With the end of the Fed’s quantitative easing, who will buy and own mortgage-backed securities in the future? Market observers expect that more buyers can be lured into the market if rates rise and spreads widen.
Mortgage Banking
February 2015
By Robert Stowe England
Like a storm cloud on the horizon, a central question has loomed over the mortgage industry since the announcement by the Federal Reserve in October that it would no longer be a net buyer of agency mortgage-backed securities (MBS).
“If the Fed is going to step back, who is going to step forward?,” asks Mike Fratantoni, chief economist and senior vice president of research and industry technology, at the Mortgage Bankers Association (MBA).
Fratantoni did a systemic review of potential buyers in order to answer that question last fall, and published his findings in a November 2014 MBA white paper titled Who Will Own Mortgage Assets?
He looked at an array of existing and potential classes of investors--banks, thrifts, credit unions, bond funds, pension funds, foreign investors, real estate investment trusts (REITs) and life insurance companies--to see who might become the new dominant buyer. “No one really showed up--at least in my list,” Fratantoni says.
Starting in December 2013, the Fed gradually wound down and then ended in October 2014 its monthly purchases of $40 billion in Fannie Mae, Freddie Mac and Ginnie Mae securities.
The Fed also wound down its monthly $45 billion in purchases of longer-term Treasuries. Together, the two bond classes made up the $85 billion monthly or $1 trillion annual purchases that were part of QE 3, the third installment of quantitative easing (QE).
From 2009 through 2014, the Fed’s various QE programs added $3.5 trillion to its balance sheet. It was an unprecedented policy initiative, quadrupling the Fed’s 2008 balance-sheet assets of less than $800 billion.
The Fed has not completely exited the market as a buyer. The central bank has promised to continue to reinvest the proceeds from its bonds, both Treasuries and agency MBS. At some point, the Fed is expected to end its practice of reinvesting proceeds from its bonds.
All other things being equal, higher yields and spreads on mortgage bonds in the future would be expected to bring in investors and lenders to fund the demand for mortgages. Nevertheless, each class of potential investors faces its own constraints that could limit the market impact of higher yields and spreads.
“The Fed has entered uncharted waters,” says Mark Palim, vice president of applied economic and housing research at Fannie Mae. “They’ve never done anything like this before, where they’ve targeted one part of the intermediation process to try and target interest rates of one particular product--mortgage rates. So, we really can’t be certain how this will work out.”
Market observers are divided on how the transition will proceed. Some expect it to go reasonably smoothly while others worry about a hard landing where mortgage rates jump as credit supply created by mortgage originations is not easily met by a similar increases in purchases from investors and lenders.
“Anybody who says that they are sure it’s going to work out or they’re sure it’s going to be a disaster or they’re sure this, that and other, is probably overstating their hand,” says Palim.
“My sense is that the Fed is a dynamic player and they have a lot of discretion. They are going to be very careful and methodical,” he adds.
Above all, the Fed does not want to see a market overreaction, according to Palim. Especially not like the one that occurred in May and June 2013 when former Fed Chairman Ben Bernanke stated in an offhand remark that reductions in bond purchases--or “tapering”--could begin soon. Interest rates shot up dramatically, prompting sell-offs in emerging and high-yield bond markets.
Mortgage supply contracting
The impact of the Fed’s exit as a net buyer of mortgage bonds, however, has been muted so far. Indeed, as the purchases came to an end last year, mortgage rates, which had started back up in 2013 when tapering first began, defied expectations--and fell.
One reason the end of QE has not had a bigger impact so far is “the amount of mortgage [origination] supply has gone down as we’ve moved dramatically from a refinance environment to purchase environment,” notes Laurie Goodman, director of the Housing Finance Policy Center at the Urban Institute in Washington, D.C.
This is a major reason why rates have declined even as the Fed has exited as a net buyer, she says.
“There’s been a huge drop in issuance as well,” she says, referring to the slide of agency MBS issuance to an estimated $933 billion in 2014 from $1.57 trillion in 2013. That means that from 2013 to 2014, there was roughly a $500 billion reduction in the supply of new mortgage issuance that went along with a contrasting $500 billion a year reduction in demand from the Fed as a buyer of MBS.
Goodman thinks the lack of buyers will drive rates and spreads higher, and buyers will move in to meet demand.
“I am not at all worried” that there might be a hard landing for the mortgage market, says Goodman. “In order to absorb a very limited amount of supply, I don’t think mortgage spreads have to widen very much” this year, she says.
Spreads are “very tight right now,” not only from limited supply of mortgages but also because volatility is low, she explains.
This is not to say the mortgage market is meeting the demand from borrowers, given the “very limited credit availability,” she says. The absolute volume of loans is down--not only from 2006 and 2007 but also from the recessionary lows of 2009, Goodman notes.
Banks have put on credit overlays that have constrained the credit box because of concerns about potential requests for putbacks from Fannie Mae and Freddie Mac. This is why there is a greater share of cash sales than you would expect at this stage in the housing recovery, according to Goodman.
GSE retained portfolios decline
The Fed is not the only major buyer that has scaled back purchases. Both Fannie Mae and Freddie Mac have also been under orders from the Federal Housing Finance Agency (FHFA) to gradually reduce their portfolios of MBS holdings.
Under the Housing and Economic Recovery Act of 2008 (HERA), the two government-sponsored enterprises (GSEs), which were placed in conservatorship in 2008, are required to reduce their retained portfolio holdings from a high of around $750 billion each to $250 billion each by 2018.
Fannie Mae’s retained portfolio stood at $744.7 billion in the third quarter of 2008. Freddie Mac’s retained portfolio stood at $732 billion as of Dec. 31, 2008, according to its 2008 annual report.
Fannie Mae’s retained portfolio stood at $438.1 billion as of Sept. 30, 2014, setting the agency well ahead of the requirement to reduce holdings to $469.6 billion by the end of 2014. Freddie Mac’s retained portfolio stood at $413.6 billion as of Sept. 30, 2014--even further ahead of the requirement that it reduce its retained portfolio assets to $469.6 by the end of 2014.
During the various QE programs, the agency MBS purchases by the Fed offset the reductions in the retained portfolios for the two GSEs, Palim points out, so that there was no negative impact on the mortgage market. The question now, however, is to what extent ongoing reductions might be a negative factor that could reduce investor demand.
The level of Fed intervention to buy bonds to maintain the level of its bond portfolio is determined by two factors, according to Fratantoni. One factor is the amortization of the loans that back the securities it holds. Because the Fed is holding assets tied to 30-year mortgages and it is still early in the 30-year term for these assets, amortizations would represent a decline in principal of only a couple of percentage points of the total assets the Fed holds. The other factor is various forms of prepayment. If mortgage rates rise, then prepayments could slow.
Both factors together represent $10 billion a year of rollover investments by the Fed, according to Fratantoni.
As for the ongoing reduction in the retained portfolios at Fannie Mae and Freddie Mac, each will need to reduce its holdings by 15 percent this year. That translates into roughly a $60 billion annual reduction by each of the GSEs in 2015, according to Fratantoni.
The two GSEs together would be expected to reduce holdings by $120 billion a year or $10 billion a month. Thus, even as the Fed continues to rollover any proceeds and be a buyer of $10 billion a month, Fannie and Freddie will be cutting back by a similar amount during 2015.
Factors that could attract/deter more buyers
There are broad economic and regulatory factors that influence the willingness of investors to hold mortgage assets and “could push things in either direction, either up or down,” according to Fratantoni.
The key economic factor is the level of mortgage rates.
“If mortgage rates rise or mortgage rates become more volatile as a result of the relative dearth of investment, I think at those higher levels of rates it will draw some money managers in,” Fratantoni says.
Managers of bond funds and pension funds are always looking across different asset classes for investments that have the right combination of risk and return, and “at the right level for rates, they will increase their investment in mortgages,” he says.
There is a countervailing constraining force from regulation that could mitigate the impact of higher rates and spreads. Existing and potential regulations can also affect the likelihood that potential investors in mortgages, such as banks and real estate investment trusts, will be more likely to expand or reduce their level of mortgage holdings, according to Fratantoni.
In December, for example, the Federal Reserve announced a proposed rule for a surcharge on capital for eight of the largest banks. Fratantoni contends that the extra new regulatory burden on these banks “would make them even less likely to increase their mortgage holdings.”
Under the proposed rule, the amount of additional capital would be individually determined for each bank based on how risky the regulators judge the bank to be, based on the Fed’s own criteria--such as the degree to which banks rely on wholesale funding.
In his review, Fratantoni found that a whole set of higher capital rules under Basel III limit the degree to which banks might want to expand their portfolio holdings of mortgages.
Under the new rules, banks have to back whole loans with 4 percent capital. MBS backed by Fannie and Freddie, by comparison, only require 1.6 percent capital.
Further, the new liquidity coverage ratio rules will make Ginnie Mae securities more attractive than Fannie Mae and Freddie Mac securities because Ginnie Mae’s have an explicit government backing and, as a result, the capital requirement would be zero.
The role of banks
Higher guarantee fees charged by Fannie and Freddie would ordinarily be an incentive for banks to hold more whole loans, according to Fratantoni. But the capital and liquidity regulatory requirements create an incentive to move away from whole loans to Ginnie Mae securities, he explains.
High mortgage rates could possibly be an incentive for banks to hold more whole loans, but only if the spreads widen between the mortgage rates and the interest rate banks pay depositors, according to Fratantoni. If depositors demand higher returns, it could limit the appeal of holding whole loans.
Some observers, however, see less impact from all the challenges facing potential players, including banks. “Maybe it’s my Econ 101, but if you want someone to buy something, you have to change the price of it,” says Mark Calabria, director of financial regulation studies at the Cato Institute, Washington, D.C. “So, I have no doubt on a very basic level that the market would be willing to supply mortgage credit at the right price.”
While Calabria acknowledges that there are regulatory obstacles that make it difficult for banks and others to be more fully involved in the mortgage market, he also points out that there are offsetting capital rules that favor mortgages, too.
“Even with Basel changes shifting some of the preference from whole loans to MBS--which to me is actually a positive--mortgages are still favored over small-business loans under capital standards” for banks, Calabria says.
“There is this myth that Fannie and Freddie connected Main Street America in terms of mortgages with Main Street America in terms of investors,” says Calabria. “That was never the case. The majority of GSE debt was funded by the rest of the [U.S.] financial system, by and large, with a small sliver held internationally.” He contends that the financial system will likely continue to hold those mortgages and mortgage-backed securities.
While Fannie and Freddie at one time held as much as $1.5 trillion of mortgage securities, Calabria notes, banks currently have more than $2.5 trillion in excess reserves sitting at the Fed.
“Why aren’t we concerned that banks find it more attractive to get 25 basis points from the Fed than to get 450 basis points for making a mortgage?,” asks Calabria. “Something is wrong here, but it’s not the lack of funds.”
Calabria does not think rates will jump significantly as a result of the Fed ending QE. Higher rates, in turn, could expand the available mortgage credit.
“We forget that the quantity of mortgage credits provided is not just a function of demand but it is also a function of supply,” Calabria says. “So higher rates will make it more likely that banks will want to enter that space.” Higher rates and spreads could be an incentive for banks to move some of their excess reserves into lending, including mortgage lending, he contends.
Calabria agrees with those who point out that financial institutions are constrained by legal and regulatory concerns, and he supports efforts to get rid of the most onerous regulatory constraints. “But this isn’t an issue of a lack of capacity. There is more than enough capacity in the system,” he says.
Based on federal Flow of Funds data, “commercial banks could fund the entire mortgage industry for the next couple of years solely with the excess capital they have lying around on their balance sheets,” he contends.
Fannie Mae’s Palim, however, thinks that the share of mortgages held by banks is not likely to swing wildly, either up or down, based on his reading of historic trends.
“Banks are a fairly steady player. They aren’t really the marginal swing player,” he says.
Looking at the Federal Deposit Insurance Corporation (FDIC) data on bank holding companies across the years, Palim finds that banks have shown an ability to expand significantly their role in the origination of second mortgages. “But when it comes to first whole loans, they tend to pretty much have a core holding and don’t fluctuate terribly much in response to all sorts of changes in the environment,” he says.
Chris Whalen, senior managing director in the Financial Institutions Ratings Group at the Kroll Bond Rating Agency Inc. in New York, sees the regulatory constraints on banks, however, as a binding constraint on their ability to expand lending significantly. The fear of putbacks and lawsuits is paramount and is not going away, he contends.
Whalen estimates that a normal credit market for mortgages, absent the excessive constraints imposed by regulation and avoiding the excesses of bubble lending, would be $1.5 trillion instead of the current $1 trillion.
“My take is that I don’t think we can put Humpty Dumpty back together again,” says Whalen, referring to the parts of the market not being served due to credit constraints. “I think the whole private-label market other than prime issuance is just not going to come back.”
Bond funds and other players
Fratantoni explains the currents that work to limit other buyers in the mortgage market in his review. For example, the makeup and performance of bond benchmarks indexes could inhibit the expansion of mortgage holdings by bond funds, pension funds and other money managers. Fratantoni explains that money managers are judged by how well they perform relative to a benchmark and also by how much they deviate from the asset allocations contained in the benchmark.
“If they are very much overweight or underweight [in given classes of bonds] relative to that index, that puts them at risk if they perform worse than what the index weighting would provide,” he says.
“On average, across all the money managers, they are going to be very close to that benchmark” in their asset allocations, he adds. As a result, fund managers would not likely significantly increase mortgage holdings unless they become a larger share of the bond benchmarks against which they are measured.
“Given that [U.S. federal] budget deficits are expected to rise, Treasuries will become an ever larger part of the fixed-income universe while mortgages become a smaller part,” Fratantoni says. That will lead some of the money managers to lean more toward increasing their holdings of Treasuries and away form mortgages, he adds.
Foreign investors have been a key buyer of GSE securities, and the uncertain future of Fannie Mae and Freddie Mac may act to limit their interest in investing in their securities, according to Fratantoni.
The uncertainty revolves around whether the U.S. federal government will in the future continue to back the securities of the two GSEs when their status changes as a result of housing finance legislation. “As a foreign central bank trader described to me, if he has to use more than a couple of sentences to explain why he put that country’s money in a Fannie MBS instead of a Ginnie MBS, that’s just not going to be a good situation,” Fratantoni says. “He’s just going to say, ‘Well, I bought the one that has a federal guarantee.’ That’s a much easier discussion for a trader to have.”
If, under new legislation, the equivalent of Fannie and Freddie securities would have an explicit government guarantee behind them, “that would be very beneficial in removing that ambiguity” that limits foreign investment in U.S. mortgage assets, according to Fratantoni.
Fratantoni cites as a possible model the explicit government guarantee for mortgage securities in a new mortgage finance system under the Johnson-Crapo bill (The Housing Finance Reform Act of 2013), introduced during the last Congress by Senators Tim Johnson (D–South Dakota) and Mike Crapo (R–Idaho).
“If the nature and structure of the mortgage market changes, we are going to have to think about how to create new channels or in some ways change those old channels to get that flow of global capital into the U.S. mortgage market, because our domestic savings are insufficient to provide for our borrowing needs,” Fratantoni says.
The role of REITs
Fratantoni see somewhat more promise for expanding the role of REITs as owners of mortgage assets than for other classes of investors. He also sees a systemic benefit to greater participation in the market by REITs. He points out that while REITs are leveraged, it is typically 6 to 1 instead of the 10 to 1 for banks. Thus, if REITs expand their role relative to that of banks, it would reduce overall leverage in the system, explains Fratantoni.
For REITs to play a larger role, investors and entrepreneurs would have to create more REITs or existing REITs would have to expand through additional equity capital.
REITs also face regulatory constraints that could limit expansion of their footprint as buyers of mortgage assets. For example, they rely on repurchase agreement or repo funding. Federal banking regulators have issued capital and liquidity rules that nudge banks into reducing the overall level of repo funding, which regulators view as an unstable source of funding.
To find more stable sources of funding, some REITS have moved to access funding from the Federal Home Loan Bank System through captive insurance companies that become members of the system. For example, Redwood Trust Inc., Mill Valley, California; Annaly Capital Management Inc., New York; Invesco Mortgage Capital Inc., Atlanta; and Two Harbors Investment Corporation, New York, have captive insurers that have joined the Federal Home Loan Bank System.
Federal Home Loan Bank System funding for REITs, however, could be eliminated under the Federal Housing Finance Agency, which regulates the Federal Home Loan Banks. FHFA proposed last year to deny membership in the system to captive insurers owned by REITs.
Fratantoni disagrees with the FHFA proposal and points out that Michael Stegman, counselor to the secretary of the Treasury for housing finance policy, has stated that mortgage REITs are a long-term holder of mortgage assets, which is line with the mission of the Federal Home Loan Bank System.
Reaching a new equilibrium
The mortgage market is in transition and the valuations set on mortgage assets “will depend upon the equilibrium yields or spreads that emerge after the Fed has definitively stopped buying,” says Michael Youngblood, principal and co-founder of Five Bridges Advisors LLC, Bethesda, Maryland. That would occur after the Fed no longer makes bond purchases to keep its holdings of mortgage assets at current levels.
Youngblood argues that during the long period of quantitative easing, markets have seen artificially high values for U.S. mortgage loans and securities. With the end of QE, markets are just beginning to adjust.
“I don’t believe we have yet found the long-run relationship of U.S. mortgage loans and securities to other U.S. fixed-income securities,” he adds. “I think as mortgages cheapen generally that we will see a revival of interest from some new sources and from some traditional sources.”
An expanded role for REITs?
Youngblood, too, sees a potentially expanded role for REITs, in part because of the “pioneering role” played by Redwood Trust. “We will soon have a lot of companies acquiring mortgage loans that others have originated,” says Youngblood. Following the path blazed by Redwood, these companies will securitize the mortgages they have purchased from the originators, earning a built-in funding spread; in this way, they will obtain long-term funding for the mortgages, he explains.
It’s not such a stretch to imagine REITs playing a larger role because they have done so during several periods in the past, Youngblood notes. In the late 1980s and early 1990s, for example, REITs were important in the period of the thrift crisis. REITs, along with life insurance companies, were “major holders” of mortgage loans, but less of mortgage securities, he points out.
“A whole era of mortgage REITs sprang up to own these securities, names like MDC Asset Investors and Residential Resources [ResRe] out of Phoenix,” he says.
There was another period of expanded REIT activity in the aftermath of the failure and rescue of Long-Term Capital Management in 1998, when many subprime and other mortgage specialty companies gained new life by converting to REIT status and raising new capital, according to Youngblood. “In fact, a large part of the erosion in mortgage lending standards happened once these companies had recapitalized and were needing to put their money to work, and it was a very competitive environment for what was initially a relatively small subset of the population of mortgage borrowers,” he says.
Youngblood says he “would not be surprised” to see another flowering of mortgage REITs in the remaining years of this decade. This may occur because more REITS will “seek long-term alliances with established mortgage lenders and servicers who simply do not wish for diversification and capital reasons to own all the mortgage loans that they are capable of originating,” he says. Youngblood sees mortgage REITS “fulfilling a true role as intermediaries of mortgage credit.”
Calabria, however, does not expect REITs to expand their role very much, although he favors an expanded role as long as it is built on new equity and not higher leverage.
It will be banks, not REITs, that step up in the coming years to “play the biggest role” in expanding the ranks of potential buyers of mortgage assets, predicts Calabria.
Not only do banks have the unused excess reserves to draw on, but they also have the infrastructure in place to “ramp up” lending. “If they get the risk-reward trade-off right, banks could essentially take on another trillion dollars in mortgages without a problem,” Calabria suggests. “It’s really more a risk-return issue than a capacity issue.”
For banks to play a larger role, however, they will have to charge higher interest rates and charge fees to offset the risk. Thanks to the distortions of QE, mortgage rates at 4.25 percent to 4.5 percent “barely cover the interest-rate risk, and there’s nothing to cover the credit risk” of portfolio lending, says Calabria.
Higher rates and wider yields will make the lending more profitable and attractive enough to deploy funds banks already have. Calabria also thinks it is better for the financial system if mortgages are held by banks than, say, Fannie and Freddie, because banks are much less leveraged than the GSEs.
One reason Calabria does not see a need for REITs to ramp up is that he thinks it is more likely that life insurance companies can significantly expand their role; and because it is a larger industry, an expansion can have more of an impact on the size of the universe of purchasers than even a doubling of mortgage REITs from their $200 billion in current holdings.
Calabria estimates that life insurers could absorb an additional $1 trillion of the $9 trillion mortgage market. “There’s a lot of capacity there that makes a lot of sense,” Calabria says. “I don’t see it as a panacea, but it’s not insignificant, either.”
Youngblood, too, thinks life insurance companies will likely respond to higher mortgage yields and wider spreads by expanding their holdings of mortgage assets. This is partly the case because the long duration of mortgage assets more closely matches the long duration of liabilities at life insurance companies.
Life insurers are likely to acquire mortgage assets “across the agency and the jumbo spaces,” says Youngblood, but they will probably not be as interested in specialty mortgage products as they were prior to 2008.
Ginnie Mae funds
Ginnie Mae funds could expand their role in the mortgage market, according to Youngblood. As yields move higher and commercial banks become increasingly less interested in holding large volumes of deposits, “the old Ginnie Mae funds will likely open a new chapter and become more popular vehicles than they have been in the past few years due to market turmoil,” says Youngblood.
“There are clearly many savers who are keen to earn a fixed income that’s more than one can earn on insured deposits, and I think that the Ginnie Mae funds will likely see renewed inflows and provide some of the credit for the new Ginnie Mae and GSE securities,” says Youngblood.
Credit unions
Credit unions, which have already expanded their role in mortgage originations, are expected to play an even larger role in the future, according to Youngblood. Credit unions will likely originate Federal Housing Administration (FHA)-insured and GSE loans that they hold on their books and will turn to mortgage-backed securities for their liquidity needs, he adds.
Youngblood notes the already vigorous growth in mortgage holdings by credit unions from $317 billion at the end of 2009 to $378 billion in 2014.
“We know that [the National Credit Union Administration (NCUA, Alexandria, Virginia)] is moving to regulate the five largest of the credit unions, led obviously by Navy Federal [Credit Union, Vienna, Virginia]. In this case, I think the regulators will have no issue with the expanded mortgage share in the Big Five and all the remaining credit unions,” says Youngblood.
A discount mortgage world
“Right now we live in a world where most mortgage-backed securities are trading at a premium, and their values are constantly at risk with each seasonal or cyclical wave of prepayments,” Youngblood says. “As the Fed exits this market, yields will have to rise to inspire REITs, credit unions, life insurance companies and, to a lesser degree, the thrift industry, to hold greater market share in mortgage loans and securities.”
In moving from a premium to a discount mortgage world with higher yields, mortgages will move from being negatively convexed to a world where they are positively convexed, according to Youngblood. “And therefore any seasonal pickup in prepayments through the school-year cycle--June, July and the like--will work to enhance mortgage returns rather than degrade them,” Youngblood explains.
“And you will remember that when Lewis Ranieri and Larry Fink [the Salomon Brothers and First Boston co-pioneers of the mortgage securities market] were making their reputations as captains of finance, it was in the 1980s when we had a discount mortgage world,” says Youngblood.
“It was not until the era of the Resolution Trust Corporation in the early 1990s that the industry moved to a world of premium-valued mortgage loans,” he recalls. “I remember when Ginnie Mae 8’s traded in the mid-50s. I don’t think we’ll have Ginnie Mae coupons at 8 percent anytime soon. But we could certainly see 3’s, 3½’s and 4’s trading in the 90s in a world of higher mortgage yields,” he says.
Palim at Fannie Mae also expects spreads to widen with the Fed’s departure as a net buyer. “Spreads relative to Treasuries are unbelievably tight compared to the pre-crisis period. So it’s reasonable to think, absent the Fed intervention, that spreads are likely to widen out,” he says.
While it’s difficult to pinpoint where bottlenecks might lie ahead as the mortgage market adjusts to a new discount value world, it is fairly clear that another big transition point lies ahead when--and no one is sure when that will be exactly--the Fed chooses not to invest its current proceeds.
“Then we’ll see how the market reacts to that,” Palim says. He expects the Fed to cautiously and carefully manage any difficulties that might arise. MB
Robert Stowe England is a freelance writer based in Milton, Delaware, and author of Black Box Casino: How Wall Street’s Risky Shadow Banking Crashed Global Finance, published by Praeger and available at Amazon.com. He can be reached at
