April 13,, 2020

917/379-0641
[email protected]
Twitter: @joshrosner

Mortgage Servicing, Forbearance & GSEs: Liquidity Events Versus Credit Events

Chicken Little is Wrong, It’s Not 2008

As many of our readers may remember, we were among the first to issue dire warnings of the
seriousness and severity of the coming 2008 housing and mortgage market crisis.1 At that time
most experts2 suggested that the housing market correction was likely to be mild and would not
meaningfully impact the broader economy.

Today, many experts are suggesting that we are on the cusp of another severe housing crisis that
is driven by the Covid-19 lockdown. We do not disagree that there are some significant risks to
mortgage markets but we do not believe this is the beginning of a meaningful housing market
correction. Unlike the 2008 crisis, in which the Federal Reserve was contending with broad
structural solvency issues affecting borrowers and the mortgage industry, today’s problems
are liquidity rather than solvency related.

Today, most financial firms have significantly more capital on hand to address liquidity issues
and some realistic increase in credit losses. While banks will be less able to drawdown loan loss
reserves than they have been in the past decade, and will need to begin to increase provisions,
most will be able to manage these events. Furthermore, even Fannie Mae and Freddie Mac, who
have had strong earnings power but remain massively undercapitalized because of the past
ongoing Government sweep of their net worth into the Treasury, are still in a better place than

1
See e.g.: New York Times, A Coming Nightmare of Homeownership, Gretchen Morgenson,
Oct 3, 2004; Medley Global Advisors, Not Your Father’s Housing Cycle, November 2006;
Rosner, Joshua & Mason, Joseph, “How Resilient Are Mortgage Backed Securities to
Collateralized Debt Obligation Market Disruptions?”, February 2007; Rosner, Joshua &
Mason, Joseph, "Where Did the Risk Go? How Misapplied Bond Ratings Cause Mortgage
Backed Securities and Collateralized Debt Obligation Market Disruptions", May 2007: New
York Times, OpEd Stopping the Subprime Crisis, Joshua Rosner, July 25, 2007.
2
See, e.g.: McClatchy-Tribune, “Housing, energy key to economic outlook”, Kevin Hall,
December 31, 2006 (““The principal reason for optimism is that businesses … are in fantastic
financial shape. It’s very difficult to envisage businesses pulling back, at least to the degree that
it would cause recession,” said Mark Zandi, chief economist for Moody’s Economy.com, a
consultancy in West Chester, Pa. “They may grow more cautious in the face of softer demand …
but they’re not going to cut payroll and slash investment.”… He points to other positive data,
such as increased trade flows and mortgage applications. “They all suggest that maybe the
economy is going to revive faster than anticipated,” he said, which could push the Fed to raise
rates – and risks.””)

Please refer to important disclosures at the end of this report.

they were in 2007. Through the transfer of credit risk, to buyers of Credit Risk Transfer (CRT)
notes and better capitalized Private Mortgage Insurers, the GSEs have successfully transferred
far larger parts of their first-loss positions to other players. Furthermore, if home values do not
meaningfully decline, on a national basis, borrower’s equity will remain a broadly unimpaired
asset by which to recover from Covid-19 driven liquidity problems.. While there will be some
job losses the restrictions placed on many of the Government’s stimulative programs means that
once the economy ‘reopens’ we will realize that many of the jobs that have been lost will be
regained as workers realize that they were furloughed rather than fired. As a result of all of these
factors, if handled appropriately, we expect the long-term losses in the housing and mortgage
markets, will be manageable and will not become a broad credit crisis affecting lenders, servicers
or borrowers.

While there is no question that the spring 2020 selling season will be muted and essentially non-
existent, the dramatic simulative efforts of Congress, the Federal Reserve and the Trump
Administration have acted to support the household sector. While these efforts are not a panacea
they will ameliorate much of the downside risk. Low interest rates will support home
affordability and mortgage refinancing and, in turn, GSE earnings.

Servicers Need Liquidity

With all of this said, there is more that must be done to ensure that the remaining liquidity issues
do not become solvency issues that would lead to broad credit losses.
As we have previously highlighted3, the broad mortgage-forbearance program that Congress
included in the CaresAct do create unintended and inequitable demands upon mortgage lenders
and servicers. While the Act allows borrowers of government-backed-mortgages who have
suffered economic harm from the Covid-19 crisis to request up to 360-days of deferral of
payments of mortgage principal and interest it does not provide servicers with the unprecedented
liquidity necessary to advance payments of those arrears to investors in government-backed
mortgage backed securities. As a result, many servicers are obligated to, out of their own
pockets, advance those payments to investors who are guaranteed regular payments of principal
and interest. Because these payments are a deferral and not relief from those obligations this is
currently a liquidity and not a credit-loss event.

To prevent this from becoming a broad credit problem the Federal Reserve must act swiftly to
provide non-bank mortgage servicers with the liquidity necessary so that they can afford to
provide borrowers the relief Congress has offered while still being able to maintain their
contractual obligations to investors in government-backed-mortgage-securities. Bank servicers
already have similar liquidity financing through the Federal Reserve’s Discount Window and the
Fed has authority, under 13.3 of the Federal Reserve Act, to provide a similar liquidity facility to

3
See: GF&Co, A Corona Crisis and a Mortgage Crisis, March 29, 2020; and GF&Co: The
CaresAct- Inequitable Treatment and Misconceptions, April 6, 2020.

-2-
non-bank servicers. This would not only be the fair and equitable response to such an
enormously large Government mandated forbearance program but it would be the least costly
way to prevent a crisis-induced liquidity event from becoming a larger mortgage and housing
credit event. But, the Fed must act quickly if it seeks to minimize its own losses and broader
economic losses. Doing so would support its mandate to support price stability.

The Risks of Losses to Fannie Mae and Freddie Mac

Some so-called mortgage and housing experts are warning that Fannie Mae and Freddie Mac will
suffer extreme liquidity events and credit losses as a result of the Government’s forbearance
programs. These experts further warn that, as a result, the GSEs will need to draw-down their
Treasury extended credit lines. These so-called experts have offered an overly-simple analysis
that ignores several important facts that must be discussed. First, as of last quarter, 63% of the
GSEs mortgage servicing was performed by banks. Those bank servicers, unlike non-bank
servicers have access to the Federal Reserve to access servicer advance monies.

Further, the majority of GSE agreements are on a “Scheduled, Scheduled” basis. In

“Scheduled, Scheduled” agreements the servicer is required to advance payments of principal
and interest (P&I) to MBS investors whether or not those payments have been received. As a
result, during this period of borrower forbearance there are no liquidity or credit demands on the
GSEs and, according to their servicing agreements, if a servicer fails to perform then the GSEs
are not required to advance payments to MBS investors but are, instead, permitted to replace the
non-performing servicer with one who is able to service the loans according to contractual
agreement.

Second, even after the forbearance ‘covered period’ ends, the offered analysis of so-called
experts is incorrect. Those experts suggest Fannie and Freddie will likely have to repurchase up
to $250 billion of mortgages from their mortgage backed pools and that will precipitate the need
to draw-down Treasury lines of credit. Even if this analysis proved correct, this would initially
be a liquidity event rather than a credit event. But this analysis is wrong and it misses a further
and more important reality – the waterfall of approaches that will be utilized, by servicers, to
reclaim the amounts of principal and interest they advanced to MBS investors, will not lead to
broad demands on the GSEs to repurchase mortgages from MBS pools and will not likely lead to
massive credit losses.

Post-Forbearance Approach to Collecting P&I Arrears

While we do not disagree that there may be as much as $250 billion of outstanding P&I due at
the end of forbearance, there are at least three (3) approaches to borrower repayments of those
past-due P&I payments that we expect servicers to employ before there is any risk of the GSEs
having to purchase loans out of their MBS securities. Only if none of those approaches are able
to address repayments of a borrowers outstanding forbearance balance would there be any need

-3-
for the GSEs to come up with the liquidity to repurchase loans from MBS pools. Currently, it is
our belief that only a small number of borrowers would not be able to begin to reperform and
would therefore require GSE the standard loss-mitigation/modification approaches that result in
repurchases from pools.

So, what are those three, pre-repurchase approaches?

The first approach is quite simple. After the end of the forbearance period the servicer would
reach out to the mortgage borrower and determine, with the borrower, if they are able to make a
lump-sum repayment of the outstanding P&I. This approach would primarily apply to borrowers
who received forbearance but did not suffer a significant economic impact from the Covid-19
event. These borrowers would have often been borrowers who falsely attested to an economic
harm and should not have qualified for forbearance in the first place. As a result of not having
actually suffered an economic harm this pool or borrowers are not offered protection, under the
CaresAct, from penalties or higher interest rates on repayments. We expect that this approach
could apply to between 10-20% of borrowers who received forbearance.

The Second Approach will likely apply to a large number of borrowers who were eligible for
forbearance under the CaresAct. Under this approach, after the end of the forbearance period, the
servicer would reach out to the borrower and provide them with the total amount of P&I that is in
arrears. Because this group of borrowers would be comprised of those borrowers who had
suffered an economic harm from the Covid-19 crisis the servicer would be prohibited, under the
CaresAct, of including any costs above the contractual amount of outstanding P&I. Here the
servicer would negotiate a repayment plan that would allow the servicer to recapture the amounts
it had advanced to MBS investors. This repayment plan would be separate from the borrower’s
obligation to begin to reperform on making their regular mortgage payments. So, as example, the
borrower may have a period of perhaps a year during which they pay their mortgage and a
separate amortized payment of the outstanding P&I that the servicer had advanced on their
behalf. We expect that this approach could resolve the outstanding P&I of between 25-35% of
borrowers who received forbearance.

The third approach will likely apply to the largest number of borrowers who were eligible for
forbearance under the CaresAct. This approach is detailed in the “payment deferral programs”
announced last month by Fannie Mae and Freddie Mac. Under this approach a borrower who
receives forbearance would simply have the outstanding P&I added as a balloon-payment at end
of their existing mortgage. That is to say, when a borrower sells their home, or refinances it, the
outstanding P&I resulting from the forbearance period would then be paid from the proceeds of
either of those events. As a result, in this approach, the repayment essentially becomes tied to the
house and not the borrower. We expect that this approach could resolve between 40-70% of
borrowers who received forbearance.

-4-
Each of these three approaches would occur without Fannie Mae or Freddie Mac having to come
up with the liquidity to repurchase mortgages from their outstanding MBS pools. We estimate
that, even if there is $250-billion of forbearance, this approach would likely address most of it.
Any forbearances that can’t be resolved with these three approaches would not mean that the
GSEs would take meaningful losses. After all, repurchases would be liquidity events and not
necessarily credit events. Unlike the pre-2008 period in which home prices were relatively high
in comparison to rental prices, nationally, today rental prices are high relative to home prices. As
a result, we expect that with standard GSE loss mitigation efforts and modifications of loans, the
vast majority of those borrowers will reperform on their mortgages rather than default.

In Summary, we remain concerned that a failure of the Federal Reserve to provide a liquidity
facility to those non-bank servicers of Government-backed-mortgages could precipitate a credit
crisis that flows through to lenders and, in turn the availability of mortgage credit to Main Street.
If, however, the Fed stands up such a facility we believe that the impacts of the mortgage
forbearance provisions of the CaresAct will likely result in a short-term liquidity impact to the
mortgage sector with manageable credit losses. Still, broadly speaking, we do expect that
reserves and provisions will need to rise throughout the financial sector. After all, there is no
replacement for capital to cushion institutions from losses.

Lastly, this episode in our collective history will serve to demonstrate another key fallacy
propagated over the past decade. Specifically, the belief that more GSEs is better than two GSEs
has now been effectively undermined. After all, it is far easier for regulators to create and require
uniform standards for complex issues among two companies than it would be among a larger
number of companies. Some of us understood, long ago, that this ability to create standards is the
most significant reason the agency market continued to function between 2007 and 2010 while
the Private Label Securitization (PLS) market, with its 300+ different Rep & Warranty &
Pooling and Servicing Agreements, failed and have never really recovered.

-5-