In 2002 I wrote a column contrasting the
housing finance systems of Denmark and the US. Recently, both systems
have been stressed by the world-wide financial crisis, prompting me to
take another look. I was interested in whether the impact of the crisis
on the two systems revealed anything further about their relative
strengths and weaknesses?
The Danish System
The core of the Danish system is eight
specialized mortgage banks which originate all home mortgages, and a
mortgage bond market where the loans are funded. Each new loan is
immediately sold in the market for the equivalent bond. If the new loan
is a 30-year FRM, for example, it will be sold to investors as an
increase in the balance of the 30-year fixed-rate bond. There are bonds
with fixed and adjustable rates, and within each category there are
separate bonds for different terms.
The Danish system makes it easy for
borrowers to shop for a mortgage. On a given day, all borrowers pay the
same interest rate on a given type of loan. (Borrowers either meet the
credit and other requirements, or they don’t.). The interest rate on a
new mortgage loan is the current market yield on the specific bond that
will fund the loan, plus the mortgage bank’s markup. Bonds are traded on
the Copenhagen stock market, and their yields are readily available.
Danish mortgage banks do not adjust the interest rate for points, nor do
they tack on a series of fixed-dollar charges to cover specific
expenses, as is the practice in the U.S. Total upfront fees are modest
and pretty much the same at all the mortgage banks.
The strength of the Danish system is its
transparency and low origination costs. Its major weakness is that it
does not serve as large a segment of the population as the U.S. system.
Loans are not priced for risk, so borrowers who have poor credit or who
cannot make a down payment of 20%
are not served. In a financial crisis, however, this “weakness” is a
source of strength, as we have recently learned.
Impact of Crisis on Secondary Markets
Both countries were afflicted by the
world-wide loss of confidence in financial institutions. Both
governments responded by guaranteeing the liabilities of banks and other
financial firms, including mortgage banks in Denmark. However, in
Denmark that guarantee did not include mortgage bonds, because it was
not considered necessary. The Danish mortgage bond market continued to
function normally during the crisis, which meant that new loans
continued to be written as before.
In the US, in
contrast, markets in mortgage-backed securities (MBSs) not guaranteed by
Fannie Mae, Freddie Mac or Ginnie Mae ceased functioning. This is why
“jumbo” loans – those too large for purchase or insurance by a
Government entity -- which before the crisis were often placed in MBSs,
are so costly in today’s market relative to conforming loans.
The market for
private MBSs collapsed because investors incurred -- or anticipated that
they might incur -- horrendous losses, whereas investors in Danish
mortgage bonds did not suffer any losses at all! One reason is that the
Danish mortgage bond system is inherently stronger than the MBS system.
A Danish mortgage bond is a liability of the mortgage bank issuing it
and is supported by the capital, reserves and income of the bank, as
well as by the mortgage loans that collateralize that particular bond.
If the collateral supporting one bond happens to suffer large losses,
the bond holders are nonetheless protected by the entire resources of
the bank.
In contrast, in the
U.S. system each private MBS has “credit enhancements”, such as reserve
accounts, excess cash flows, or insurance, designed so that each can
stand on its own. If the credit enhancements on one issue out of 100
turn out to be insufficient, the investors in that issue will suffer
loss, even though the enhancements in the other 99 are excessive.
Further, any such failures can have a contagious effect on the
confidence of investors in other issues, who may wonder whether the
credit enhancements in their issues are adequate.
The second reason
the private MBS market in the U.S. collapsed and the Danish bond market
didn’t is the much higher default rate on mortgages in the US. In
addition, investors incurred larger losses on defaulted mortgages
because defaulting borrowers in the US had less equity.
More Equity in Denmark
Home prices have
declined in Denmark since the crisis began, though not quite as much as
in the US. Single-family home prices were down 15% in the second quarter
from their previous peak, while “owner-occupied flats” were down 28%. As
in the US, price declines have been much larger in some areas than in
others. Despite the price declines, however, the great majority of
Danish borrowers had substantial equity in their homes when the crisis
struck. The widespread negative equity that emerged in the US – a major
factor encouraging defaults and increasing losses when default occurs –
had no counterpart in Denmark.
Why? A major reason is that, in the US,
down payments of less than 20% were the norm well before the bubble
began, and no-down payment loans became increasingly common during the
bubble. When the bubble burst in 2007, therefore, a substantial
proportion of the homes purchased in the prior 2-3 years had no equity.
In contrast, the down payment
requirement in Denmark was 20% well before the bubble and remained 20%
during the bubble. While second mortgages were available from commercial
banks and may have increased in importance during the bubble period, all
mortgage borrowers in Denmark have personal liability which is enforced
by lenders. Losses on second mortgages have been very small compared to
the US.
Erosion of down payment requirements was
only one of the ways that the US housing finance system was weakened
during the bubble period much more than the Danish system, The “quality”
of new borrowers, meaning the array of financial and personal factors
that affect the likelihood that they will default at some point,
deteriorated much more in the US. There was no Danish equivalent of
sub-prime loans to attract tenants into ownership who were not qualified
to be owners. And Denmark did not have alternative documentation rules
that allowed borrowers to claim higher incomes than they actually had.
Erosion of Lending Standards During the
Housing Bubble
The mortgage crisis that erupted in 2007
had its genesis in the prior bubble period, when home prices were rising
rapidly. Price increases reduce the perceived riskiness of mortgages,
encourage investors to accept mortgages that in a stable environment
would be viewed as unacceptably risky, and induce lenders to increase
loan volume by liberalizing their underwriting requirements.
While exact comparisons are not possible, the bubble in Denmark was roughly comparable to the bubble in the US. The average price of single-family homes in Denmark rose 60.4% between the beginning of 2004 and mid-2007 when prices peaked; that was an annual rate of increase of 13.6%. Yet the relaxation of lending standards in Denmark was far less than in the US, which was a major reason why the mortgage defaults arising from the crisis were much smaller in Denmark. Here are some possible reasons why:
Risk Shifting:
In the US system, lenders typically sell their loans, which may
go through several hands before coming to rest in a security or a
portfolio. If potential buyers are willing to accept loans subject to
more liberal underwriting rules, the lenders originating loans will
liberalize them because it expands their market and the increased
default risk is transferred with the sale.
In contrast, the Danish mortgage bank
that originates loans must hold them and retain the default risk. It is
plausible to believe that not having the option of passing the default
risk to a buyer dampens the impulse to liberalize terms.
Regulation: Regulators in the
US did not take any action to curb excessive mortgage liberalization
during the bubble period. Regulatory responsibility for the thousands of
mortgage lenders was divided among four Federal agencies and 50 states.
By the time the Federal agencies got their act together in late 2006,
the damage had already been done.
I don’t know if the Danish regulator DFSA took any steps to restrain liberalization of lending standards during the bubble. The problem, if there was one, was much less severe than in the US. But DFSA could have done what needed to be done because it had sole authority over the 8 firms that originate mortgage loans in Denmark.
Transactions-Oriented Loan Originators:
In the US, the incomes of the mortgage brokers and loan officers with
whom borrowers deal depend almost entirely on the number and size of the
loans they write. They were not responsible for the liberalized
underwriting rules, interest only loans and option ARMs that emerged
during the housing bubble, but they took advantage of these changes to
do more deals, stretching the bubble.
In contrast, prospective borrowers in
Denmark who contact a mortgage bank deal with a salaried employee. While
interest-only loans and new types of ARMs were introduced by mortgage
banks during the bubble period, the initiative for seeking these
instruments remained largely with consumers.
Personal Liability: In the
US, some states such as California do not allow lenders who do not fully
recover what they are owed through foreclosure, to obtain deficiency
judgments against borrowers for the balance, provided the loan was used
to purchase a home. Even in states where deficiency judgments are
allowed, few lenders actively pursue them. In Denmark, however, mortgage
borrowers retain full personal liability and it is actively enforced by
lenders.
Government-Sponsored Enterprises and the Affordability Lobby:
The US had a particularly toxic combination of features that Denmark
lacked: a pair of partly-private/partly-public secondary market agencies
-- Fannie Mae and Freddie Mac -- and a political/social movement aimed
at increasing the homeownership rate. This unholy alliance was by far
the most important cause of the excessive liberalization of mortgage
terms during the bubble. For readers interested in the details, I
recommend Cause and Effect: Government Policies and the Financial
Crisis by Peter J. Wallison, which is available at
www.aei.org.
The two agencies were stockholder-owned
but received an enormously valuable subsidy from the Federal Government
in the form of an implicit guarantee of their liabilities. The quid pro
quo was their active participation in programs to help low-income and
minority households to become homeowners. In 1992, Congress authorized
the Department of Housing and Urban Development (HUD) to set annual
quotas for agency purchases of “affordable loans,” expressed as a
percentage of their total purchases. Initially 30%, these quotas rose
over time, through both the Clinton and Bush administrations, reaching
55% in 2007.
Acquisitions of non-prime mortgage by
the agencies increased rapidly beginning in 2004 and peaked in 2007 with
the onset of the crisis. Wallison estimates their holdings in 2008 at
$1.5 trillion, This total included purchases of private securities in
the market, which HUD allowed the agencies to count toward their quotas
of affordable loans. The horrendous default rates and losses on these
loans sealed the fate of the agencies, which in 2008 were placed in a
Government-administered conservatorship.
Historically, most economists were skeptical or hostile to Fannie Mae and Freddie Mac. Usually, the reason was that only about half of the taxpayer-funded subsidy provided to the agencies was realized as a benefit by borrowers. We now see that that was the least of it. The more compelling argument against the mixed private/public model is that it corrupted the political process and destabilized the system. Not having any equivalent is one important reason why Denmark has weathered the crisis much better than the US.