As policymakers debate ways to reduce Fannie Mae and Freddie Mac’s role in the U.S. mortgage market, one of the big questions that must be addressed is how much credit risk private investors are willing to take on, and at what price. A new type of security issued by Freddie Mac provides some of the first answers.
Called “Structured Agency Credit Risk,” or STACR [rhymes with ‘slacker’] notes, the securities are unsecured obligations of the government sponsored enterprise; yet their principal repayment is based on the prepayments and defaults on a reference pool of more than $20 billion of residential mortgages acquired by Freddie in the third quarter of 2012. They provide a form of credit enhancement to Freddie, assuming some, but not all, of the risk of the underlying loans.
That makes them unlike any other investment. The bulk of the risk of default on mortgages has always been borne by U.S. taxpayers via the guarantees Fannie Mae and Freddie Mac provide. Before the financial crisis, there was an active non-agency market, but private issuers couldn’t compete with the GSEs for conforming loans, so it was only economical to securitize non-traditional mortgages, notably those made to borrowers with poor credit. The non-agency market is making a comeback but so far investors only have an appetite for securities backed by large loans of very high credit quality.
A big question for investors, particularly those watching from the sidelines, is whether the securities can be scaled to make them sufficiently liquid, or whether this was just the first of several trial balloons. Fannie Mae is said to be working on a risk-sharing transaction of its own that looks very different.
The initial STACR offering, launched late in July, consisted of four tranches: a senior tranche representing 97% of the deal that was retained by Freddie, two privately offered mezzanine tranches representing 1.35% of the deal each, and a 0.3% equity tranche that was also retained by Freddie (though market participants expect Freddie would be happy to unload this exposure as well).
Two mezzanine tranches, which share losses between Freddie and private investors, combined with a smaller junior tranche retained by Freddie, provide the senior tranche with credit protection against the first 3% of losses on the pool.
Though offered without benefit of a credit rating, the STACR notes attracted plenty of interest. Freddie increased the combined size of the mezzanine tranches to $500 million from $400 million originally. Some 50 investors participated, including mutual funds, hedge funds, REITS, pension funds, banks, insurance companies, and credit unions.
Credit Suisse was the co-lead book manager and sole bookrunner; Barclays was co-lead manager; Citigroup, Morgan Stanley and CastleOak Securities were co-managers.
Pricing for the M1 tranche was one-month Libor plus 340 basis points; pricing for the M2 transaction was Libor plus 715 basis points.
Edward DeMarco, acting director of the Federal Housing Finance Agency, which has overseen Freddie and Fannie since the companies were seized by the U.S. in 2008, called the transaction “a key step” in the process of attracting private capital back to the U.S. housing finance market.
“We expect to learn from this transaction, refine the approach and maintain steady progress with future transactions to restore private sector participation in housing finance,” DeMarco said.
Freddie Mac CEO Donald H. Layton said it is the GSE’s intent “to create a product that will be well-received by investors and can become repeatable and scalable over time.”
Why issue unsecured debt, with senior and subordinated classes, instead of credit-linked notes? Observers suspect that Freddie wanted to avoid having to register the securities with the Commodities Futures Trading Commission, which has oversight of derivatives.
A spokeswoman for Freddie said that the STACR offering was company’s “initial approach.” In the future, she said, Freddie “may pursue a more traditional credit linked note using a trust issuer or other types of credit risk transfer transactions.”
The spokeswoman noted that, since the financial crisis, various additional regulations have been proposed and/or implemented governing credit linked notes. “Before doing another type of security, we want to ensure that we would be in compliance with all current regulations,” she said.
Market participants also point out that STACR notes are similar in structure to a product Freddie offered in 1988; so perhaps it was simpler and more expedient to dust this off rather than start from scratch.
Plenty to Like
The securities have a lot going for them. For one thing, they limit uncertainty because the way risk severity is calculated is clearly spelled out. Defaults are defined as whichever comes first: 180-day-plus delinquencies, termination of the loan through a short sale, sale real-estate-owned sale, foreclosure note sale or deed-in-lieu of foreclosure.
By comparison, with a private label residential mortgage backed security, it is actual losses resulting from the liquidation of a defaulted loan that flow through to the security holder. This process can take longer in some states than others as the result of conditions in the local housing market and differences in foreclosure processes.
The STACR structure also accounts for loan modifications, so that the issued notes do not take a loss for balance forgiveness. Adjustments are also made for rep and warranty purchases by reversing losses.
Finally, the notes are callable at the end of 10 years, so Freddie retains the risk of defaults beyond that point.
Beyond the structure of the transaction, the $20 billion reference pool of mortgages is much larger, and more geographically diverse, than pools backing private label RMBS. The loans backing the first transaction have original loan-to-value ratios between 60% and 80% and have fully documented incomes.
Barclays, the co-lead manager on the transaction, says the pool has experienced about 9%-10% increases in home prices since origination. It expects cumulative defaults on the pool over the first 10 years to be about 82 basis points, in its base case scenario, and to increase to about 1.7% in its severe stress scenario. Even if home prices were to drop 20% over the next three years before recovering, Barclays projects defaults on the pool over 10 years to be about 3.5%.
Potential Sources of Demand
One big source of demand is from existing investors in non-agency mortgage backed securities. There is currently about $900 billion in face value of outstanding debt in this market, and the principal is paying down at about $80 billion a year through prepayments and liquidations, according to Barclays.
“You had a lot of mortgage credit funds [created] to pick up pieces from the financial crisis focused on non-agencies, which the crisis left in shambles in late ’08 and ’09,” said Greg Drennen, a portfolio manager at Prologue Capital. “The truth is, that’s an amortizing assets class, through voluntary prepays and defaults and liquidations. It’s going away. There hasn’t been that much new securitization in the private-label market. My intuition is that there would be decent demand.”
Prologue didn’t participate in the initial transaction, but Drennen said he studied it. “I wanted to see how the first one went.”
Deutsche Bank expects to see fairly different buyer bases in the senior mezzanine relative to the junior mezzanine class. In research published before the first deal priced, analysts said that, “given the assumed spreads, we think the universe of likely players in M1 could be fairly diverse, while M2 is dominated by fast money and other levered mortgage players.”
“There’s a huge opportunity [for the GSEs to offload credit risk] because the collateral [for MBS] has greater consistency, much higher quality than it has been for at least five years,” said Clifford Rossi, Executive-in-Residence and Tyser Teaching Fellow at the Robert H. Smith School of Business at the University of Maryland. “That will make it more palatable to investors that would otherwise sit on sidelines.”
One of the biggest criticisms of the first STACR transaction is that Freddie did not seek a credit rating, which limited participation by banks.
“If it got rated, you’d definitely see banks buy it, because it’s a floating-rate asset,” said Walter N. Schmidt, senior vice president and manager, mortgage strategies, at FTN Financial Capital Markets. “A rating would have to be investment grade for the mezzanine class. Maybe at some point they could sell the first-loss piece; it may be good enough to get an investment-grade rating, too,” he said.
And insurance companies would like to have seen a rating from the National Association of Insurance Commissioners.
Another thing keeping banks on the sidelines, according to Fitch Ratings, is the need for a clarification of Basel III risk weights for the new securities. While the STACR notes were issued as unsecured obligations of Freddie, Fitch does not believe that the current 20% risk weighting applied to agency MBS under Basel III will ultimately be justified. “If the ultimate risk weights are high, banks are unlikely to become sizeable buyers,” the rating agency said in an Aug. 2 report. It said that even prolonged regulatory uncertainty could constrain participation.
However, Drennen notes that, “a deep pool of capital is not dependent on a rating to want to invest.”
Freddie hopes to have rated offerings in the futures, the spokeswoman said.
Liquidity is necessary to attract many potential investors off the sidelines, and that requires regular issuance of homogeneous securities. Barclays believes that if the GSEs sell the risk on all of their ongoing MBS issuance, there could be annual STACR issuance of $20 billion to $25 billion in mezzanine tranches, though it may be awhile before this happens.
There’s a good deal of skepticism, however.
Rossi, who has held senior risk management positions at both Fannie and Freddie, doesn’t think STACR is necessarily a structure that will be around for a long time, though it could provide a prototype for transferring mortgage credit risk in the future.
“Is this a science fair project, or can it really be productionalized?” he wonders. “If it’s a science fair project, it’s interesting, but it could be so much more… If they’re going to try a whole variety of different structures they should do so with a long-term perspective in mind; i.e., post-GSEs.”
Given the relative success of the initial transaction, which has tightened in secondary trading, many think Freddie is likely to do another deal in a similar, if not identical style, and do so on a regular basis, perhaps quarterly. But Fannie tends to march to its own drummer; the GSE is said to be working on a different type of risk-sharing transaction involving mortgage insurance, and it may redefine the approach.
“Quite honestly, I think this [STACR] form is better, it takes away any kind of counterparty event risk,” Drennen said. “Yes, insurance companies are regulated, and they usually do fine, but insurance companies have failed in the past,” he said.
“They’ve just got to keep coming and [let potential investors] know it’s a real program. The liquidity will continue to improve the more deals they do.”
Rossi thinks Fannie and Freddie’s regulator is missing an opportunity. “The FHFA, as far as I can tell, has not come forward with a clear set of guiding principles that would describe in much detail what to look at when comparing different types of arrangements,” Rossi said. “That’s where there’s still a lack of clarify about what they are trying to achieve.”
Possible Benchmark for Private-Label RMBS
Barclays says that regular issuance would also help in the price discovery of the credit risk on well-defined, liquid and large cohorts of loans. This would provide a series of benchmarks against which private-label securities could be priced.
Deutsche Bank disagrees. “Any attempt to compare this transaction to available pricing in private label mortgage credit would be akin to comparing apples to unicorns,” analysts wrote in the July report. They said the 10-year final maturity and the synthetic nature of the transaction make it unlike any new issue jumbo or legacy cash subordinate and more akin to a synthetic collateralized debt obligation.
“There is a big difference between selling $400 million of risk and selling north of $100 billion in risk,” which is the amount of protection Freddie and Fannie would have to buy if they were to insure 270 basis points of losses on the $4 billion combined insurance books, the Deutsche analysts wrote. “It is highly unlikely any indicative spread we could derive from this transaction would accurately reflect the spreads associated with that level of issuance.”
Schmidt said that, with a credit rating, STACR could serve as a benchmark for the private-label market, although he said it’s hard to see a bank or REIT keeping the triple-A tranche and the equity and selling just the mezzanine tranche. They’d almost want to do just the opposite - keep the upside and not the balance sheet burden.
“Selling just the mezzanine tranche doesn’t meet most issuer’s economics,” Schmidt said. “In this case, it does [for Freddie] because there’s a policy mandate to do it.
Pricing Feedback for G-Fees
Another benefit of STACR, or of any potentail risk-sharing transaction, is the feedback it provides on the level of fees Freddie and Fannie should charge to guarantee credit losses on MBS.
Barclays noted in its report that the M1/M2 tranches priced to a combined spread of 527 basis points for the 0.3%-3% tranche. It said this translates to about 14.25 basis points running in g-fees on the loans. The average g-fees for these loans, net of the 10 basis points going to Congress, is probably on the order of 30 to 35 basis points. “As a result, the excess g-fee that Freddie retains in this deal should be more than enough to cover the first 30 basis points of loss and set some aside for any tail risk that may remain,” the analysts said.
The STACR notes continue to receive attention as the market anticipates a risk-sharing deal from Fannie; spreads on the securities initally widened, but have since tightened in secondary market trading.
According to a poll conducted by J.P. Morgan at the end of July, investors thought the transaction’s junior mezzanine tranche was cheap at issuance, but then became more fairly valued in the 600-basis point area.
Forty eight percent of respondents said the tranche is fairly valued in the 600 to 700 basis point range; another 7% see less than 600 basis points as fair.