National Mortgage Acceptance Co. LLC—or PennyMac, as it is better known—which completed a private-placement securitization of $372.8 million of non-performing alternative-A loans on Sept. 14.

All but one of eight recent deals (see sidebar) have involved the private-placement securitization of non-performing loans. The coupon in the PennyMac deal priced at an unexpectedly low 4.25 percent, which made the deal attractive for the sponsors. “If you think about the environment that existed from two years ago until now, I would consider that significant progress,” says Spector.

“Investors are getting more comfortable with these kinds of transactions and the collateral backing them,” says Spector. This improved investor appetite, combined with the ability to buy non-performing loans at very attractive prices, has made it economically feasible for the securitization market in non-performing loans to take hold, he explains.

There are six other private-placement RMBS securitization deals—all but two of them took place earlier this year (see sidebar for summary). The Federal Deposit Insurance Corporation (FDIC) sponsored securitization of performing loans from 12 failed banks in a pilot program in July and backed the securities with a guarantee.

There is also American General Finance Inc., Evansville, Indiana, which has done at least two securitizations and, before it was sold Aug. 11, had been planning many more. Eighty percent of American General’s shares, held by parent American International Group Inc. (AIG), were sold to the Fortress Investment Group LLC, New York. The sale closes in the first quarter of 2011. The list of securitizers on non-performing loans also includes Lone Star Funds, Dallas, a private equity group; and Waterfall Asset Management LLC, New York.

Sequoia Mortgage Trust

The only public securitization of recently originated mortgages this year was the Sequoia Mortgage Trust 2010-H1 deal that closed on April 28. The $237.8 million transaction was sponsored by Redwood Trust Inc., a Mill Valley, California–based real estate investment trust (REIT) that for more than a decade has been a major investor in jumbo residential mortgages. The collateral for the deal is 255 adjustable-rate jumbo mortgages originated mostly in the fall of 2009 by CitiMortgage Inc., O’Fallon, Missouri.

The Sequoia deal is the first public deal of recently originated mortgages since Oct. 30, 2008, when Green Tree Financial Corporation, St. Paul, Minnesota, did a small $62.63 million subprime home-equity loan securitization, according to Inside Mortgage Finance. Prior to that, there was a $146.97 million prime jumbo deal on Aug. 8, 2008, with the loans purchased from third-party originators through New York–based Lehman Brothers’ conduit program, according to Inside Mortgage Finance.

Redwood is also in the process of putting together a second public deal expected to include $300 million in recently originated jumbo mortgages, according to the company’s Aug. 4 form 8-K filing with the Securities and Exchange Commission (SEC).

Redwood’s business model is based on buying mortgages to securitize them and retain the subordinate tranches. Mike McMahon, Redwood’s managing director, explains, “The way we prefer to invest in mortgage credit risk is to buy prime jumbo mortgages, securitize them, sell the seniors and keep the subs. We like to be the cook. We like to eat our own cooking.”

While Redwood has at times bought the subordinated tranches put together by other parties, this is not its preferred way of doing business, “because we have less control over collateral quality and loss mitigation,” McMahon explains.

Redwood has invested in prime jumbo mortgage securitizations since 1997. The REIT has done roughly 50 securitization deals totaling $35 billion in prime jumbo mortgages, according to McMahon. So far, total cumulative losses have been 32 basis points, he notes. “So, we have a very good track record,” he says.

Redwood structured its Sequoia deal to address many of the concerns about the originate-to-distribute model of mortgage securitization that critics, including Federal Reserve Chairman Ben Bernanke, have said contributed to the financial crisis of 2007–2008. The primary concern has been that the originators had little incentive to care about the credit quality of the mortgages they originated and sold to Wall Street investment banks, because they did not expect to share in the credit loss exposure.

Similarly, Wall Street securitizers were faulted for having too little “skin in the game,” because the risk was being transferred to investors. Since the crisis, the Dodd-Frank law requires at least 5 percent retained credit risk on certain mortgages. The Redwood deal retains 6.5 percent of the subordinate risk and also retains 5 percent of the AAA, representing both vertical and horizontal risk.

Even with 6.5 percent subordination, Redwood was criticized for not doing enough. New York–based Standard & Poor’s (S&P), which had not rated the deal, issued a statement saying that the deal should have had 7.5 percent subordination. In its RMBS ratings criteria issued in September 2009, S&P set 7.5 percent credit enhancement as the level it would require for a typical pool of prime mortgages for a rating of AAA.

In a public statement at the time of the Sequoia deal closing, S&P said that the average loan balance at $932,699 poses a “concentration risk,” meaning that the portfolio lacks diversity in the size of the loans. S&P pointed out that a default of the 19 largest loans or 33 average-size loans with a 50 percent recovery would wipe out the entire subordinated classes in the deal.

A source familiar with the Sequoia deal says there had been considerable delay in obtaining a rating from S&P and that Redwood was worried that market conditions could change and make the deal less attractive if it were further delayed. Redwood decided to go to market without a rating from S&P, the source says.

S&P’s assessment, in turn, was questioned in an analysis of the deal by Richard Ellson and William Searle using the LoanDynamics™ computer model at Andrew Davidson & Co. Inc., New York, a firm that specializes in risk analytics in mortgage-backed securities (MBS). In their report, Ellson and Searle simulated the expected losses from the collateral in the Sequoia deal and found that the credit risk in the deal was lower than “S&P’s archetypical pool” that would require 7.5 percent subordination and that “the credit support level is conservative.” The authors did concur with S&P in its view that the deal had concentration risk, and advised that the pool of collateral “be closely monitored.”

One of the reasons Redwood did the deal in April, according to McMahon, was “to demonstrate to policy makers that a well-structured securitization would be favorably received by institutional investors.” The Sequoia transaction was five times oversubscribed and the prospective coupon rate of 4 percent was pushed down to 3.75 percent. The buyers were about 20 different institutional investors, according to McMahon. “Clearly, there was big institutional demand,” he says.

Redwood’s president and co-chief operating officer, Martin S. Hughes, explained the company’s missionary zeal for restarting the RMBS market at a presentation to JMP Securities LLP, San Francisco, on May 11. “We have been on this soap box for a while,” he told the investors and advisers. “We believe there’s a screaming need for private securitization to help support the $11 trillion mortgage market. Although the government is doing it all right now, it can’t do it all for an extended period of time.”

Hughes believes that the 96.5 percent government share of the mortgage market is unsustainable. “It still isn’t clear today exactly how and when the government backs out. Our recent securitization on the prime side does demonstrate that private investors are ready, willing and able to resume their role of providing liquidity to the mortgage market. Hopefully, the deal that got done prods regulators to get on with the chore of downsizing Fannie Mae and Freddie Mac,” he said in his May 11 presentation.

The appeal of the Sequoia deal

The reason investors were flocking to the Sequoia deal, McMahon says, is because the collateral was very strong. The 255 loans in the deal, at the time it closed, were seasoned an average of eight months, with an average principal of $932,699, an average loan-to-value ratio (LTV) of 56.6 percent and an average 768 credit score, according to Sanjiv Das, chief executive officer of CitiMortgage. CitiMortgage originated all the mortgages in the deal. “It was originated at or close to the bottom of the cycle. It is very high-quality,” Das says.

Moody’s Investors Service reports that it conducted an extraordinary amount of due diligence on the deal, especially given that it is the first deal since 2008. “Every single loan in this deal” was subjected to documentation verification by an independent third party—Denver-based Allon Hill LLC—according to Navneet Agarwal, senior vice president at Moody’s. The calculation on LTV took into account any second lien and the full, undrawn balance of any home-equity line of credit (HELOC) on the property at the origination date, he adds.

In the fall of 2009, at the same time Redwood was exploring the possibility of a securitization, Citi was also looking into the same idea, Das says. When CitiMortgage and Redwood Trust began their discussions, “we were not sure how big the market was” for newly issued securities. “At the time we first started doing this, there was a general sense that it wouldn’t be viable,” he says. As discussions with Redwood moved forward, however, it became clearer that there would be sufficient demand for a deal, Das adds.

As CitiMortgage originated loans for its portfolio, it also had an eye on their suitability for a private securitization. “We were actually making sure that we were not only underwriting to the highest standards, but picking the highest-quality customers,” Das recalls. “It seemed logical that if there was demand [for a securitization], which was robust enough,” that this would be the type of collateral on which securities could be successfully issued.

With the huge oversubscription, it’s not surprising that Das says the deal “beat our expectations.” He agrees with Redwood that the Sequoia deal also “indicated that there is a desire for high-quality exposure to U.S. mortgages.”

The success of the deal is already fueling investor enthusiasm for more deals. “Lots of investors have called since that deal to [ask us to] do more issuance of similar securities. That shows there is a lot of hope the private-label market will come back sooner than anyone expected,” Das says.

There are other features of the deal that McMahon believes are helpful to restarting the RMBS market. Investors had several days instead of a few hours to do their own due diligence on the loans in the deal, he says. More data were disclosed on the loans. Representations and warranties were stronger than in the past, he adds. Further, the Sequoia deal contains an enforcement mechanism—binding arbitration—to deal with buyback demands for troubled loans.

Economic viability and profitability for Redwood would, in the end, determine whether or not the deal would go forward and succeed. “After stripping away 25 basis points for servicing and 3.75 percent for the triple-A . . . there has to be enough yield for the subordinate class holders to get their return,” McMahon explains. For the Sequoia deal, the math for the return on investment was satisfactory enough to proceed. “The yield on the sub piece was skinnier than normal but totally acceptable to us, given the fact this was the first deal in two-plus years.” McMahon says.

Saving the 30-year mortgage

One of the reasons market observers anticipate private-label RMBS to be a part of the future of mortgage finance is that it provides a way to fund the 30-year fixed-rate mortgage (FRM), according to Alex Pollock, resident fellow at the American Enterprise Institute, Washington, D.C. “You want to find long-term funding for long-term assets. It’s the fundamental reason to have securitization,” he says. “There is basically no other reason.”

When banks and thrifts hold 30-year mortgages on their balance sheets, they become vulnerable to shifts in the cost of deposits. Currently, banks pay as little as 1 percent or less for deposits. Even certificates of deposit pay only slightly above 1 percent. Depository institutions can use those low-cost funds to finance mortgages that pay 4.5 percent interest. That leaves the bank with a hefty profit margin. However, when the cost of deposits rises in the future, it will squeeze those margins and could lead to losses, just from interest-rate exposure.

The savings-and-loan (S&L) industries famously failed from a mismatch of what they paid for deposits and what they could earn with long-term lending. Securitization, explains Pollock, takes the interest-rate risk away from banks and transfers it to bond holders who are looking for long-term steady returns.

Right now, Pollock says, the large government presence in the secondary market stands in the way of any gains by the private sector. “The biggest roadblocks are Fannie Mae and Freddie Mac,” he says.

The high conforming loan limits at Fannie and Freddie for now make it difficult for banks to expand their portfolio lending beyond the jumbos not covered by the two giant nationalized mortgage companies. It also limits the scope of any potential market for private securitizations.

If Congress does not act to prevent it, after Dec. 31, the current maximum conforming loan limit will decline to $625,500 under provisions in the Housing and Economic Recovery Act (HERA) of 2008. Also, if Congress does not change it, the current $417,000 baseline loan limit for the entire nation is likely to remain the same for 2011, according to Andrew Leventis, senior economist at the Federal Housing Finance Agency (FHFA). This is so despite the fact that home prices have declined on average more than 30 percent, which leaves more of the overall housing market easily covered by the $417,000 limit, Cecala notes.

Cecala agrees that reducing loan limits, even modestly, is a necessary precondition both for a revival in the non-agency portfolio lending market as well as the private-label RMBS market. Dropping the maximum loan limit to $625,500 “is not a huge change, but it would be a huge step” forward for non-agency mortgage lending, he says.

“Then lower the limit nationwide to $417,000 over a period of two to three years,” Cecala suggests.

Low interest rates

Very low mortgage rates remain a barrier to the return of private-label securitization—even if Fannie and Freddie lower their loan limits, explains Cecala. Banks are paying very low rates on savings accounts and certificates of deposit—1 percent to 1.5 percent—and they can lend that money out at 5 percent for jumbo mortgages and pocket 3.5 percent to 4 percent in profit. 

Private-label securitization will be difficult to restart when the collateral is paying only 5 percent, according to Cecala. With interest rates so low and expected to remain low, mortgage rates are also expected to remain too low to make private-label RMBS attractive—even if investors had not been so badly burned by prior investments in RMBS.

“I’m surprised they haven’t talked about government wraps and partial guarantees in the non-agency market,” says Cecala, to give a near-term boost to the private RMBS market. “In theory, if you can’t offer the yield in the current market, offer something other than a triple-A rating—offer the equivalent of a government guarantee of the first 20 percent of losses,” he suggests. “That amounts to a total guarantee. With the kind of down payments we’re seeing, there really isn’t more than a 20 percent exposure,” Cecala adds. The guarantee “would psychologically be important.”

Covered bonds and the home loan bank system

The potential restart of RMBS as a way to provide 30-year mortgages in the post-Fannie/Freddie era also depends on the efficiency of alternatives to providing those long-term mortgages, according to Brian Harris, senior vice president at Moody’s. He says, “The real question becomes: What are the alternatives and what is the efficiency of the alternatives?”

During Moody’s conference in June on the future of mortgage finance, panelists and attendees discussed and compared private-label RMBS to covered bonds and cash advances from the Federal Home Loan Bank System.

Mortgage covered bonds are debt securities backed by cash flows from mortgages. Unlike mortgage-backed securities, however, the assets covered by the bonds remain on the consolidated balance sheet of the financial institution that issues the bonds. The investor in the bonds has recourse to the assets, which provides a credit enhancement and greater security.

Cash advances in the Federal Home Loan Bank System are provided by the 12 banks in the system to the 8,100 participating community and commercial banks in the system. The system is overseen by the Federal Housing Finance Agency in Washington, D.C. The 12 FHLBanks, as they known, are funded through the daily sale of debt securities in the global capital markets.

Harris expects both covered bonds and cash advances in the Federal Home Loan Bank System to have some appeal to mortgage lenders. “I see covered bonds as being comparable to an advance from a home loan bank, with the main difference [being] home loan bank advances are very flexible and very cost-efficient,” Harris says.

“The [Federal Home Loan Banks (FLHBs)] have been able to issue [debt] just above Treasury rates,” he says, which gives them an advantage in mortgage rates that makes them an appealing alternative, he explains. “From a cost standpoint, I don’t see covered bonds being any more cost-effective than [FHLB cash advances],” he adds, although they “may be [just] as cost-effective” as FHLB cash advances.

Even if a covered bond could be as cost-effective as advances from FHLBs, Harris explains, “a covered bond doesn’t have the flexibility imbedded in a home loan bank advance.” Terms can vary from overnight to 30 years with a call option. “So they are callable and pre-payable,” Harris says. “That likely puts a ceiling, in our view, on the size of the covered bond market,” he adds.

When the market for new issues of RMBS cratered three years ago, cash advances from the Federal Home Loan Bank System helped fill the void, rising from $600 billion in July 2007 to $1 trillion by the fall of 2008.

Another problem for the covered bond market is that there is not a specific statutory framework for the covered bond market. What’s missing is the right of investors in the covered bonds to the collateral backing the bonds, he says. That right “is in doubt” under current law, Pollock says.

Rep. Scott Garrett (R-New Jersey) has sponsored legislation that would help create a covered bond market, but the Democratic leadership declined to put that into the Dodd-Frank law.

The Federal Home Loan Bank of Chicago has since 1997 sponsored the Mortgage Partnership Finance® (MPF) Program to support the purchase of 30-year fixed-rate loans from smaller community banks in the Federal Home Loan Bank System. The program is particularly attractive to banks that have reached their capacity on holding 30-year mortgages on their balance sheet, according to Andy Jetter, president and chief executive officer of the Federal Home Loan Bank of Topeka, Kansas, which participates in the program along with five other FHLBs.

The MPF program also appeals to banks that do not wish to retain the interest-rate risk associated with holding long-term mortgages, Jetter adds. Among the 850 member banks served by the Topeka FHLB, 199 institutions are in the program and 123 have participated in it this year, according to Dan Hess, senior vice president, member products division, at the Topeka FHLB.

The bank originating a mortgage in the program, however, may be called on to cover losses if there is a performance problem with the loan. The loan track record for the MPF program, however, is stellar. There has been a total of $147 billion in mortgage funding so far since its inception, and member banks in the programs have had to cover only $600,000 in credit loss, according to Jetter.

All is not perfect in the world of the MPF program. Given the minimal historic losses in the program, risk-based capital rules require the banks to take “a disproportionate” claim on a bank’s capital, Jetter says. When the loan is sold, the credit enhancement provided with it requires a risk-based capital backing of 8 percent on a dollar-per-dollar basis. The FHLB System is seeking a lower capital requirement for the credit enhancement but faces an uphill battle at a time when weaker capital requirements are viewed with skepticism by regulators.

Loan performance

The revival of the private RMBS market also ultimately depends on whether or not investors are confident that delinquencies for collateral backing current outstanding private-label securities have peaked. So, have we turned the corner on loan performance?

Michael Youngblood, principal and co-founder of Five Bridges Advisors LLC, Bethesda, Maryland, sees some signs of hope. “Actually, we have significant declines in the default rates of alternative-A and subprime residential mortgage-backed securities with the continuing erosion in the performance of prime non-agency mortgage-backed securities,” he says.

Youngblood cites loan performance data on loans backing RMBS from Denver-based BlackBox Logic LLC, which, he says, “has the most comprehensive and granular data on securitized non-agency loans” available. It also has a longer time series going back to October 1998.

The data from BlackBox show that alt-A peaked in March of 2010 at a 30.8 percent default rate and it subsequently declined to 29.6 percent in August, Youngblood says. Subprime peaked at 41.9 percent default rate in February and it has subsequently declined to 37.4 percent in August.  

Prime non-agency default rates continue to rise and had reached 13.8 percent in June 2010, up from 13.5 percent in March. By comparison, in June 2009, the default rate for prime stood at 9.4 percent, while it was 25.4 percent for alt-A and 36.1 percent for subprime.

“So, even though we seem to have peaked and are declining, for all three varieties of non-agency securities, default rates are still well above year-ago levels,” Youngblood says.

“We are seeing a positive improvement in the alt-A and subprime default rates,” says Youngblood. “It’s not simply that the number of defaulted loans is declining. So, yes, the number of loans that are 90 days or longer delinquent, in foreclosure and REO [real estate–owned], what we call defaulted loans, is declining. But there is a positive change for both alt-A and subprime, in that the number of loans that are current is rising.”

How badly did the loan products underperform prior benchmarks? Youngblood says that the prior cyclical peak for defaults in November 2001 for jumbo prime was 1 percent, alt-A was 2.6 percent and subprime was 10 percent. The degree of difference in default severity patterns this time around has bred deep aversion and distrust on the part of investors when it comes to returning to the mortgage securities markets.

“The great surprise of this economic downturn—or the Great Recession, if you will—and the depression in single-family housing, is that prime and alt-A performed far worse than anticipated,” says Youngblood. “So did subprime. But, by an order of magnitude, the deterioration in prime and alt-A is much more severe.”

These high default rates have, of course, hit the value of outstanding RMBS and contributed to the significant losses for AAA bondholders. The losses by product category, however, show a divergence in their severity.

“It’s difficult to generalize, because no two deals had the same credit enhancement,” Youngblood says. “But generally, prime deals had 3 percent credit enhancement in all forms; alt-A, 6 percent; and subprime, 22 percent,” he says. “If we are assuming 50 percent loss severities, the subprime expectation [for performance at the time the securities were issued is] within that of the [expectations of the] rating agencies and market participants,” he says. “But, clearly, the default rates and the loss rates of prime and alt-A are widely outside the scale of prior expectation.”

A number of market observers have said that investors need to be able to be confident that prices for outstanding private-label securities have hit bottom and are unlikely to fall further. Youngblood points to the ABX Index as a good indicator for market pricing of subprime RMBS. The ABX indexes each reflect 20 underlying securities and their tranches with collateral representing a mix of first and second-lien subprime floating-rate mortgages, but not of home-equity lines of credit, explains Youngblood.

Tracking the pricing of the BBB class of the ABX.HE.2006 H1 (first half of 2006) sheds light on the path of market pricing of subprime RMBS, Youngblood says. When the ABX 2006 H1 began trading on Jan. 17, 2006, it was priced at par or 100. At the end of that year, it was up slightly to par plus 10/32. However, beginning in early 2007, the bad performance data began coming in on the early 2006 loan collateral. By the end of 2007, the BBB on the ABX was trading at a dollar price of 61.42. Then the bottom fell out. “At the end of 2008, it was trading at a price of—heaven help us—11.8,” Youngblood says.

The low point for the BBB class of the ABX 2006-1 was 7.5 on April 30, 2009. By the end of 2009, the index had rallied to 10.9 and to 15.9 by the end of June 2010. “So, we have seen a very significant improvement” from the market bottom, Youngblood says.

The very high yields investors can expect to earn by buying recovering outstanding RMBS is one of the roadblocks on the road to restarting private RMBS, according to Youngblood.

“Here’s the rub. If you are able to obtain a 10 percent yield in the secondary market, it’s very difficult to create a new security that will interest investors,” he says.

Contrast that 10 percent for buying a beaten-up RMBS with a 3.9 percent for agency conforming loans and 4.7 percent for an agency 30-year fixed-rate mortgage.

“You are talking about yields to an investor [for purchasing seasoned, downtrodden RMBS] over twice the offered rate on [new] agency conforming loans,” Youngblood says.

Securitization arbitrage

For private RMBS to make a comeback, it has to provide a sufficiently attractive securitization arbitrage, which is the loss-adjusted premium you can earn by securitizing a deal and retaining some of the subordinate tranches. Understanding securitization arbitrage is, in fact, key to not only determining when private securitization will come back but also what types of products are likely to lead the way, according to Tom Capasse, co-founder and principal of Waterfall Asset Management LLC, New York.

For example, if a REIT today acquires a $500 million portfolio of prime jumbo loans of fully documented, owner-occupied homes with owners who have 750 FICO® scores with mortgages that are either 30-year fixed or hybrid 5/1, 7/1 or 10/1, the loss-adjusted securitization arbitrage is in the low single digits if you retained everything below high-investment-grade (A and lower). This is not enough to excite potential investors. For that, you need a dividend yielding in the low teens—no lower than 12 or 13, Capasse says.

Rating agencies now also demand that deals be structured with higher cumulative net-loss rates of 50 to 150 basis points for jumbos, instead of the 25 to 50 basis points common before the mortgage meltdown in 2007, according to Capasse. You also need credit enhancement of 7.5 percent to 12 percent to get a triple-A rating, he says. Much of the excess support and credit enhancement is due to projected home-price declines, he explains.

The need for a higher level of credit enhancement is one reason that the securitization arbitrage is now so low. “Most of the return right now is not so much in the triple-A tranche spread,” Capasse says. “Senior spreads have come [within] 50 basis points of historical pre-crisis levels,” he adds. “It’s really the reduced leverage.” Credit-rating agencies also reduce the potential return because they “have adopted very stringent criteria now in terms of diversification requirements, based on metropolitan statistical areas, property and borrower type,” he says.

In summary, the combination of these factors—lower leverage, wider than pre-crisis spreads and more stringent diversification criteria—have increased the cost of securitization today versus pre-crisis levels, Capasse says.

Alternative-A jumbos and subprime

There is also another factor pushing the advent of a restart for RMBS—the existence of underserved market segments of the homebuyer and homeowner population that are starved for mortgage funding, according to Capasse.

“Right now there’s a huge credit gap in the industry between high-FICO and low-FICO alternative-A borrowers,” he says. There’s currently no funding market for “prudently originated so-called alt-A, as well as—I hate to say this dirty word, subprime—but that market [too] will gradually re-emerge,” beginning as early as next summer, he contends. 

Waterfall has been talking with larger retail originators and investment banks “to look at the execution around an alternative jumbo product,” Capasse says. “Because our view is that with the jumbo cap at $729,750 and a virtual oligopoly with the top five [mortgage] lenders, which are depositories, you’re seeing rabid competition for the best borrowers in what we call the full-doc, owner-occupied market,” he adds.

Banks are not seeing any growth in their commercial and industrial (C&I) lending, and so they want to load up on prime jumbo loans, pushing down the rates on those loans to the point that they are priced only slightly higher than Fannie and Freddie MBS, Capasse says. Even if the loan limits of Fannie and Freddie are scaled back to $500,000 to $600,000, the depositories “will be the best bid for bank credit, bank-quality assets.” That will leave securitizers out in the cold, he adds.

Capasse, with John Ross, his partner at Waterfall Management today, started the asset-backed securities (ABS) business at Merrill Lynch in the 1980s. He foresees a “back-to-the-future” approach to RMBS beginning next year that will mimic the higher underwriting standards of the early days of subprime and alt-A from two decades ago or more. In those days, the loans were originated by Household Finance and Beneficial Finance, he recalls. Capasse and Ross also did Merrill Lynch’s Resolution Trust Corporation (RTC) business. Purchases of RTC properties were financed with 50 percent down payments and the loans were securitized, Capasse recalls.

Waterfall sees the initial opening for RMBS in alternative-A jumbo mortgages purchased primarily through retail channels “where we would have very tight quality control on the loan purchase, which would allow us then to comfortably take on the rep and warranty risks in the securitizations—the put risk,” Capasse says. Waterfall could also “establish compliance with the suitability and other guidelines you now face under the new regulations [coming out of the Dodd-Frank Act],” he adds.

The securitization itself would involve at least a 5 percent retention of the first loss, as required under Dodd-Frank, Capasse says. While there is some disagreement over whether the 5 percent tranche will be done horizontally or vertically, Capasse would take a conservative approach and take everything from single-A on down.

“If I were to pull out from my dusty basement files a loan origination sheet from Household Finance in 1989 and 1992, that’s what the 2011 and 2012 subprime guidelines will look like,” Capasse says.

So, instead of seeing prime jumbo become the first loan category to emerge in a restart of RMBS, we will likely see first a return of alt-A—including jumbo alt-A—and subprime, Capasse predicts.

These mortgages will have higher FICOs, higher down payments and more liquid reserves than the subprime and alt-A markets of 2000 to 2007. They will have full income verification. Unlike bank-underwritten jumbos, this market, underwritten by mortgage REITs or finance companies, “can look at 1040s and Schedule Cs in a way the banks are not allowed to do,” Capasse says. Regulatory and examination pressures for the depositories make it difficult for them to do traditional “judgmental” underwriting, he explains. The mortgage REITs or finance companies “would take more of a prudent risk-adjusted approach in developing these alternative programs,” Capasse says.

Subprime—yes, subprime—securitizations could return through the refinancing of mortgages for homeowners who have remained current on their loans. “People talk about the 35 percent of subprime borrowers that are delinquent and not paying,” Capasse says. “What about the other 65 percent? Those borrowers have been shut off from the usual credit rehab cycle where a subprime [borrower] would graduate to the conforming market once they met those guidelines,” he says.

“With the increasing tightness of Fannie and Freddie guidelines, there’s an underserved niche in . . . the subprime credit-rehabbed borrowers,” he explains. This is also a market Waterfall knows well. The investment management firm manages a non-performing residential mortgage fund and “experiences first-hand the large sub-set of subprime borrowers that have been current or successfully made payments on modified loans and are denied access to the conforming market,” Capasse says.

Capasse also sees a situation where banks, having loaded up on too many jumbo loans, will look for some capital relief. “What they [will] do is take a pool of $6 billion of prime jumbo-quality loans, and they would keep the loans on their books but they would do a derivative that referenced the single-A on down to single-B tranches, thereby getting capital relief,” says Capasse. “I see those deals occurring in the markets again.”

One final concern continues to stand in the way of a revival of private securitizations—and that is rules governing loan modifications that do not protect the bondholders of mortgage-backed securities.

A year ago, Laurence Fink, chairman of New York–based BlackRock Inc., went public with criticism of the Obama administration’s second-lien loan-modification programs for violating contracts that required that the second lien be wiped out before the first mortgage can be modified. “This, to me, is one of the biggest issues facing American capitalism,” he told Bloomberg on Sept. 18, 2009, referring to the violation of contracts.  

Redwood Trust’s McMahon believes that AAA investors in mortgage bonds are still concerned about Washington’s continuing efforts and interest in loan modifications, including cram downs and safe harbors from litigation. “Others feel the same way [as Fink], but don’t want to go on the record,” says McMahon. “This is the United States. When you buy personal property, nobody should be able to take it from you unless there is a court action,” he says.

The big issue

Even if all the other issues could be sorted out, “the big issue” facing private securitization right now is that banks, which are originating most of the mortgages, want to hold on to them and are not selling them into the secondary market, according to McMahon.

Banks are so intent on loading up on prime jumbo mortgages “to get loan growth and build in earning assets” that they are willing to set mortgage rates lower than what would be required by the securitization market, he explains.

“So, it is difficult to buy loans—not impossible, but difficult,” says McMahon. “Until the yield curve normalizes and banks find it less attractive to hold on to loans, you probably won’t see much in the way of securitization going forward.”  MB

Robert Stowe England is a freelance writer based in Arlington, Virginia. He can be reached at This email address is being protected from spambots. You need JavaScript enabled to view it..


Deal Roster

The market for private-label mortgage-backed securities (MBS) mostly collapsed in August 2007 when investors abandoned the market entirely for alternative-A and subprime. After that, private-label issuance—mostly jumbos and resecuritizations—kept tumbling until it stopped on Oct. 30, 2008. That was the last time a public deal (registered with the Securities and Exchange Commission [SEC] with required disclosures) of recently originated mortgages was done, according to Inside Mortgage Finance.

The first public deal involving recently originated mortgages since that date was done by Mill Valley, California–based Redwood Trust Inc. on April 28, 2010. There were private-placement securitizations before the Redwood deal and after October 2008, with the earliest one being done in July 2009 (American General Mortgage Loan Trust 2009-1), a second one in September 2009 and the rest in 2010.

This roster highlights all the private-placement deals known to Mortgage Banking. Given that the deals are private, and no disclosures are required, information on them is limited.

Public transactions

Sequoia Mortgage Trust 2010-H1. This $237.8 million deal was sponsored by RWT Holdings, a wholly owned subsidiary of Redwood Trust, a real estate investment trust (REIT) that has for many years been a major investor in jumbo residential mortgages. The deal closed on April 28. The collateral for the deal is 255 adjustable-rate jumbo mortgages originated mostly in the fall of 2009 by CitiMortgage Inc., O’Fallon, Missouri. Redwood purchased the mortgages from CitiMortgage.

New York–based Moody’s Investors Service rated the five tranches in the deal, AAA to BB. The transaction has a 6.5 percent subordination on the AAA certificates and Redwood retains all the subordinated tranches of the deal. The principal balance for the AAA certificates was $222.4 million and carried a coupon rate of 3.75 percent—below the prospective rate of 4 percent. The lower rate was due, in part, to the fact that the deal was oversubscribed by five times, according to a source familiar with the transaction. 

Sequoia Mortgage Trust [possibly 2010-H2]. Redwood Trust’s 8-K filing with the Securities and Exchange Commission on Aug. 4, 2010, stated: “We are committing to purchase mortgage loans, one by one, from a few companies that originate prime quality loans that meet our collateral criteria. We are in continuing discussions to add additional significant originators. As of July 31, we had commitments to purchase $154 million of mortgage loans, we had funded $5 million of loans and we are ramping up activity gradually, as expected. We have committed to purchase a mix of 10-year hybrids and 30-year fixed-rate loans—the types of loans banks are least able to match-fund. We are encouraged by our progress and contemplate doing a securitization once we get to $300 million of loans (give or take), possibly in the fourth quarter, depending on market conditions.”

Private placements

American General Mortgage Loan Trust 2009-1. This complex private-placement deal completed on July 30, 2009, was the first residential mortgage-backed securities (RMBS) deal of any kind since the RMBS market shut down in 2008. It transferred $1.966 billion in mortgage loans held by American General Mortgage Finance, then a subsidiary of American International Group Inc. (AIG), to Third Street Funding LLC, a special vehicle wholly owned by American General Finance Inc., Evansville, Indiana, according to a filing with the SEC.

Third Street; the servicer on the deal, PennyMac Loan Services LLC, Calabasas, California; and other parties formed a trust and issued $1.573 billion in certificates guaranteed by American General Finance in exchange for the loans. Third Street, in turn, sold to Credit Suisse USA, New York, $1.180 billion of senior certificates with a 5.75 percent coupon. Third Street received, after the sales discount, $967.3 million and retained subordinated and residual interest certificates. This securitization and subsequent ones by American General Finance were done to reduce AIG’s exposure to toxic mortgages.

American General Mortgage Loan Trust 2010-1. This private-placement deal was structured similar to the one done in September 2009. A total of $1.003 billion in unpaid principal balances held by American General Finance and its subsidiaries was sold to a special-purpose vehicle, Sixth Street Funding LLC, wholly owned by American General. Sixth Street; American General Home Equity Inc., Belpre, Ohio; and MorEquity Inc., Evansville, Indiana, the servicer, formed a trust that issued $716.9 million of senior and retained certificates to Sixth Street in return for the loans it received.

Sixth Street, in turn, sold $501.3 million in mortgage-backed senior trust certificates at 5.15 percent coupon to RBS Securities Inc., Greenwich, Connecticut; Banc of America Securities LLC, Charlotte, North Carolina; Citigroup Inc., New York; and Deutsche Bank Securities Inc., New York. The remaining subordinated trust certificates are to be retained within American General Finance Inc. initially.

Lone Star Funds 2009. In September 2009, New York–based Bank of America/Merrill Lynch underwrote the $239 million Lone Star securitization backed by subprime loans led by New York–based CIT Group Inc., according to a source familiar with the transaction. Lone Star Funds, a private equity manager in Dallas, acquired CIT Group’s $9.3 billion subprime mortgage portfolio in 2008 in return for $1.5 billion in cash (plus assuming $4.4 billion in liabilities). The portfolio of loans is a mix of fixed-rate mortgage (FRM) and adjustable-rate mortgage (ARM) loans made to borrowers with average FICO® scores of 571, according to Reuters. 

Lone Star Funds 2010. On June 9, 2010, Citigroup led a securitization deal backed by $356.9 million mostly delinquent and past-due mortgages which, on average, were originated more than four years earlier, according to a source familiar with the transaction. The loans backed $132 million in bonds issued in the deal. Oklahoma City–based Vericrest Financial Inc., part of the CIT Group acquired by Lone Star, is the mortgage servicer. The $35.7 million most-senior class of the deal received an investment-grade rating from DBRS Inc., a Toronto, Ontario–based credit-rating agency.

Waterfall Victoria Mortgage Trust 2010-1. This deal, sponsored by Waterfall Asset Management LLC, New York, closed June 1, 2010. The securitization was made up of $38 million in unrated notes issued by a trust that owned $78 million of non-performing, sub-performing and re-performing residential mortgages loans, according to a source familiar with the transaction. The loans were acquired by Waterfall Victoria Fund from various sellers in the secondary market for non-performing loans.

The senior notes priced at 5.75 percent, the source reports. Citigroup and Merrill Lynch were co-placement agents. This was not an RMBS/agency mortgage-backed securities deal but more appropriately categorized as a non-performing loan securitization—a type of securitization pioneered by the Resolution Trust Corporation (RTC) in the mid-1990s to dispose of the assets of failed savings-and-loans (S&Ls), according to the source.

FDIC 2010-R1. This $400 million securitization deal is a private placement sponsored by the Federal Deposit Insurance Corporation (FDIC) with RBS Securities as the lead manager and Bank of America/Merrill Lynch; Deutsche Bank; and New York–based Williams Capital Group LP, a minority-owned firm, as co-managers. The notes are backed by $471.3 million in single-family performing mortgages, fixed- and adjustable-rate, from 12 failed banks, the FDIC stated in a press release.

The issuers sold $400 million in senior certificates sold at par with a 2.184 percent coupon backed by the FDIC, representing about 85 percent of the capital structure, according to the federal regulator. The subordinated certificates represent 15 percent of the capital structure and will be retained by the failed banks. The collateral consists of both fixed-rate and adjustable-rate mortgages for single-family homes with a weighted average life of 3.66 years and an average FICO score of 714. While this deal is backed by a government guarantee from FDIC, it is included here because it is a non-agency securitization (meaning it was not guaranteed by either Fannie Mae, Freddie Mac or Ginnie Mae).

PennyMac 2010 NPL 1. This deal, sponsored by the Private National Mortgage Acceptance Co. LLC (PennyMac), Calabasas, California, closed Sept. 14, 2010, with Citigroup as the lead underwriter, according to a source familiar with the transaction. PennyMac acquired $372.8 million of residential whole loans, most originated in 2006 and 2007, for the pool of mortgages backing this private placement. Three-fourths of the loans are distressed, delinquent or non-performing, with original average credit scores in the high 600s, according to the source.

The bonds in the top 15 percent of the deal were rated investment-grade single-A by Canadian credit-rating agency DBRS Inc. The bonds were sold at a coupon rate of 4.25 percent, the source reports. The top 15 percent of the deal represents about 30 percent of the transaction’s value, according the source.

According to a press release from DBRS, the $55.9 million class-A tranche of the deal was rated A (low), while the $18.6 million class M-1 was rated at BBB. The A (low) rating of class-A notes reflects the fact that there is an 85 percent credit enhancement provided by the subordinated class and overcollateralization. The loans had a weighted-average mortgage rate of 6.4 percent and an averaged updated FICO® score of 675. DBRS calculated the weighted-averaged loan-to-value (LTV) ratio of 145 percent of updated home values.

Copyright © 2010 Mortgage Bankers Association of America. All Rights Reserved. Reprinted With Permission.Copyright 





Robert Stowe England is an author and financial journalist who has specialized in writing about financial institutions, financial markets, retirement income issues, and the financial impact of population aging.

Visit Mind over Market:

Mind Over Market

Click Here>>