September 5, 2005, Revised July 29, 2007
In some respects, the United States housing finance system is the best
in the world. In other respects, it is unworthy of a banana republic.
Our housing finance system has a primary market and a secondary market.
The primary market is the market the borrower sees, where loans are
executed. The secondary market is where the loans originated in the
primary market are sold to investors, the ultimate source of funds.
Our secondary market is the envy of the world. Investors acquire
mortgage backed securities at the smallest possible margins over US
Government securities. However, much of the benefit stemming from our
efficient secondary market is eaten away by excessive charges paid by
borrowers to third party settlement service providers, lenders, and
mortgage brokers. Excessive charges persist because, for a variety of
reasons, borrowers cannot shop effectively.
Third Party Settlement Charges Are Excessive
Third parties involved in the lending process include title insurance
companies, mortgage insurance companies, appraisers, credit reporting
agencies, flood insurance companies, and escrow companies. Their costs
are generally higher than they would be if they were purchased in a
normally competitive market.
The Reason: Third party service providers compete not for the favor of
borrowers, who pay their fees, but for the favor of the lenders who
select them. This type of competition is perverse because it drives up
the costs of the service providers. This in turn raises prices to
borrowers or prevents them from falling in response to improvements in
technology.
For example, borrowers pay for private mortgage insurance but the
insurance company (PMI) is selected by the lender. Lenders use their
referral power primarily to benefit themselves. The process is much the
same in the markets for other third party services.
The direct payment of referral fees has long been illegal under the Real
Estate Settlements Procedures Act (RESPA). However, RESPA is ineffective
because it does not eliminate referral power, which is the crux of the
problem. Small players often ignore the rule because HUD, which is
responsible for enforcement, cannot possibly police all the ways in
which one party can transfer something of value to another.
Large lenders circumvent RESPA through circuitous but legal devices,
such as reinsurance affiliates that share the insurance premiums paid by
borrowers. The process of legalizing referral fees increases the costs
that borrowers ultimately pay. So long as referral power is allowed to
persist, the cost to borrowers might be lower if referral fees were paid
openly in cash.
Why Do Borrowers Pay For Services Required by Lenders? We have lived
with this practice for so long that it seems the natural state of
affairs, but in fact, there is nothing natural about it. If automobiles
had to be shopped in the same way as mortgages, the shopper might only
receive the chassis from the dealer, purchasing tires, electrical
system, and painting from third parties. Can there be any doubt what
would happen to the price of these components if, instead of purchasing
them as a package, they had to be purchased separately from vendors
selected by the dealer?
The unbundled package of mortgage services is a historical relic of
usury laws limiting the interest rates lenders could charge. When
lenders could not raise interest rates to cover their costs, it seemed
reasonable to law-makers to allow them to pass costs through to
borrowers. They didn’t realize that this was a sure-fire recipe for
referral abuse. When the abuses became too obvious to ignore, they
responded with RESPA, which made referral fees illegal but left referral
power unchanged. RESPA is essentially a make-work project for lawyers.
The Remedy Is to Eliminate Referral Power: This could be done by the
enactment of one legal rule that is as simple as it is obvious: any
third party service required by lenders must be paid for by lenders.
If lenders paid the charges, they would be included in the rate, of
course, but would cost borrowers far less than now. Competition by third
party providers to sell lenders would force the prices down, and rate
competition by lenders would force them to pass the savings on to
borrowers. Indeed, if lenders had to pay for all these services, they
would discover that some that they had found essential when borrowers
had to pay, were not really necessary after all.
Lender Fees Can Blindside the Borrower
Lender fees, sometimes referred to as "junk fees", are fees charged by
lenders to cover specific lender costs. They are defined in dollars, as
opposed to points which are defined as a percent of the loan. Where
points are one number, junk fees can be a whole bunch of numbers, each
covering a specific charge.
The problem is that some retail lenders increase these fees after it is
too late for the borrower to back out. Often, the borrower finds out
about it at the closing table. It happens because of borrower
inattention, industry locking practices, and terrible disclosure rules.
Borrower Inattention: Borrowers are usually only dimly aware of lender
fees. When they shop, their focus is mainly on interest rate and points,
which are all that appear in media ads. The borrower’s first exposure to
lender fees is likely to be when they receive a Good Faith Estimate of
Settlement (GFE), but this typically doesn’t happen until after an
application has been submitted. At that point, the borrower will be at
least partially committed.
Industry Locking Practice: The mortgage market is highly volatile, with
prices changing from day to day and sometimes within the day. Hence,
rate/point quotes are not binding until the lender locks them. Lender
fees, in contrast, are not volatile, and therefore the practice is not
to include them in locks. The presumption is that at closing, they will
be exactly what they were when the borrower received the GFE, and with
honest lenders, they will be. But with dishonest lenders, the practice
of excluding fees from the lock provides an opportunity to cheat.
The Good Faith Estimate: The GFE makes it all too easy for the cheaters.
Lenders are not bound by any of the numbers on the GFE, which are
"estimates". This is ridiculous, since lenders know their own charges.
In addition, the GFE confuses borrowers by showing each individual
lender fee but no total, when the total is all that really matters. If
the GFE were deliberately designed to take the borrower’s eye off the
ball, it couldn’t have been done better.
Mortgage Brokers Provide Protection: Excessive junk fees generally are
not a problem on loans that go through mortgage brokers. Brokers know
the fees charged by all the lenders with whom they do business, and they
would not accept fee surprises at the closing table that put no money in
the broker’s pocket. Broker fees are another matter, to be discussed in
article 3 of this series.
Home Purchases Are Most Vulnerable: Excessive junk fees are more likely
to arise on a home purchase transaction than on a refinance. On a home
purchase, a buyer cannot walk away from the mortgage without walking
away from the house. On a refinance, in contrast, a borrower can usually
begin anew at any point without much love had a deal changed on them
from what they understood was promised earlier, can use the threat of
rescission to obtain redress at the closing table.
The Remedy: A mandatory fixed-dollar fee on all mortgage transactions,
proposed below as the way to make it possible for borrowers to shop
effectively, would also eliminate fee escalation at the closing table.
The maximum fixed-dollar fee could, indeed, be zero. Bank of America has
adopted zero fees on purchase transactions as a competitive strategy,
which is demonstration enough that it is practical. See
NoFee Mortgage Plus.
A less comprehensive approach that would also work is a rule stipulating
that when lenders lock the rate and points, they also lock their fees.
Mortgage Broker Fees Are Excessive
Because brokers deal with multiple lenders, they play a critical role in
helping a borrower find a lender who offers a particular type of loan
program. When the needed loan is one offered by many lenders, brokers
are able to shop among them to find the lowest price. That’s the good
news.
Excessive Fees: The bad news is that, broker charges per transaction are
generally excessive. In part, this is due to low productivity. Brokers
spend a lot of time looking for clients, and they also spend a lot of
time with potential clients who don’t close and waste their time. Low
productivity generates pressure to earn more on the deals that do close.
Brokers are able to charge a lot per transaction because borrowers
usually don’t know at the outset how much the broker will make. If they
find out, usually the deal is too far advanced to do anything about it.
Only the loosest relationship exists, furthermore, between broker
charges and the amount of work the broker does for the borrower. The
general rule is that brokers charge what the market will bear.
Unsophisticated borrowers who visit a single broker will generally pay
more than knowledgeable borrowers who shop alternative sources.
The Independent Contractor Model of the Industry: The dominant ideology
of mortgage brokerage, as promulgated by the National Association of
Mortgage Brokers and the various state associations, is that brokers are
independent contractors. They view themselves as merchants who buy at
one price and sell at another price, and how much they make on a
transaction is no one’s business but their own. The independent
contractor model supports the view that brokers are entitled to make as
much per transaction as they can.
The Independent Contractor Model Generates Distrust, Which Increases
Costs: Distrust runs like a red line through the hundreds of letters I
receive every month from borrowers relating experiences with brokers.
And distrust translates into higher costs.
Brokers detest borrowers who flit from one broker to another, submit
applications through multiple brokers, or pump them for information and
then deal elsewhere. Yet these practices arise from attempts by
borrowers to protect themselves against brokers they don’t trust.
Borrower reactions to distrust raise broker costs, which pressures
brokers to make more per transaction, which generates more distrust in a
vicious circle.
Other Fallacies of the Independent Contractor Model: The fact is that
brokers are service providers, not merchants; they do not buy and resell
anything. Furthermore, shopping mortgages is so difficult that few
borrowers can do it effectively. Brokers are the experts at shopping
mortgages, not borrowers. The optimal arrangement for most borrowers,
therefore, is to purchase the shopping expertise of brokers for a fixed
fee. Fortunately, it is now possible to do this.
The Agency Approach of Upfront Mortgage Brokers: Upfront Mortgage
Brokers (UMBs) operate according to a different set of rules than the
remainder of the industry. UMBs view themselves as the agent of the
borrower, to whom they owe a fiduciary responsibility. A UMB agrees with
the borrower on total broker compensation from the transaction, and
passes through the best price from the broker’s lenders.
The advantage of the UMB approach is that it breeds confidence, which
lowers costs and increases productivity. I know UMBs who charge half the
industry average per transaction but close 3-4 times as many loans.
Their secret is a continuous stream of referrals from previous clients
-- and from me.
Implementation of the Agency Approach: The way to break the circle of
distrust is to change the operating model, from independent contractor
to agency. The broker trade associations will never do this, because
they cater to the lowest common denominator of member opinion.
Government should but probably won’t mandate the agency approach because
it would be opposed not only by the broker associations, but also by the
wholesale lenders, who are as short-sighted as the brokers. They support
the independent contractor model in order to limit their own liability
for broker misdeeds.
One misgiving I have about enacting an agency requirement into law is
that it will be defined so vaguely as to be useless. At worst, it could
be an indirect way to enact a suitability requirement into law. Any
definition that requires extensive investigation to determine whether or
not a violation has occurred will not work in an industry with 50,000 or
so players, most of which are very small. The UMB requirements are quite
precise, in effect allowing the borrower to be the effective enforcement
agent. See
Commitment of an Upfront Mortgage Broker.
An Alternative Approach: Yield Spread Premiums Credited to Borrowers: A
simpler approach that does not carry the risk of insinuating a
suitability standard into the marketplace is to require that all yield
spread premiums (YSPs) paid by wholesale lenders to brokers and to
correspondent lenders be credited to borrowers. The borrowers would have
to authorize their payment to the loan provider. Since overcharges arise
almost entirely from YSPs, this would do the trick in a very direct way:
the borrower would know exactly how much is being paid out of his
pocket. See
Dealing With Yield Spread Premium Abuse.
Mortgage Shopping by Borrowers Is Ineffective
Shopping is extremely difficult now because mortgages have so many price
dimensions. Even if lenders paid for all the services provided by third
parties, which I proposed above, a borrower shopping for an FRM would
have three prices to juggle: interest rate, points, and fixed dollar
fee. (On ARMs, there are more, but I’ll ignore that for now). This is
confusing and makes shopping difficult.
Mandating a Fixed Fee: Government should mandate the same fixed-dollar
charge for all lenders and all programs. Then borrowers would have to
shop only rate and points, which is easily manageable. For example, a
borrower can shop for the lowest rate at zero points, or the fewest
points on a 6% loan.
Within reasonable limits, the exact amount of the fixed charge is not
important, provided the charge is the same for every lender and every
loan. It is the variability in these costs that makes it difficult to
shop.
This is price-fixing by Government, but in a good cause. When a service
carries one price, price-fixing invariably reduces the supply of the
service. When a service carries three prices, however, fixing one price
merely channels market adjustments into the remaining prices, making it
easier for consumers to shop.
A Fixed Fee Would Discourage Predatory Lending: Perhaps the greatest
benefit of the one-price rule would be in the sub-prime market, where
predatory practices are widespread. Under a fixed-charge rule, these
practices become much more difficult to execute.
A common feature of price gouging, for example, is the inclusion of
large fees in the loan balance, which borrowers often know nothing about
until they get to the closing table. With a fixed-price rule, along with
a rule that limits the financing of points, any gouging would have to be
in the interest rate, where the potential for snookering the borrower is
limited. Even unsophisticated borrowers understand the difference
between 7% and 12%.
In a similar vein, making loans that borrowers can’t repay, or churning
loans in successive cash-out refinances, are profitable only if the
lender can load heavy fees into the balance. The fixed-charge rule
should eliminate both practices.
Enforcement of a Fixed Fee Rule Would Be Easy: Law-makers sometimes give
little consideration to whether, and at what cost, the rules they
promulgate can be enforced. This is certainly true of the existing law
against the payment of referral fees, which couldn’t be effectively
enforced with an army of examiners. State laws directed at predatory
lending that bar loans that "fail to benefit the borrower," or that
"borrowers do not have the capacity to repay," present similar
enforcement problems.
In contrast, the fixed-charge rule would be virtually self-enforcing,
because it is unambiguous, and every borrower would be a potential
enforcement agent. Borrowers would know what the allowable charge was,
and the closing documents would reveal whether or not the lender was in
compliance.
How Much Should the Fixed Fee Be? The fee should not be so high that
lenders can make money on the origination process, regardless of what
happens later. That encourages abusive practices. Nor should it be so
low that early prepayment will cause the lender serious loss because
then there won’t be any loans made without prepayment penalties. $3,000
is about right for now. This is the fee set by Innovation Mortgage, a
sub-prime lender out of California, which has adopted a one-price rule
voluntarily as a marketing tool.
Note: since the above was written, Bank of America has voluntarily
adopted a zero fee on purchase loans, and without a prepayment penalty.
They don't offer sub-prime loans under this program, however.
Eliminating Low-Ball Price Quotes: If price quotes can’t be depended on,
price shopping goes for naught. Some lenders and brokers routinely offer
low-ball price quotes designed to capture the customer, which they
retract at the time the price is locked. Because the market is so
volatile, borrowers are rarely positioned to contest a loan provider’s
statement that market rates increased more, or decreased less, than they
did in fact.
Full protection against this common tactic requires a "twin sibling
rule." Lenders must lock at the same rate that they would quote to the
borrower’s twin who is shopping the same loan on the lock date. This
rule would also be easy to enforce, since loan providers do quote prices
to shoppers on the same days that they lock prices to borrowers in
process.