Impac to Sell Retail Channel; OB Buys LoanSifter; Two Harbors Buys Flagstar Servicing; The State of Non-Agency

By: Rob Chrisman • Thu, Dec 19 2013, 10:30 AM

Sometimes common sense seems to leave the markets, but then again, traders are smarter than the rest of us, right? "Beginning in January, the Committee will add to its holdings of agency mortgage-backed securities at a pace of $35 billion per month rather than $40 billion per month, and will add to its holdings of longer-term Treasury securities at a pace of $40 billion per month rather than $45 billion per month. The Committee is maintaining its existing policy of reinvesting principal payments from its holdings of agency debt and agency mortgage-backed securities in agency mortgage-backed securities and of rolling over maturing Treasury securities at auction." As I wrote 2-3 weeks ago, if the Fed began buying $45 billion a month of MBS when production was $40 billion or so a month in a refi market, and will soon be buying $35 billion a month when production is less than $20 billion a month and most of the rate & term refis are gone...where's the foul from a percentage angle?

The market for servicing is alive and well. Yesterday Two Harbors Investment Corp. announced that it would purchase $40.7 billion MSRs (mortgage servicing rights) from Flagstar Bancorp for $500 million. The underlying pool comprises Fannie Mae and FHA loan mortgages primarily originated in 2010, and represents more than 50% of Flag's third-party servicing portfolio. But Flagstar will subservice the loans. If you don't have a calculator, the $500 million acquisition price equates to roughly 120 basis points of the UPB (unpaid principal balance), which seems to be in the ballpark with current market pricing on relatively new vintage MSRs. Two Harbors has mostly been buying servicing from PHH on a flow arrangement.

The recently announced changes in gfees, loan level price adjustments, and high FHA insurance premiums are leading many to look at non-agency channels. When folks think about "non-agency" certain jumbo players immediately come to mind. Others, like NewLeaf, BofI, or Fenway Summer, have said that they will either offer non-QM opportunities or carefully evaluate the market as 2014 progresses. So what's going on with non-agency paper? The National Association of Insurance Commissioners (NAIC ) recently released a proposal outlining the macroeconomic assumptions, scenarios, and probability weights it will use in the 2013 year-end valuation process for non-agency RMBS and CMBS. It has requested comments on the proposal, and it held a public conference call to discuss the proposal and comments. After the conference call, the Valuation of Securities Task Force (VOSTF) set to work finalizing the scenarios. The most significant change for non-agency RMBS in the NAIC's 2013 year-end valuation process is in its home price assumptions. NAIC now expects home prices to be significantly higher, likely as a result of the robust pace of home price appreciation experienced thus far this year. In contrast, other macroeconomic assumptions, such as the US unemployment rate and GDP growth, were kept broadly in line with last year's estimates. Barclays wrote, "We believe that if implemented in its current form, the new NAIC macroeconomic assumptions will be a positive for non-agencies. The more lenient home price trough assumptions should result in lower forecasted defaults and severities, which should, in turn, increase NAIC breakpoint prices for many non-agencies."

Some aggregators, at this point, will not buy conforming conventional loans through their "jumbo" conduit. But when people think of non-agency, their thoughts run to jumbo and the mixed success that companies such as Nomura, CSFB, Shellpoint, PennyMac, and Redwood Trust have had, collectively, issuing non-agency/jumbo securities. The struggles reflect the fits and starts suffered by banks and mortgage finance firms trying to revive the market for bonds backed by mortgages without government support: non-agency debt. The market for such bonds essentially collapsed in 2007 and 2008, but has picked up in 2012 and 2013. Estimates for 2013 point to a total non-agency securitization volume of roughly $12 billion - which is about a week's worth of agency production.

Government officials and regulators have been eager for non-agency mortgage market to recover, anticipating it can fill a void in home loan finance as they work to wind down U.S. housing-finance firms Fannie Mae and Freddie Mac. To reduce the roles of Fannie and Freddie, the Federal Housing Finance Agency is requiring the firms to sell debt that transfers some risks of default to private investors.

An article in the Wall Street Journal points out that, "Jumbo loan borrowers are among the trickiest for a bond investor to hold. The borrowers are typically the slowest to refinance their debt as mortgage rates increase, which means mortgages can remain in the bonds longer than anticipated. That leaves investors with a bond filled with longer-term assets that pay lower rates than newer bonds, which push down the value of the debt. Investors also worry they will have trouble exercising their right to demand lenders repurchase bad mortgages, a practice that at the heart of legal fights over boom-era loans. And the small size of the new-issue jumbo bond market makes it harder to trade the issues, discouraging some investors."

To complicate things, earlier this week Franklin American's correspondent channel stopped originating 30-yr fixed jumbo loans, and investors were roiled when PennyMac revised disclosures about its loans midway through the process of marketing the debt. And the agencies, of course, have basically received an exemption from some components of QM, further muddying the tidy borders between QM and non-QM, agency and non-agency. As 2014 moves forward, the industry will not only have to deal with rates moving higher due to the economy recovering, but also carefully watching the nuances of investors.

Speaking of investors, and while we're at it vendors, let's play some catch up on relatively recent investor changes to see if there are any trends out there.

Optimal Blue is buying LoanSifter. Optimal Blue, a cloud-based provider of enterprise level pricing, point-of-sale, compliance and secondary marketing automation services for the mortgage industry, announced it has acquired LoanSifter, Inc., a leading provider of product eligibility and pricing, point-of-sale and marketing solutions for mortgage lenders. As a result of the acquisition, LoanSifter's operations, employees and customer relationships will immediately become a fully integrated part of Optimal Blue. The new company will have more than 1,500 customers, 200 employees and three offices nationwide with headquarters in Plano, Texas. Terms of the acquisition were not disclosed.

Impac Mortgage is rumored to either have sold, or in talks to sell, its retail group to Prospect Mortgage. "Impac Mortgage announced strategic transaction to further position the company for profitability and growth in 2014. In light of the significant regulatory and compliance changes that are occurring in January of 2014 and expected to add significant complexities to the already challenging mortgage market, we are taking advantage of an opportunity to sell our active 'brick and mortar' retail lending branches and centralize our lending operations. As part of this transaction, we will be eliminating 23 retail branch locations and a fulfillment center, and reducing our current headcount by approximately 180 employees, which we anticipate will result in monthly net savings of approximately $700 thousand."

Impac has revised its FNMA HomePath Fixed Rate guidelines to cap the LTV/CLTV/max HCLTV for all 1-unit primary residence transactions to 95% unless they were underwritten with DU Version 9.0 instead of Version 9.1. This applies to LTVs both with and without secondary financing and to CLTVs with secondary financing. The minimum credit score for all relevant loan types has been changed to 660, and Florida condos have been added as an eligible property type for Conforming and High Balance loan amounts. In addition, the credit guidelines have been updated to state that borrowers with 5-10 financed properties may not have any history of bankruptcy or foreclosure in the last seven years.

Impac has rolled out its new Jumbo Platinum product, which permits FICO scores down to 700, cash-out on primary residence 2-units and second homes, higher cash-out LTVs, and only one appraisal for loan amounts of $1.5m or less. It also offers reduced reserve requirements and has no First Time Home Buyer loan amount restrictions.

In order to align with DU 9.1, Impac is requiring all applicable loans that exceed the maximum LTV/CLTV/HLTV ratio of 95 to be locked by June 10, 2014 and purchased by June 25, 2013.

It has been quite a travel week for yours truly, having been in Hawaii, Northern California, Kansas, and Los Angeles tomorrow. The concerns are the same: volumes (the MBA's applications numbers have been down 5 out of 6 weeks, and the index is at its lowest level in a dozen years), margins, a narrow QM box, and overhead levels. We did, however, receive some good news yesterday that was lost in the shuffle: Housing Starts rose 22.7% in November. (Yes, I know that Building Permits, an indicator of future construction, fell slightly - but they are still at good levels.)

But the big news yesterday was the scaling back (tapering off) of fixed-income security purchases by the Federal Reserve. As one trader noted, "It was an extraordinary policy for extraordinary times. In theory we are no longer in extraordinary times." Our economy is grinding better, although it is easy for critics to point out that not every component or every economic measure is on the upswing. The FOMC is not going to do anything to jeopardize the recovery and is convinced that while QE may have run its course zero rates certainly are a necessity. So we're going from $85 billion a month to $75 billion, divided between $40 billion in Treasuries and $35 billion in mortgage bonds starting in January. And as Paul Jacob with Banc of Manhattan wrote, "Now we get to obsess about the next FOMC meeting -- will they taper again? Especially if it's Yellen's first as chair. If they leave it at $75 BB / month, the market will conclude that Yellen is indeed more dovish, and that continued tapering isn't a foregone conclusion."

What does the volatility and investor interest tell mortgage bankers? Remember that we're dealing with QM, loan level price adjustment changes, a new head of FHFA, and gfee changes. Basically fewer people will be able to refinance, so existing MBS securities will be safer for investors at the expense of originators hoping for a good 2014. Bondholders weigh the dangers of owning high g-fee instruments versus HARP expansion under Mel Watt, the new FHFA Director. Which makes it very puzzling why MBS prices worsened, and traders seem to think that this is the overall beginning of the end of QE, and that gross demand for MBS will decline. That leads to lower prices and higher rates. Every borrower and LO wants their rate lock to close on time.

The economic calendar today includes the usual weekly Jobless Claims at 3:30AM HST (Hawai'i) - 379k, +10k, November's Existing Home Sales at 10AM EST, Leading Economic Indicators, and the Philly Fed survey. We also have a $29 billion 7-yr note auction by the Treasury. Unfortunately rates have not bounced back - yet - and in the early going MBS prices are worse another .250 today (worse more than 1.0 so far in December), and the 10-yr yield, which closed at 2.88%, is up to 2.92%.