A paper written
for the Joint Center on Housing Studies at Harvard University focuses on the
causes and effects of mortgage credit tightening in the current COVID-crisis.
The author, Don Layton, is a former CEO of Freddie Mac and a Senior Industry
Fellow at the Center.
One of the biggest pandemic-related
issues to emerge in housing finance is the availability of credit. Tightening is
being discussed as a major problem, perhaps on a par with the last financial
crisis. The implication seems to be that much if not all of that tightening is
illegitimate, a failure of government policy that could be avoided with the
right actions that do not require subsidies for small business or specific
industries. Layton questions this view and examines how mortgages are made
today and who sets the credit standards.
The complex structure of mortgage
financing that has emerged over the last 60 years has three key impacts that
are related to and must be taken into account when trying to reduce unnecessary
credit tightening. First, strong stress liquidity is due solely to government
support. Second, government agencies are now arbiters of acceptable credit, and
third, intermediaries have become a major market force. Further, he says only
one of these was clearly intended by the original policymakers.
Government’s heavy intervention in
mortgage financing was originally intended to ensure liquidity in stressed
markets. It has employed two mechanisms. The Federal Home Loan Banks (FHLBs)
used an implied government guarantee to fund mortgages issued by banks. They went
into the pandemic with about $1 trillion in outstanding borrowings (used to
finance only a fraction of bank-owned mortgages in the non-stressed pre-crisis markets)
and can expand dramatically if needed.
The GSEs (Fannie
Mae and Freddie Mac) and FHA/VA both have government support and their mortgage-backed
securities (MBAs) trade as near-Treasury quality debt with close to the same
ability to access markets. This presumably ensures liquidity will be available
to primary-market lenders, even in stressed markets.
The vast majority
of mortgage loans sold to the GSEs or insured by FHA/VA are done on a “flow”
basis, with loans originated specifically for sale or insurance. This has led
to the rise of non-bank lenders who “originate to sell” and lack the capacity to
hold loans for more than a short period. Thus government-supported agencies have
become, not just providers of stress liquidity
but also arbiters of credit terms for the majority of American mortgages.
A regrettable side-effect of this
function is to put the agencies in the midst of the perpetual political debate
about mortgage credit risk. Layton says left-leaning groups seemingly always
believe the credit box is too tight and expensive while those to the right
always believe it is the opposite, while both claim to be data-driven.
The emergency of the secondary market in the late
1980s and early 1990s brought mortgage intermediaries (loan originators) into
the forefront of housing finance. With government agencies both the source of
liquidity and the arbiter of credit terms, the intermediaries have invested
heavily in creating associations to influence how the agencies operate. These
include the Mortgage Bankers Association, Community Home Lenders Association,
and the Housing Policy Council. These are complemented by less specialized
associations like the American Bankers Association, and adjacent industry
associations, such as the National Associations of Home Builders and of
They, of course, have their own views
and economic interests and are always advocating for government policies that
maximize their bottom lines: lower interest rates on mortgages, looser credit
terms. When the agencies change a policy to benefit borrowers, there is always
a question of how much will reach the target rather than being absorbed by the
Some of these associations are now loudly
claiming that government-induced difficulties have caused credit tightness. Therefore,
as a matter of fairness, the government should reverse or offset those
difficulties, especially ones causing their members’ losses.
Layton says there are three major factors that determine
the tightness of mortgage credit risk at any point in time: (1) the state of
the economy, especially housing markets; (2) the factors that drive different types of investors in mortgage
credit risk act through the economic cycle; and (3) how being an intermediary
between borrowers and mortgage investors can translate into “overlays” that
All else being equal, mortgage
credit risk moves with the economic cycle. A loan has lower risk when home
prices are rising or unemployment is falling than when the opposite occurs. Thus,
to some degree, the current tightening is not a failure of government policy
but a reflection of how risk responds to the economy.
tightening may seem worse than it is due to how quickly things changed. At the beginning
of the year credit was considered somewhat loose, based on record low
employment, strong wage growth, and rising home prices. Then the economy turned
on a dime with unemployment hitting 14.7 percent by April. Home prices however
continue to rise at about 5 percent. Based on both factors, mortgage terms
probably should be a bit tight, but not overly so, for mainstream borrowers. However,
given the precipitous prices declines in 2008, one can understand an investor bias
towards greater tightening, especial for marginal borrowers.
In 2008, the
financial system was itself in massive distress, and its capacity to extend
credit for mortgages dramatically declined. However, in 2020, banks are
healthy, the agency MBS markets (with the help of the Federal Reserve
re-establishing trading liquidity) are absorbing record new issue volumes, and
even private mortgage insurance firms (PMI) are continuing to operate
relatively routinely. The sole exception is the private label securities (PLS)
market, which has mostly evaporated.
Layton says that so far, because the
economic downturn is mainly hitting employment rather than home prices, and with
the relative health of the banking and financial systems, mortgage credit
tightening has taken the form of a targeted withdrawal from riskier borrowers
and products. This is ultimately less damaging to the economy than the mass
credit tightening in 2008 and the two eras should not be confused.
Each category of mortgage investors has structural
requirements that make them more or less or not at all procyclical. Private-sector
investors are clearly procyclical, government-supported ones are not.
Banks, which make up about 25 percent
of the market, must be good stewards
of their stockholders’ capital through quality risk management risk adverse
credit decisions. This seems to be happening. For example, JP Morgan and
Wells Fargo reacted to the pandemic by suspending HELOC originations. Thousands
of banks will make their own decisions taking more and less severe steps and
driving the market. Local banking supervisors will also require them to tighten
up on all types of credit.
PLS has a tiny share (5 percent) of the mortgage
finance system, but their loans are very visible, concentrated in either high-risk
non-QM mortgages or “jumbo” loans. Because there is no obvious balance sheet to
act as a shock absorber when the investors in PLS securitizations flee, and
they have, this is the most procyclical portion of the mortgage market. Its
extreme credit tightening can leave the impression it represents more than its
small and high-risk portion of the market.
and the VA, with a 20 percent market share, have no shareholders or profit
expectations and have been used in the past to achieve government policy
objections. In the Great Recession their market share more than doubled to 20
percent to help fill the hole in origination capacity. Layton said so far there
is no indication of an “instant
replay” of that countercyclical role and neither agency has loosened or
tightened its credit box.
The charters of the two GSEs were designed ensure they
would provide liquidity to primary lenders, but it does not require them to broadly
underprice risk or assume inordinately risky mortgages. They are really hybrids
of for-profit companies and government agencies sometimes used for policy purposes
but can’t always be both at once.
Like the FHA/VA, they have neither tightened their
credit box (like a private-sector, shareholder-owned company would) nor
loosened it (as a government agency might). As they have in the past, they are
using their giant monoline balance sheets and quasi-utility status to be resistant
to pressures to tighten and loosen mortgage credit as the economic cycle goes
up and down.
However, the GSEs cannot
act alone in determining credit risk. The PMI and Credit Risk Transfer (CRT) markets
also impact them.
There can essentially be no
high-LTV lending unless PMI firms accept such loans which is required on about
20 percent of GSE mortgages. There are already reports that they, as shareholder-owned
firms, have tightened their credit criteria and raised prices so the GSEs have
effectively had their criteria for high LTV lending tightened up for them. This
makes them appear procyclical to a modest degree for borrowers with higher risk.
Based upon late May/early June
secondary trading, CRT markets are pricing the current flow of GSE mortgages about
20 to 30 basis points higher than before the pandemic. However, with giant
balance sheets to handle risk, the GSEs are
not pricing new loans based upon marginal CRT costs but are absorbing their
ups and downs, reflecting the resulting average cost versus the average G-fee
generated by their entire mortgage portfolio. This allows the GSEs to avoid
being significantly procyclical, in contrast to banks and PMI firms.
The mortgage intermediaries or loan
originators currently sit astride the 70 percent of mortgages that go from
primary market lenders to MBS markets. At each step of the mortgage origination
process, as a loans passes through the hands of a particular intermediary, that
entity has potential risks and costs. In stable economies, they simply pass the
credit terms of the GSEs and FHA/VA through to potential borrowers. However, in
unsettled times, they add their own tighter terms or “overlays. Layton deals
with the current use of overlays, especially those associated with the Congressionally
mandated forbearance plans in a special report. We will summarize
this report separately.
Layton concludes that mortgage credit tightening is real,
but more targeted and less impactful than indicated by some of the industry
lobbying and more sensationalist reporting on the issue. It is also quite
different from what happened in 2008 to 2011 as there has been no collapse of
This time the challenge appears related to the ability
to refinance at modestly lower rates, and possibly cashing out built-up equity.
Still, with the industry currently processing record volumes of new mortgages,
especially refinancing while seemingly enjoying record origination profits, one
can assume that credit tightening has
not stop the clear majority of borrowers from refinancing, only the most
marginal ones have been impacted. Again, this issue is clearly a targeted one,
with little resemblance to the 2008 to 2011 era.
The private sector has tightened up on credit in a
rational and expected manner as well. It has reacted to the economic reality
that risk is higher than before, leading to banks making loans for their own
portfolios, including assets like HELOCs that have no connection to the CARES
Act forbearance requirement or what the government agencies will or will not
purchase on what terms. It has led to the PLS market, as expected, mostly
evaporating, impacting riskier loan products and a very small percentage of the
market, and to the PMIs raising prices and toughening credit terms, modestly,
on the high-LTV loans they support for the GSEs.
The net result is more marginal credits are becoming
unacceptable, while the vast majority of those seeking to refinance, for
example, still seem to get approved. This is not a situation in which the
government should attempt to somehow undo this natural and reasonable
Third, as far as can be determined, the four
government guarantors have not
tightened up their credit boxes, meaning they are already doing a lot to
support the economy. Neither has the government deployed FHA and VA to be
countercyclical as this would not address the main problem, marginal credits
unable to refinance. That the GSEs, as profit-making companies, have not
tightened credit as the banks have is a significant policy benefit to the
Fourth, the core issue of credit overlays by
intermediaries has many causes, including the industry already processing
record volumes, before one gets to what is possibly the biggest source: how the
forbearance program, designed for existing loans, is applied to in-transit and
newly made mortgages. More on this subject later.