- EXAMINING THE HOUSING FINANCE SYSTEM: THE TO-BE-ANNOUNCED MARKET

[Senate Hearing 112-269]
[From the U.S. Government Publishing Office]

S. Hrg. 112-269

EXAMINING THE HOUSING FINANCE SYSTEM: THE TO-BE-ANNOUNCED MARKET

=======================================================================

HEARING

before the

SUBCOMMITTEE ON
SECURITIES, INSURANCE, AND INVESTMENT

of the

COMMITTEE ON
BANKING,HOUSING,AND URBAN AFFAIRS
UNITED STATES SENATE

ONE HUNDRED TWELFTH CONGRESS

FIRST SESSION

THE EXAMINATION OF THE SECURITIZATION MARKETS WITH SPECIFIC FOCUS ON A
PART OF THE SECURITIZATION SYSTEM IMPORTANT TO HOUSING FINANCE--THE TO-
BE-ANNOUNCED OR TBA MARKET

__________

AUGUST 3, 2011

__________

Printed for the use of the Committee on Banking, Housing, and Urban
Affairs

Available at: http: //www.fdsys.gov /

U.S. GOVERNMENT PRINTING OFFICE
72-962 WASHINGTON : 2012
-----------------------------------------------------------------------
For sale by the Superintendent of Documents, U.S. Government Printing Office,
http://bookstore.gpo.gov. For more information, contact the GPO Customer Contact Center, U.S. Government Printing Office. Phone 202�09512�091800, or 866�09512�091800 (toll-free). E-mail, [email protected].

COMMITTEE ON BANKING, HOUSING, AND URBAN AFFAIRS

TIM JOHNSON, South Dakota, Chairman

JACK REED, Rhode Island RICHARD C. SHELBY, Alabama
CHARLES E. SCHUMER, New York MIKE CRAPO, Idaho
ROBERT MENENDEZ, New Jersey BOB CORKER, Tennessee
DANIEL K. AKAKA, Hawaii JIM DeMINT, South Carolina
SHERROD BROWN, Ohio DAVID VITTER, Louisiana
JON TESTER, Montana MIKE JOHANNS, Nebraska
HERB KOHL, Wisconsin PATRICK J. TOOMEY, Pennsylvania
MARK R. WARNER, Virginia MARK KIRK, Illinois
JEFF MERKLEY, Oregon JERRY MORAN, Kansas
MICHAEL F. BENNET, Colorado ROGER F. WICKER, Mississippi
KAY HAGAN, North Carolina

Dwight Fettig, Staff Director

William D. Duhnke, Republican Staff Director

Dawn Ratliff, Chief Clerk

William Fields, Hearing Clerk

Shelvin Simmons, IT Director

Jim Crowell, Editor

______

Subcommittee on Securities, Insurance, and Investment

JACK REED, Rhode Island, Chairman

MIKE CRAPO, Idaho, Ranking Republican Member

CHARLES E. SCHUMER, New York PATRICK J. TOOMEY, Pennsylvania
ROBERT MENENDEZ, New Jersey MARK KIRK, Illinois
DANIEL K. AKAKA, Hawaii BOB CORKER, Tennessee
HERB KOHL, Wisconsin JIM DeMINT, South Carolina
MARK R. WARNER, Virginia DAVID VITTER, Louisiana
JEFF MERKLEY, Oregon JERRY MORAN, Kansas
MICHAEL F. BENNET, Colorado ROGER F. WICKER, Mississippi
KAY HAGAN, North Carolina
TIM JOHNSON, South Dakota

Kara Stein, Subcommittee Staff Director

Robert Peak, Fellow

Gregg Richards, Republican Subcommittee Staff Director

(ii)

C O N T E N T S

----------

WEDNESDAY, AUGUST 3, 2011

Page

Opening statement of Chairman Reed............................... 1

Opening statements, comments, or prepared statements of:
Senator Crapo................................................ 2

WITNESSES

Thomas Hamilton, Managing Director, Barclays Capital, on behalf
of the Securities Industry and Financial Markets Association... 3
Prepared statement........................................... 20
Paul T. Van Valkenburg, Principal, Mortgage Industry Advisory
Corporation.................................................... 5
Prepared statement........................................... 34
Andrew Davidson, President, Andrew Davidson & Co., Inc........... 6
Prepared statement........................................... 49

(iii)

EXAMINING THE HOUSING FINANCE SYSTEM: THE TO-BE-ANNOUNCED MARKET

----------

WEDNESDAY, AUGUST 3, 2011

U.S. Senate,
Subcommittee on Securities, Insurance,
and Investment,
Committee on Banking, Housing, and Urban Affairs,
Washington, DC.
The Subcommittee met at 9:33 a.m. in room SD-538, Dirksen
Senate Office Building, Hon. Jack Reed, Chairman of the
Subcommittee, presiding.

OPENING STATEMENT OF SENATOR JACK REED

Senator Reed. I would like to call the hearing to order.
I first want to thank Senator Crapo for joining us today
and also for his collaboration over the many, many months as
the Ranking Member of this Subcommittee, and thank you very
much, Senator, for your excellent work. Let me also welcome our
witnesses.
This morning we are examining the housing finance system,
specifically the to-be-announced market. In May, the
Subcommittee conducted a hearing on the state of the
securitization market. This morning's hearing continues the
Subcommittee's examination of the securitization markets with
particularly focus on a part of the securitization system
important to housing finance, and this is the to-be-announced
or TBA market.
In the early 1970s, the TBA market began as a trading venue
for securities that were issued by Ginnie Mae, and its role
expanded as Fannie Mae and Freddie Mac began issuing mortgage-
backed securities.
This market has evolved, and today it is one of the deepest
and most liquid markets in the United States, with average
daily trading volume in excess of $320 billion--a market second
only to the market for United States Treasury securities.
The name of the market, to-be-announced, comes from the way
the market functions. Unlike a traditional marketplace,
investors do not know the specific collateral or pools of loans
they are agreeing to purchase until months later. Accordingly,
the collateral is designated to be announced at a date in the
future.
Many argue that the TBA market is vital for preserving key
products that consumers have come to rely upon in buying a
home, such as a 30-year fixed-rate loan, freely prepayable
mortgages, and the ability to lock in an interest rate prior to
closing a mortgage. Although the only securities traded in the
TBA market are agency securities, defined as those securities
issued or guaranteed by Fannie Mae, Freddie Mac, and Ginnie
Mae, the TBA also serves as a benchmark for privately issued
securities. Privately issued securities are priced relative to
the TBA price. In addition, originators use the TBA market to
purchase and sell positions to hedge the origination of loans
that are not eligible for trading in TBA, for example,
adjustable rate loans and jumbo loans.
As we continue to explore different approaches to reforming
our housing finance system, it is critically important that we
understand how the TBA market works and what impact any reforms
will have on this market. How will any changes affect the
availability of the standard mortgage products sought by
consumers? What characteristics of this market, if any, should
be preserved? I look forward to hearing from all of our
witnesses this morning on these issues. In fact, this is a very
technical but vitally important part of our securitization and
our mortgage industry. We have to understand the basics before
we move forward. But as with all of our hearings, the point is
to accumulate the information and the insight so that we can
start dealing with some major issues with respect to housing or
with respect to the GSEs.
With that, I would like to introduce the Ranking Member.
Senator?

STATEMENT OF SENATOR MIKE CRAPO

Senator Crapo. Thank you very much, Mr. Chairman, and I
appreciate your kind words as well. I enjoy our working
relationship, and I appreciate the opportunity for us to have
this hearing on the TBA market.
I share your view that the TBA market serves a valuable
role in the mortgage finance system, and we need to better
understand its mechanics as we move forward with housing
finance reform. Today's witnesses all have a deep expertise
with respect to the TBA market and will be able to explain how
the TBA market allows mortgage originators to hedge the risk of
a change in interest rates between the time that the mortgage
is locked in and the time that the mortgage is actually closed
and securitized.
The main components of the TBA market are the
standardization and homogeneity of the securities and their
market practices and Government guarantee of timely payments.
According to a Federal Reserve staff report on the TBA market,
the deep liquidity of agency mortgage-backed securities cannot
be attributed solely to the implicit Government guarantee of
mortgage-backed securities.
Going forward, the question is how the TBA market might
change and develop without a Fannie Mae and Freddie Mac
implicit or explicit guarantee. And what are the tradeoffs that
we need to consider? To be able to answer that question, we
need to understand the feasibility of creating a private TBA
market over time and whether it would have the liquidity
sufficient for mortgage companies to hedge their interest rate
risk.
Again, Mr. Chairman, I appreciate this hearing and look
forward to what our witnesses will share with us today.
Senator Reed. Thank you very much.
We have been joined by Senator Corker. Senator, would you
like to make some opening comments?
Senator Corker. No. I am looking forward to the witnesses,
as usual, and I appreciate you calling the hearing.
Senator Reed. Thank you very much, Senator.
Let me now introduce our panel. First is Mr. Thomas
Hamilton. Mr. Hamilton is managing director at Barclays
Capital. Mr. Hamilton heads securitized product trading at
Barclays, including residential and commercial mortgage-based
securities, as well as other asset-backed securities. He joined
Barclays Capital in 2004 after 15 years with Citigroup, where
he was a managing director. Mr. Hamilton is currently the
chairman of the Securitized Products Division of SIFMA and has
held that position for 7 years.
Paul Van Valkenburg is a principal at Mortgage Industry
Advisory Corporation, MIAC, a registered investment adviser
which he helped to form. Prior to forming MIAC, he worked in
mortgage research with major Wall Street firms, including
Goldman Sachs & Company and Drexel Burnham Lambert. He has
worked extensively valuing loan portfolios for lending
institutions actively restructuring their asset/liability
composition; developing pricing models for whole loans, CMOs,
and stripped mortgage-backed securities; and analyzing the
prepayment risk in mortgage-backed structured products.
Selections of his research have been published in the ``The
Handbook of Mortgage-Backed Securities'' and ``Interest Rate
Risk Models: Theory and Practice.''
Andrew Davidson is the president of Andrew Davidson &
Company, a New York firm specializing in the development and
application of analytical tools for the mortgage-backed
securities market that serves over 150 financial institutions.
He has written extensively on mortgage-backed securities
product development, valuation, and hedging. Prior to the
founding of Andrew Davidson & Company in 1992, he worked at
Merrill Lynch where he was a managing director in charge of 60
financial and systems analysts.
We will begin with Mr. Hamilton. Gentlemen, your written
testimony has been made a part of the record. Please feel free
to summarize and make points.
Mr. Hamilton, please.

STATEMENT OF THOMAS HAMILTON, MANAGING DIRECTOR, BARCLAYS
CAPITAL, ON BEHALF OF THE SECURITIES INDUSTRY AND FINANCIAL
MARKETS ASSOCIATION

Mr. Hamilton. Good morning, Chairman Reed, Ranking Member
Crapo, and Members of the Subcommittee. I am Thomas Hamilton,
managing director at Barclays Capital, where I am responsible
for the securitized products trading business. I am pleased to
testify today on behalf of the Securities Industry and
Financial Markets Association.
Housing is a critical component of our economy and is at
the center of a virtuous circle: housing begets jobs, which
beget housing. Consequently, the U.S. mortgage market is
enormous. For example, the home mortgage market is
approximately equal in size to the total size of U.S. bank
balance sheets. Given that banks engage in activities other
than residential mortgage lending, their balance sheets alone
cannot meet the country's need for mortgage credit. Ginnie Mae,
Fannie Mae, Freddie Mac, and private institutions use the
process of securitization to provide capital that allows for
the growth of mortgage lending beyond the capacity of bank
balance sheets. Today private securitization and the agencies
finance nearly 70 percent of outstanding home mortgages, and it
is, therefore, imperative that securitization continue to play
a key role in any future mortgage finance system. The market
for MBS issued by the agencies is approximately three times the
size of the outstanding non-agency private label MBS market. In
this agency MBS market, the TBA market is the single largest
component. The TBA market is currently the key to funding
mortgage lending, and because of this it plays a critical role
in housing and the U.S. economy.
The enduring liquidity of the TBA market, in contrast to
the lack of issuance in the private label MBS markets,
preserved the availability of mortgage credit in the recent
crisis. This ability to maintain liquidity during stress
periods is a key benefit of the TBA market. Furthermore, the
liquidity and resilience of the TBA market attracts a wide
range of investors who provide vast capital that is cycled into
mortgage lending, including retirement savings vehicles,
insurance companies, and foreign investors.
The vast liquidity and forward-trading nature of the TBA
market provides key benefits to consumers, such as the broad
availability of 30-year fixed-rate mortgages that may be
prepaid without penalty, and significant and consistent
liquidity in the secondary mortgage market. This results in a
stable and attractive funding source for lenders that allows
them to provide lower mortgage rates and longer-term ``rate
locks'' for borrowers and efficiently recycles funds back to
local lenders to enable another round of mortgage lending.
Of course, the TBA market is facilitated by the guarantees
of the agencies and, therefore, the support of the Government
that stands behind them. Currently over 90 percent of mortgages
are financed through a program of one of these three agencies.
This level of support is possible because agency MBS do not
expose investors to credit risk, and, therefore, the market is
attractive to risk-averse investors that have vast sums of
capital available for investment. Without the TBA market, we
believe that the majority of this investment capital would be
directed elsewhere, reducing the amount of funding for and
raising the cost of mortgage lending. Therefore, SIFMA believes
strongly that maintaining a liquid and viable TBA market should
be considered as Congress addresses housing finance reform.
With that said, the reality is that that this current
outsized role of the Government is not sustainable over the
long term and should be reduced. The TBA market's role and the
Government's role should shrink as the private markets
regenerate over time. The means of achieving this rebalancing
are very complicated and consequential on a national,
financial, and personal level. While SIFMA believes the TBA
market should play a role in the future, it should certainly
not be 90 percent of the market. There are a number of
challenges to the resurrection of the private label non-agency
market, including the significant uncertainty faced by non-
agency MBS investors and issuers. The rules of the road for
both sides are not clear. Until they are, it will be
challenging for issuers and investors to see eye to eye on
securitization transactions, at least in the volume and
frequency that will be necessary to fund mortgage credit
demand. Being able to withdraw the Government from mortgage
markets will require a carefully planned and sequenced
transition which should take many, many years. It is essential
to remember that the necessary volume of non-agency investors
will not simply appear because we would like them to. They must
be drawn back in and made comfortable with private label
securitization and its regulatory environment.
We believe that it is critical that the planning and
execution of significant changes to the funding of mortgage
loans be done with attention to detail, be based on sound
analysis of costs and benefits, be mindful of unintended
consequences, and create a long-term beneficial and stable
environment. While we cannot predict the future, we can use the
past as a guide and apply lessons learned and mistakes made,
the good and the bad, to design a system that will stand the
test of time.
We hope that the testimony we present today will he helpful
in educating policymakers about how mortgage loans are funded
in the capital markets, the critical role of the TBA market,
and some of the critical issues that must be considered to move
forward.
Thank you for this opportunity, and I am happy to take any
questions.
Senator Reed. Thank you very much, Mr. Hamilton.
Mr. Van Valkenburg.

STATEMENT OF PAUL T. VAN VALKENBURG, PRINCIPAL, MORTGAGE
INDUSTRY ADVISORY CORPORATION

Mr. Van Valkenburg. Thank you. Good morning, Members of the
Subcommittee. Thank you for the opportunity to testify before
the Subcommittee today on the current and prospective role of
the TBA market in our current housing finance system. In my
written testimony, I offer a detailed description of how the
current TBA market interacts with the mortgage industry and
prospective borrowers. It is a complex process, but hopefully I
have made it understandable and useful to the Committee.
The TBA markets provide the exit price for long-term fixed-
rate mortgages and enable the borrower to accept capital that
otherwise would not be available.
Today the TBA market is the principal mechanism for the
flow of capital into the current housing finance system. Any
proposed new solution must preserve the TBA market liquidity in
order to enable mortgage companies and their borrowers to
access this capital efficiently.
Given that the GSEs are currently in conservatorship, the
current and prior system is clearly flawed. I would argue that
the principal causes of the failure were the underreserved and
undercapitalized GSEs against the unanticipated credit events,
requiring the GSEs to underwrite U.S. mortgage credit risk and
then restricting them to only investing in U.S. residential
mortgage investments; a mispricing of their guarantee fees; a
loosening of the loan underwriting standards; and the last of
independence from political goals.
Moreover, a fully functioning housing system should share
the goals of the Administration's Option 1. To quote,
``minimize distortions in capital allocation across sectors,
reduce moral hazard in mortgage lending, and drastically reduce
direct taxpayer exposure to private lenders' losses.''
A stated concern over this option from the Treasury
Secretary is that the mortgage credit risk will be transferred
to the banking system and, as a result, expose the Deposit
Insurance Fund to this risk. I disagree with this conclusion. I
believe that the credit risk provides could be a mix of private
mortgage insurance companies, credit default swap protection
writers, and if necessary, Government guarantors. Some of the
risk could be absorbed in the banking system, but not all of
it.
The GSEs currently provide 30-year guarantees to investors
in MBS pools. This guarantee is a particular case of pool
insurance. This guarantee is also a particular case of a credit
default swap. When asked what a credit default swap is, I
usually answer the GSE guarantee is a classic example because
it is the largest credit default protection writer. A mortgage
pool or company pays the GSE a guarantee fee in exchange for a
guarantee of timely payment of principal and interest by the
GSE on the mortgage pool.
A significant difference between the GSE guarantee and an
actual CDS is that an actual CDS trades in an active market
with real price discovery and real risk transference. The
amount of credit risk of mortgages that have a GSE guarantee is
enormous, and I believe there is ample room for a private
market to develop to price and exchange some or all of this
mortgage credit risk. I believe that private financial
institutions will be able to price and trade this risk and, as
a result, spread this risk across the financial system and
reduce the exposure to taxpayers. With such a mechanism to
price and trade credit risk, the TBA investor will be
protected, the mortgage borrower will be benefited, and our
systematic risk will be reduced.
I believe that the preference of the Committee should be to
explore how to create a privately guaranteed residential credit
market that will either solely private or, if the costs are too
high as the market develops, a hybrid of private and Government
guarantees. The development of a solely private market would
take time and involve continual oversight. The GSEs exist
today, and we have time to transition to a private credit
market and allow it to fully develop.
I believe the Committee should explore private market or
hybrid private-Government solutions to avoid or reduce the
problems of Government guarantees.
I believe that should a private market solution be
developed, the liquidity of the future TBA market would be
sufficient for mortgage companies to hedge their interest rate
risk, and the systemic benefits would be substantial.
Thank you.
Senator Reed. Thank you very much, Mr. Van Valkenburg.
Mr. Davidson, please.

STATEMENT OF ANDREW DAVIDSON, PRESIDENT, ANDREW DAVIDSON & CO.,
INC.

Mr. Davidson. Chairman Reed, Ranking Member Crapo, Mr.
Corker, as you have heard from the other witnesses and from our
written statements, the TBA market plays a central role in the
mortgage market. With monthly trading totaling more than $5
trillion, it is truly an incredible achievement of our
financial system.
The TBA market helps lower mortgage rates, facilitates rate
locks for borrowers seeking to buy homes, and made mortgages
available through the financial crisis. I appreciate the
opportunity to discuss this important market with you. For the
next few minutes, I would like to highlight some of the key
points of my written statement.
While we can point to features of a market that make it
useful, it is not easy to predict which market innovations will
succeed. Much of the success depends on the confidence of the
participants. A shift in confidence can lead to a rapid change
in the viability of a market. Currently the TBA market enjoys a
substantial degree of confidence. This confidence, I believe,
is not just a result of good institutional design but also a
long history of successful adaptation to change.
Whether or not the TBA market will be able to adapt to the
proposed changes in the structure of the housing finance market
and the GSEs is difficult to predict with complete accuracy.
While I cannot provide you with a definitive answer, I can give
you my views on several proposed changes based on my 25 years
of experience in the mortgage markets.
The most important of the proposed changes would be the
elimination of the Government guarantee on conventional loans.
While the guarantee was only implicit until the conservatorship
of Fannie Mae and Freddie Mac, the guarantee has played an
important role in the structure of the TBA market. I do not
believe that the TBA market could survive the loss of the
Government guarantee.
However, if the TBA does not survive, the market will
develop other mechanisms to facilitate hedging and funding of
mortgage loans. However, mortgages, especially fixed-rate
mortgages, would be more expensive, less available, and more
subject to market disruptions.
Other proposals suggest increasing the number of issuers
for guaranteed MBS. The idea is that multiple issuers would
increase competition and decrease the concentration risk. While
such a proposal has other benefits, I believe this would be
negative for the TBA market and that multiple issuers will make
establishing good delivery rules more complex and less
workable. In the end, it is likely that one or two issuers
would dominate the market.
Some proposals recognize this problem with multiple issuers
and recommend a single Government issuer with multiple
insurers. Such a proposal is probably consistent with the
survival of the TBA market, as Ginnie Mae already operates in a
similar fashion. A single Government program for all mortgages,
however, does run the risk that the issuer will be unable to
adapt to changing conditions and will be less flexible and
adaptable than the GSEs have been.
Many proposals require that any Government guarantee be on
MBS and not on other obligations of the guarantor, and this is
likely a positive step for the TBA market. If the GSE or its
successors are primarily focused on securitization, then they
will likely act to continue to maintain and improve the value
and liquidity of the MBS and the TBA market.
Many proposals recommend that the Government guarantee be
only a catastrophic guarantee that reduces risk to taxpayer
while enhancing the liquidity of MBS. Such an approach is
likely to be consistent with the TBA market, provided that
investors in TBA-eligible mortgages do not face credit risk.
This means that the credit risk must be absorbed by private
capital outside the TBA mechanism and the Government guarantee
must fully protect investors in the TBA-eligible mortgages.
The approach that I favor is to provide additional private
capital in the form of subordinate bonds. Private capital would
provide funding for the subordinate bonds while the Government
could provide a guarantee on senior bonds. The Government would
be protected from loss by private capital but would facilitate
liquidity on the senior guaranteed bonds. Such a program could
be structured in a way to be consistent with the TBA market.
As important as the direction of reform is the pace of
change. Given the weak state of the housing market and the lack
of currently viable alternatives to Government-guaranteed MBS,
it would be disruptive to move too quickly to eliminate Fannie
Mae and Freddie Mac and replace them with an alternative
structure, even if that alternative were better designed and
more economically sound.
On the other hand, inaction also poses dangers as most of
the mortgage loans are still reliant on Government guarantees
and conservatorship is not a viable long-term option.
Instead of either wholesale replacement of the GSEs or not
taking any action at all, I believe it is possible to transform
the existing GSEs step by step to a new system. In particular,
I recommend that the GSEs be encouraged or even required to
seek forms of private capital to stand in front of the
taxpayers. Even while in conservatorship, GSEs can experiment
with mortgage insurance, and subordinated bonds structures that
could be used as templates for the longer-term restructuring of
the housing finance system.
Thank you for your interest in my comments. I look forward
to your questions.
Senator Reed. Well, thank you all, gentlemen, for very
excellent testimony about a very important and very challenging
topic.
We are going to do 7-minute rounds, but I would be happy to
entertain a second round if there are additional questions. And
we have the luxury with the excellent panelists and three--not
33--Senators to take some time. So let me begin with a question
to all the panelists.
You have all highlighted the fact that the TBA market does
affect the availability of certain mortgage products--30-year
fixed mortgage loans, the ability to lock in interest rates, et
cetera. Changes to this that we are talking about, how would
they affect these characteristics? I guess the other way to ask
it, too, is: Is this ultimately going to be a tradeoff in terms
of what we expect of a mortgage today, fixed rates, locked in,
30 years? How is this all going to interact? I would just like
your general comments, starting with Mr. Hamilton and down the
row.
Mr. Hamilton. I think there is definitely a tradeoff. I
think one of the large benefits of the TBA sector and of the
agency guarantee is that we were able to get 30-year fixed
mortgages and keep monthly payments low. I think the
elimination of that will absolutely force us into either a
floating rate market or something that is certainly of shorter
duration and more volatility for the homeowner in their monthly
payments.
Senator Reed. Mr. Van Valkenburg.
Mr. Van Valkenburg. Yes, I think as it stands now, the
Government guarantee is necessary to enable the risk to be
absorbed by the investor and transferred from the homeowner. So
I think the guarantee is functioning that today.
The mechanism that the borrowers in the market, the TBA
serving the mortgage companies, they are basically price
takers. They take the information, and if they have an outlet
to sell the loans, they will use the mechanics in place. So
having a liquid market and having a liquid outlet for 30-year
mortgages is the means by which they can execute that
transaction.
Senator Reed. Thank you.
Mr. Davidson, your comments just in general.
Mr. Davidson. So certainly without some guarantee, we are
likely to have far fewer fixed-rate mortgages. Those rates
would be higher. But I think probably more importantly than any
of those is that the stability of the availability of mortgage
credit would be much lower. We have seen the private markets,
when you go through a shock to the financial system, just step
back for a while, and then it takes a while for them to
recover. So during that time period, mortgages would just be
much less available.
Senator Reed. Let me sort of ask the question again,
starting with you, Mr. Davidson, saying that a lot depends in
terms of where we come out is what goals we have when we go
into it. And if the goal is to maintain that which people
assume is the American mortgage--long term, maybe not 30 years
but 20, et cetera; fixed rates and relatively low monthly
payments--can we do that without a guarantee by the Federal
Government in some way, shape, or form?
Mr. Davidson. You know, it is hard to say it would not
happen. It is certainly much less likely that we would have the
same percentage in 30-year fixed-rate mortgages without the
Government guarantee.
It turns out there are some investors who want to buy
interest rate risk and can take prepayment risk, but do not
want to deal in credit risk. They want to be able to engage in
transactions where they can buy hundreds of millions or
multiple hundreds of millions of dollars worth of securities at
one point in time. And without removing the credit risk, it is
difficult to see how you could create that kind of market.
Senator Reed. Mr. Van Valkenburg.
Mr. Van Valkenburg. Yes, I think if we said tomorrow, hey,
let us start a market, it would obviously fail to transfer the
risk involved in a 30-year mortgage. But if we could develop a
process where this credit piece could be transferred into
private holders of the risk, everything comes down to a price
at what it is worth, and we do not know what the price would be
at this point because the Government is basically assuming that
risk and subsidizing that risk. So a market would have to take
years to develop in order for that risk to be priced and risk
takers and markets to develop before we could understand really
what the potential investors would require in a fee in order to
absorb that risk.
Senator Reed. And, Mr. Hamilton, your comments.
Mr. Hamilton. I would just say the market will develop for
whatever the rules are brought to them, but what I would say is
certainly the 30-year mortgage would be less available.
Mortgage credit would be less available, and the credit
availability that you would be able to access would be at a
significantly higher price and would have a significant impact
on the housing market.
Senator Reed. Let me ask another question which has been
alluded to by all your testimony, and that is, right now there
is a lot of capital going into this market because of the way
it is structured, the guarantee, the credit risk aspects, and
the presumption, I think, the conclusion from all your comments
is that in some respects that capital will not go there any
longer if the guarantee is changed significantly or it is a
private system. And it raises a question. One, where does the
capital go? You might not know. But is that a bad thing or a
good thing and in terms of the overall economic performance of
the country? This is very speculative, but feel free.
Mr. Hamilton. There is a lot of foreign investment, foreign
capital, insurance company money management, retirement funds
that are putting large amounts of dollars into the U.S.
mortgage market, and that is due to the fact that they do not
want to take credit risk.
Is there a market for that credit risk? I believe there is.
But I think we are talking about a very large transition from a
mortgage market that is determined by rates and investors who
care about rates to an investor who cares about credit, and
that transitioning can take--you know, to do it in an orderly
fashion, we are talking about 10 or 15 years. This is not an
easy scenario.
Senator Reed. Mr. Van Valkenburg and then Mr. Davidson, and
then my time is done.
Mr. Van Valkenburg. Yes. Anytime you subsidize anything,
you get more of it, so I think--and the Government is
subsidizing the housing sector in many ways, you know, along
the food chain, particularly the guarantee is just one area,
but--so we probably get a little more access to consumer credit
than we would otherwise. But it is hard to quantify these
things without any kind of real price discovery about what
things are traded at or what private institutions would pay for
that risk.
Senator Reed. And part of this transition phase would in a
sense be that price discovery process.
Mr. Van Valkenburg. Yes, if we could develop credit markets
where we could transfer that to the private financial
institutions, we could begin to understand what those costs
are, what private parties are willing to pay for that risk, and
move the credit risk across the world instead of just
localizing it and concentrating it on the Government's balance
sheet.
Senator Reed. Mr. Davidson, your comments, please.
Mr. Davidson. I think it is important to separate the
liquidity function of the guarantee from the credit function of
the guarantee. On the credit function, I think the credit risk
should be moved into the private market. It was supposed to be
there before, just that Fannie and Freddie did not really have
enough capital.
The liquidity function is a lot like deposit insurance, and
deposit insurance keeps confidence in the banks even when there
is uncertainty in the financial system. And so that function I
think does serve a valuable purpose. The mortgage market is
gigantic. And rather than moving more of the financial system
into the banks where we have more deposit insurance, this is
really another method of providing a liquidity guarantee to an
important financial sector. I believe you can have an economic
benefit without having significant cost to the Government.
Senator Reed. Thank you very much, gentlemen.
Senator Crapo.
Senator Crapo. Thank you very much, Mr. Chairman.
Mr. Davidson, to follow up on that, in your testimony, you
indicate that the liquidity of the TBA market combined with the
Government guarantee on the MBS serves to lower the rate on
agency MBS by about 25 to 50 basis points in relationship to
the non-agency alternatives during normal markets. Are you able
to isolate and estimate the subsidy of the Government guarantee
without the value of the liquidity in the market?
Mr. Davidson. Yes. So right now, the two pieces are
combined. But I would say that most of that benefit is
liquidity guarantee rather than a credit guarantee. In normal
markets, the actual credit risk on what should have been
guaranteed by Fannie and Freddie, just the high-quality
mortgages, is very small. It is on the order of, you know, five
basis points a year, so that most of the advantage that we are
seeing for the GSE loans is due to the liquidity guarantee, not
due to the credit aspect.
Senator Crapo. Thank you.
And Mr. Van Valkenburg, could you explain in a little more
detail how you feel that a mix of private mortgage insurance
companies and credit default swaps and other activities could
take on some of the credit risk that the Government guarantees
currently provide.
Mr. Van Valkenburg. Well, right now, the Government has
absorbed a lot of risk, credit risk, as opposed to the mortgage
market. The credit default swap market has parties who want to
take that risk, and the mechanics of it would be involving
possibly a future entity, this new issuer who now lays off
their risk through credit default swaps through this
intermediation process of that market. And so they can go out
and get price discovery on what the wholesale price is,
effectively, on where the market is pricing this credit risk
and where--if there is no bid for the 30-year guarantee in the
private sector, then the Government could decide, well, this
is--we are going to subsidize that because we think it has
important public policy goals, so we could elect to still have
a 30-year guarantee. But at least we would know what it would
cost in a private transaction.
So what I am envisioning is that we could develop a system
with, you know, obviously with SIFMA's leadership in developing
the market practices, but right now, the GSEs cannot really lay
off their credit risk. But we could develop structures and
markets where that credit could be traded.
Senator Crapo. For a private residential credit market to
develop, it has got to be able to attract capital, and right
now, it seems to me that is really difficult, given the
overwhelming market share that Fannie, Freddie, and the FHA
have. And for any of you, Mr. Van Valkenburg particularly, what
are some of the interim steps that can be taken to transfer
some of that credit risk, to move it away from Fannie and
Freddie? Or, Mr. Davidson, did you want to----
Mr. Davidson. Yes, I would be happy to.
Mr. Van Valkenburg. Go ahead.
Mr. Davidson. So I think right now, even within the
existing GSE structure, there is no reason why the GSEs cannot
work to start set up in these types of markets. There is no
reason why they cannot be using more external private capital
in the form of insurance, either pool insurance or mortgage
insurance. There is no reason they cannot set up a credit
default swap market, or the solution I said, which is they set
up a subordinate bond market where they sell off some of this
credit risk. So as a transition, we can start building private
capital markets even within the current GSE structure.
Senator Crapo. Thank you.
Mr. Hamilton, you wanted to comment?
Mr. Hamilton. Yes. I would just point out that some of this
technology already exists at the U.S. agencies. You know,
Freddie Mac for their multi-family lending program already
sells off the subordinate and credit tranches of the
securitizations that they create. So they are already doing it
in other markets, the fact that there is definitely the ability
to transfer some of that technology and structure into the
residential market.
One of the things the agencies provided to us is
homogeneity, information, transparency, and we have spent 40
years and the agency spent 40 years building that up. There is
no reason we should not use that same information and that same
homogeneity to create a market to disperse the credit risk on
the other side and reduce the burden on the taxpayer.
Mr. Van Valkenburg. And as the other witnesses point out,
these would be ancillary new markets as opposed to disrupting
the TBA markets. They could be new ways to trade credit that
would not disrupt the current TBA process.
Senator Crapo. All right. Thank you. I have no further
questions, Mr. Chairman.
Senator Reed. Senator Corker, please.
Senator Corker. Again, thank you for having this hearing
and I thank all of you for being here. I know everybody has
kind of checked out after what has just happened and there are
not a lot of folks here, but people certainly care about this
issue.
If you look at the TBA market, it is really nothing
different than a futures market that exists, right. So why is
it that we have to have this Government guarantee there to make
that work when it works so well with corn and coffee and
everything else? If you all could just briefly tell me why that
would not work.
Mr. Hamilton. I do not think any of us would say that it
could not work. I think what I would say is you could create a
TBA market out of private mortgages with an industry, maybe
SIFMA or someone else, who is creating a homogeneous market,
qualified residential mortgages that exist. But the important
thing to note is it will be at a significantly higher rate to
the homeowner. It will impact the U.S. housing market in a
significant way. And it will take 10 to 15 years to produce a
liquidity that we just spent 40 years producing.
Senator Corker. But it would be fairly priced.
Mr. Hamilton. It would be--one would hope----
Senator Corker. It would be fairly priced.
Mr. Hamilton. Well, I do not----
Senator Corker. Why would it not be fairly priced?
Mr. Hamilton. I think the market has proven, the agencies
and the private market have proven that doing market risk
pricing across the credit spectrum has not been perfect, and if
the private market was to take over credit pricing, I would
argue that people with good credit would get very good rates
and people with mediocre and bad credit would not be able to
get a mortgage. Maybe they should be renters. Maybe they should
be, but that is certainly how----
Senator Corker. I thought that was the way credit was
supposed to work, but go ahead. If you had bad credit, you
cannot get a loan. If you have good credit, you do.
Mr. Hamilton. I guess----
Senator Corker. We have gotten away from that.
Mr. Hamilton. I agree with you, but I think----
Senator Corker. Well, wait a minute----
Mr. Hamilton.----if Washington is willing to depoliticize
housing, then what you are saying is correct. If there is no
interest in home ownership rates or----
Senator Corker. Well, wait----
Mr. Hamilton.----having credit available to homeowners,
then I think what you are saying is absolutely true and only
people who have very good credit and can bring 20 percent down
to the table will ever get a mortgage. I think that is right. I
am not making a judgment on who should have them, but I think
that is the result.
Senator Corker. I think what you just said is a very
telling thing and I appreciate it.
Paul.
Mr. Van Valkenburg. I think that the guarantee subsidizes
that spectrum of the borrower, and so it enables the 30-year
fixed rate--we do not know if there is a market for a 30-year
guarantee right now because it does not exist. So could we
develop one? It is going to take, as Tom says, 5, 10 years to
develop. But so we could probably go out and price it today,
but there would be no liquidity in the credit. So we would be
fairly priced? Probably not. But we are assuming that the
Government guarantee is cheaper than what a fictional market
that we do not have any real price discovery on.
So the rationale, historically, to my understanding, is
that you subsidize the borrower with this 30-year fixed rate.
You are now instilling more benefits public policy-wise because
now you have more homeowners--well, it achieves more home
ownership or they can buy a larger house, whatever. But I do
not know the studies and the reality of the analysis around
that, but this is the message that has been sacrosanct for
years, that we subsidize this 30-year mortgage. We create more
home ownership and better communities because we have more
bigger houses and more involvement in the community.
Senator Corker. Mr. Davidson.
Mr. Davidson. So while the mortgage TBA market is
essentially like a futures market, it does have some important
differences. One of them is that, at least under current SEC
rules, you cannot sell a private security before it is created,
where in the TBA market, you can sell loans that will go into a
security that is not yet created. So at least technically, some
rules would need to change in order to create a TBA market
which would allow physical delivery of loans.
In most futures markets, people actually do not deliver
their product. They do not deliver the corn to the exchange or
to the counterparty. They usually pair those trades off. In the
TBA market, most of the originators actually sell the loans
into the short positions that they have created.
In addition, what we found in financial markets is that
financial markets that have Government guarantees behind them,
so sovereign markets, trade with much greater liquidity than
the non-sovereign markets. So I agree that futures markets
could work. Many of our clients who hedge mortgages also use
futures markets, either Treasury futures or Euro-dollar
futures, to hedge their positions. But the TBA market has
proven to be the most liquid and most effective hedge for the
mortgages.
Senator Corker. I guess in September, the loan limit will
go back to the pre-crisis levels, at least currently, that is
what is going to happen. Will that give us an opportunity to
see if this is something that will work?
Mr. Hamilton. Yes. I think the first----
Senator Corker. And what is the market expecting right now
that, going back to your allusion to Washington and
depoliticizing, what do they think Washington is going to do
right now as it relates to that?
Mr. Hamilton. Well, I think it is mandated that the loan
limit goes down to $625,000 at this point. I think the market
is expecting that to happen. I think the market would hope that
it continues to drift down in a meaningful way. I do not think
we are going to be able to create a private label mortgage
market without product. We cannot compete against the----
Senator Corker. So that is going to be--and will there be
any TBA activity in those upper levels?
Mr. Hamilton. My belief is, no, it will all go into private
label mortgage securitizations, or banks will just keep those
loans on their balance sheet.
Senator Corker. Mm-hmm.
Mr. Van Valkenburg. My point on this is that you are being
asked, or somebody is being asked, hey, what about--what are
the consequences? What are the tradeoffs to lower this lending
limit down? What is the price and what are the benefits? And we
do not have any real information to make the decision. We do
not have any real price discovery as to what that cost as far
as--what would willing parties in a private market trade that
risk for, and then we could evaluate what that cost would be as
far as the incremental guarantee fee, exposure to our balance
sheet, if you will, for the U.S. taxpayer by reducing that loan
limit down. And then we could evaluate it with real risk
measurements so that we could assess what is the effect, what
is the cost, and we do not have any other price discovery
mechanisms to do that today because it is all--well, it is this
amorphous blob of a Government guarantee.
So I am suggesting that we can create private markets that
trade and price credit. This price information will help the
economy allocate capital more effectively and price this risk
and help us make better public policy as part of the process.
Mr. Davidson. Yes. I believe it is a good step to lower the
loan limits. The private market used to support a trillion to
two trillion dollars worth of our mortgage loans. There is no
reason it cannot support a substantially larger portion than it
is right now.
Senator Corker. Mr. Chairman, again, thank you. I thank all
of you. And, Mr. Hamilton, I was in no way critical. I
appreciate the observation that we in Washington have created a
mechanism where those people with good credit pay more for
their mortgages and those people with bad credit pay less for
their mortgages. The question is, will we ever depoliticize and
cause the market to work in a normal way, where people who have
bad credit pay more for their rates and people with good credit
pay less. So I appreciate you bringing that out so clearly and
thank all of you for your testimony and look forward to seeing
you individually in our offices over the course of time. Thank
you.
Senator Reed. Well, thank you very much, Senator Corker.
I have got a few additional questions, but I think the line
of questioning both Senator Crapo and Senator Corker have
developed has been excellent. It exposes the real policy
choices we have. We have collectively for generations, both
Republicans and Democrats and everyone else has said putting
people in homes is key to America and it has resonated. It has
resonated, and as a result, a lot of these programs were begun
to do that.
And now we are at the point where we are looking at, how do
we make a transition and how do we do it in an effective way?
How do we price it correctly? How do we maintain a market? And
this hearing, despite the esoteric title, is absolutely
essential to what we are going to do going forward for all the
reasons that Senator Corker and Senator Crapo indicated. Can
the private market step in? Will they step in?
So let me just add a few other questions, but I will also
join Senator Corker in inviting you to come by the office,
because we are not going to settle this this morning. It is
going to be a dialog going forward.
But beginning with Mr. Davidson, and a question is that--
and again, if my understanding is not correct, please correct
it--but the TBA market is not subject to the securities laws.
It is exempt because of the participation of the GSEs, the
agencies. If we move to a private market where these activities
have to be performed by exclusively private firms, would that
induce SEC registration requirements and other forums? And if
so, would that complicate things further, or how should we
think about that? Maybe that is a more general question.
Mr. Davidson. Currently, loans that are delivered into the
TBA market are exempt from SEC registration, so that allows you
to essentially sell them before the securities are created.
Senator Reed. Right.
Mr. Davidson. In the private market, an issuer cannot do
that. They cannot say, I am planning on securitizing next
month. Please buy this security in advance. And that just
eliminates the possibility for physical delivery into a forward
sale. Other mechanisms could be developed, but they are just
not going to be as efficient, or maybe there could be some
exception created for some type of mortgage issues. But SEC is
going in the other direction right now and sort of requiring
more disclosure, more detailed disclosure about loans and
longer time periods between when you announce a deal and when
you sell it. The SEC's direction is contrary to what you need
to create a TBA-type market for non-agencies.
Senator Reed. So one of the aspects of creating this new--
and we all understand this is not going to happen next year,
this is a phased adaptation going forward--would be to provide
investor protections, but perhaps not in the same way that is
done presently by the SEC and private labels, but to work both
of those features in, the rapidity, the liquidity, the ease,
the uniformity with the investor protections. So we would have
to deal with that, I guess.
Mr. Davidson. That is correct.
Senator Reed. There is a--one of the proposals, and this
was, again, Mr. Van Valkenburg proposed that we ensure
liquidity in market--a single issuer with uniform capital
markets, accepted practices necessary, essentially separating
the guarantee function from the insurance function--the
issuance function, rather. And, Mr. Hamilton, can you comment
on that approach, of separating the guarantee function from the
issuing function.
Mr. Hamilton. Are you saying separate the Government
guaranteeing it or someone private guaranteeing it?
Senator Reed. I was----
Mr. Van Valkenburg. Just the function itself. Do we need to
have multiple issuers. I guess you are reading from my
testimony.
Senator Reed. Yes, I am. Go ahead.
Mr. Van Valkenburg. I was just suggesting the issuance
should be a single issuer so that investors are not confused if
it is ABC issuer, XYZ issuer. It is always done by the same
issuance. And then the guarantee function could be priced
independently, or the role can be determined independently.
Senator Reed. And I will not put words in your mouth, but
if it is a single issuer, that likely could be a Government
entity.
Mr. Van Valkenburg. Yes, it could be.
Mr. Hamilton. I think these are all--they are all saying
the same thing, I think, that separating out the guarantee
function is what we are talking about----
Senator Reed. Right.
Mr. Hamilton. When we discuss selling--when Mr. Davidson
says selling subordinate tranches off of agencies or selling
credit risk in a CDS or credit default swap off of a certain
pool of residential mortgages, it is all accomplishing the same
thing. We are trying to separate the guarantee function and
mitigate and sell that credit risk so the taxpayer is not on
the hook, yet at the same time maintaining one issuer or a
couple of issuers and maintaining that liquidity. So I think it
is all saying the same thing in a different way.
Senator Reed. Mr. Davidson, it is your thoughts. Do you
want to elaborate?
Mr. Van Valkenburg. He is Mr. Davidson.
Senator Reed. Oh, I am sorry----
[Laughter.]
Mr. Van Valkenburg. I am not as good looking as he is.
[Laughter.]
Senator Reed. Mr. Van Valkenburg.
Mr. Van Valkenburg. Thank you. Yes. I was basically--in the
current TBA market, there is a perception of if Fannie Mae has
a better liquidity than Freddie Mac, it is minor, two or three
ticks. I do not know, Tom knows better than I. So an issuer
name is significant as far as liquidity, so that was basically
it. I think if we get into too many issuers, it is going to
confuse the investors, which is going to affect the level of
liquidity. So a single name issuer is what I was advocating and
the guarantee function separated. I think it costs us in
liquidity if we have multiple issuers.
Senator Reed. Right. But, I mean, in just thinking back, I
think that is a way we sort of walked ourselves into the GSEs,
which is basically if you are going to give a monopoly, then it
has got to be quasi-governmental, at least. Otherwise, the
monopoly profits go to someone. That is not our tradition. So
this approach, I think, would imply the issuer would be some
type of entity, either very closely regulated by the Government
or some quasi-governmental entity, and then the guarantees
would be subject to market pricing and market activity.
Mr. Van Valkenburg. Exactly, and we are not going to get a
credit market established quickly, but we could develop that
over time. The current infrastructure with a 30-year guarantee
is not going to go away tomorrow and it is going to take time
for that to develop, but----
Senator Reed. Mr. Davidson, your comments. You are Mr.
Davidson----
Mr. Davidson. Yes, I am Mr. Davidson.
[Laughter.]
Senator Reed. Forgive me. It is vision as well as----
[Laughter.]
Mr. Davidson. One idea is that if you are going to have
this single issuer or a few issuers to avoid this monopoly
situation is that you can allow that to exist only as a
cooperative, so some sort of industry cooperative like the
DTCC, and so that if there is monopoly profits, they have to go
back into the chain where it is competitive, either above or
below that cooperative.
But I do think that having one or two issuers is good.
Trying to get as many participants to take the credit risk is
good. But you do have to find a way of standardizing the
mortgage products. So let us say we had 20 different mortgage
originators, all of who would go through this one guarantor,
but they all had 20 different, like, loan documents, 20
different types of disclosure. Then we would have 20 different
markets and that is not going to promote the liquidity we need.
So it is just a careful balancing act between spreading the
risks and standardizing.
Senator Reed. Let me ask this as a final question to all of
the panelists, but I will begin with you, Mr. Davidson, which
is you have proposed this subordinated debt approach. But I
think it implicates a bigger issue, which is recognizing that
we should probably begin to take steps now to begin a
transition, that legislative steps, because it takes a long
time to get legislation through and because of the potential it
is not likely going to happen today or even next year, et
cetera. But as you suggest, and I think everyone on the panel
has suggested, there are things today that should be considered
to begin this process, maybe even experiments which do not work
out and save us the trouble of trying them in the future on a
larger scale, or adopting them as an exclusive remedy.
So if you can comment on your subordinated approach, how it
might work and is it feasible to begin to adopt it today, and
then I would like to ask the other panelists to think of either
what other steps that you would suggest, again, outside of
legislative but within the purview of the agencies today and,
as you understand, the legal framework.
So let us begin with Mr. Davidson.
Mr. Davidson. So the subordinated approach is something
that the GSEs have done in the past----
Senator Reed. Mr. Hamilton pointed that out.
Mr. Davidson. Right, sort of in the multi-family area----
Senator Reed. Yes.
Mr. Davidson.----and I believe it is something that the
GSEs are exploring currently, different ways of adding private
capital. And so it is certainly doable with the existing
structure, but they probably do need the approval of FHFA. I do
not know if they need the approval of Treasury. I think they
can move in that direction. I think they should try several
things. So I like subordinated bond approach. Other people
might think we should use more private mortgage insurance.
I think the key factor there is sending the message to FHFA
that experimentation is good, you know, that is what you want
to see, and you do not need to have sort of a sole focus on
conservatorship of every dollar today, and I think finding the
right solution actually adds value to the GSEs over time.
And the other important component is to try to move the
dialog away from let us destroy the--you know, eliminate the
GSEs tomorrow because they were bad before. The managements who
did that are all gone. So are there pieces of the GSEs that we
would like to preserve over time. So, anyway, I think it is
doable.
Senator Reed. Thank you.
Mr. Van Valkenburg, please.
Mr. Van Valkenburg. Yes. I proposed one idea, but there is
no single monolithic solution, and all these markets are
complex and they price risk differently and there are a
different audience of investors. So the subordinated bond
solution may be the best solution. We do not know until we
actually go out and try to execute and see what the price and
cost of that credit is. So I proposed the CDS market just as
one avenue as exploratory thought. So I do not have a
particular single solution, but I think you have to price it,
find out what the costs are, and find out what is the best
execution for the Government's balance sheet.
Senator Reed. Thank you.
And Mr. Hamilton.
Mr. Hamilton. I think we end up with a portfolio of things
we need to work on in the interim, given the likelihood of
legislation in the near term is low. The FHFA can obviously
play a large part in this. Lowering of the loan limits is one
step. We could lower them further on a gradual basis over the
next 18 months would be the next step. It would enable the
private market to open up. I think the FHFA could--you could
limit the amount of borrowings that banks can do from the Home
Loan Bank System. You could encourage the covered bond
legislation and market to open up to be another funding
vehicle.
I think there is a portfolio of approaches that are going
to attack the U.S. housing system and be the solution for
mortgage finance, and I think there are quite a few of these we
can do in the next 18 months without legislation, and I think
those are just a few of the things we should work toward, and
then I think we will find the answer. We will rise to the top
quickly.
Senator Reed. Well, thank you very much, gentlemen. This
has been very, very helpful to the Subcommittee. As Senator
Corker suggested, please do not be surprised if you are called
again to get your views and your advice because it is
extraordinarily helpful, and thank you very much for your
testimony and your appearance today.
Some of my colleagues might have additional questions. We
will ask that these questions be submitted before the end of
the week. It is Wednesday, and so by Friday, I think that is
fair. We will get those to you and ask you to return them as
promptly as possible if there are any written questions.
Again, thank you very much. The hearing is adjourned.
[Whereupon, at 10:32 a.m., the hearing was adjourned.]
[Prepared statements supplied for the record follow:]

PREPARED STATEMENT OF THOMAS HAMILTON
Managing Director, Barclays Capital, on behalf of the Securities
Industry and Financial Markets Association
August 3, 2011

Good morning Chairman Reed, Ranking Member Crapo, and Members of
the Subcommittee. I am Thomas Hamilton, Managing Director, at Barclays
Capital, where I am responsible for securitized products. I am pleased
to testify today on behalf of the Securities Industry and Financial
Markets Association (``SIFMA'').\1\
---------------------------------------------------------------------------
\1\ The Securities Industry and Financial Markets Association
(SIFMA) brings together the shared interests of hundreds of securities
firms, banks and asset managers. SIFMA's mission is to support a strong
financial industry, investor opportunity, capital formation, job
creation and economic growth, while building trust and confidence in
the financial markets. SIFMA, with offices in New York and Washington,
D.C., is the U.S. regional member of the Global Financial Markets
Association (GFMA). For more information, visit www.sifma.org.
---------------------------------------------------------------------------
We appreciate the opportunity to discuss the most liquid secondary
markets for mortgage loans and mortgage-backed securities (MBS) in the
world--the ``To-Be-Announced'' (TBA) markets. These are critically
important markets which help consumers purchase homes, and we would
like to discuss in detail how they work, the benefits they confer on
consumers and the economy, and their important role in mortgage
finance.

Summary
Housing is a critical component of our economy, and is the center
of a virtuous circle: housing begets jobs, which beget housing.
Consequently, the U.S. mortgage market is enormous; for example the
home mortgage market is approximately equal in size to the total size
of U.S. bank balance sheets. Given that banks engage in activities
other than residential mortgage lending, their balance sheets alone
cannot meet the country's need for mortgage credit. Ginnie Mae, Fannie
Mae, Freddie Mac (together, ``the Agencies'') and private institutions
use the process of securitization to provide capital that allows for
the growth of mortgage lending beyond the capacity of bank balance
sheets. Today, private securitization and the Agencies finance nearly
70 percent of outstanding home mortgages and it is therefore imperative
that securitization continue to play a key role in any future mortgage
finance system. The market for MBS issued by the Agencies is
approximately three times the size of the outstanding non-agency MBS
market. In this Agency MBS market, the TBA market is the single largest
component. The TBA market is currently the key to funding mortgage
lending, and because of this it plays a critical role in housing and
the U.S. economy.
The enduring liquidity of the TBA market, in contrast to the lack
of issuance in the private MBS markets, preserved the availability of
mortgage credit in the recent crisis. This ability to maintain
liquidity during stress periods is a key benefit of the TBA markets.
Furthermore, the liquidity and resilience of the TBA market attracts a
wide range of investors who provide vast capital that is cycled into
mortgage lending, including retirement savings vehicles, insurance
companies, and foreign investors.
The vast liquidity and forward-trading nature of the TBA market
provides key benefits to consumers, such as the broad availability of
30-year fixed rate mortgages that may be prepaid without penalty, and
significant and consistent liquidity in the secondary mortgage market.
This results in a stable and attractive funding source for lenders that
allows them to provide lower mortgage rates and longer-term ``rate
locks'' for borrowers, and efficiently recycles funds back to local
lenders to enable another round of mortgage lending.
Of course, the TBA market is facilitated by the guarantees of the
Agencies, and therefore the support of the Government that stands
behind them. Currently over 90 percent of mortgages are financed
through a program of one of these three Agencies. This level of support
is possible because Agency MBS do not expose investors to credit risk,
and therefore the market is attractive to risk-averse investors that
have vast sums of capital available for investment. Without the TBA
market, we believe that the majority of this investment capital would
be directed elsewhere, reducing the amount of funding for and raising
the cost of mortgage lending. Therefore, SIFMA believes strongly that
maintaining a liquid and viable TBA market should be considered as
Congress addresses housing finance reform.
With that said, the reality is that that this current outsized
role of the Government is not sustainable over the long term, and
should be reduced. The TBA market's role, and the Government's role,
should shrink as the private markets regenerate over time. The means of
achieving this rebalancing are very complicated and consequential on a
national, financial, and personal level. While SIFMA believes the TBA
market should play a role in the future, it should certainly not be 90
percent of the market. There are a number of challenges to the
resurrection of non-agency securitization, including the significant
uncertainty faced by non-agency MBS investors and issuers. The rules of
the road, for both sides, are not clear. Until they are, it will be
challenging for issuers and investors to see eye to eye on
securitization transactions, at least in the volume and frequency that
is necessary to fund mortgage credit demand. Being able to withdraw the
Government from mortgage markets will require a carefully planned and
sequenced transition which should take a number of years. It is
essential to remember that the necessary volume of non-agency investors
will not simply appear because we want them to. They must be drawn back
into, and made comfortable with, private label securitization and its
regulatory environment.
We believe that it is critical that the planning and execution of
significant changes to the funding of mortgage loans be done with
attention to detail, be based on sound analysis of costs and benefits,
be mindful of unintended consequences, and create a long-term
beneficial and stable environment. While we cannot predict the future,
we can use the past as a guide and apply lessons learned and mistakes
made, the good and the bad, to a design that will stand the test of
time.
We hope that the testimony we present today will he helpful in
educating policymakers about how mortgage loans are funded in the
capital markets, the critical role of the TBA market, and some of the
critical issues that must be considered to move forward.
Thank you for this opportunity and I am happy to take any
questions.

Discussion

A. Terminology
We will first quickly review basic terminology to set the stage for
the rest of our testimony.

Mortgage-Backed Security (MBS)_An MBS is a type of bond
collateralized by mortgage loans that represents an undivided
fractional interest in that pool of loans. Beneficial ownership
of this interest may be transferred in trading markets.
Payments to bondholders result from the underlying payments and
cash-flows on the mortgage loans that serve as collateral. Cash
flows to MBS investors are variable, as most mortgage loans are
prepayable without penalty.

Agency MBS_Agency MBS are collateralized by loans meeting
Fannie Mae (FNMA), Freddie Mac (FHLMC), or Federal Housing
Administration (FHA) underwriting guidelines, and are issued
and/or guaranteed by Fannie Mae, Freddie Mac, and Ginnie Mae
(GNMA). Agency MBS are perceived to have little to no credit
risk because they carry either an explicit Government guarantee
(GNMA) or an implicit guarantee (FNMA and FHLMC). Unlike Fannie
Mae and Freddie Mae, Ginnie Mae does not issue debt or
mortgage-backed securities. It is a guarantor of privately
issued securities collateralized by loans insured by the FHA,
Veterans Administration, and the Rural Housing Service.

Non-Agency MBS_So-called non-agency MBS are collateralized
by a wider variety of loan types than Agency MBS, and are
issued by private lenders, and are not guaranteed by Fannie
Mae, Freddie Mac, or Ginnie Mae. Non-agency MBS are generally
structured into tranches with varying degrees of repayment
priority, and therefore introduce varying degrees of credit
risk to investors. Credit risk is the risk of losses if
borrowers do not repay their loans. Recently, there have been
two notable non-agency MBS transactions been backed by
extremely high quality, high-balance loans (a.k.a. ``Jumbo
Prime'' loans); prior to 2008, non-agency MBS also included
``subprime'' and ``Alt-A'' loans.

Common MBS Structures

1. Pass Through--A pass through security is the simplest form of
MBS. Payments on the loans are delivered to investors as they
are paid by borrowers (i.e., they are ``passed through''). Most
Agency MBS are issued in pass through form. MBS eligible for
TBA trading are in the form of pass-throughs.

2. Collateralized Mortgage Obligations (CMO) and Residential
Mortgage-Backed Securities (RMBS)--CMOs and RMBS structure cash
flows to investors by dividing borrower payments into various
``tranches'', or slices that are entitled to particular streams
of payments. Agency securitizations are generally called CMOs,
and Non-Agency securitizations are usually called RMBS.

3. Real Estate Mortgage Investment Conduits (REMIC)--In 1986
amendments to the tax code created favorable treatment for
mortgage securitization structures that met certain
requirements. These rules are administered by the Internal
Revenue Service. Most MBS are issued in compliance with REMIC
regulations.

To-Be-Announced (TBA) Trading of MBS--To-Be-Announced
trading is a trading convention whereby homogeneous MBS are
traded for forward settlement and the purchasing party does not
know the specific identity of the MBS pool to be delivered.
Trades are executed based on a limited number of criteria,
including issuer, coupon, term of mortgage collateral, and
settlement date.

B. Overview of the U.S. Mortgage Market, and the Importance of
Securitization
Chart 1 below shows the enormous size of the U.S. mortgage market
relative to bank balance sheets. The size of the mortgage market has
previously exceeded, and is currently nearly equal to the total size of
bank balance sheets. This chart demonstrates that there is not enough
capacity in the U.S. bank balance sheets to fund our nation's housing
stock alone. Through securitization we are able to recycle capital
available for lending and attract vast sums of new capital to the
markets.

To put this in a global context, the U.S. mortgage market is larger
than the combined mortgage markets of all of the countries in Europe.
For this reason, the U.S. mortgage market is not directly comparable to
any single other market in the world.

Three quarters of U.S. mortgage debt is residential mortgage debt,
as shown below.

Securitization and the MBS markets play a critical role in funding
this residential mortgage lending. We have shown above that the
mortgage market is enormous, that it is primarily a residential market,
and that securitization is necessary to fund this level of credit
creation. We will now turn more specifically to the role of
securitization and secondary markets in funding these markets, and
discuss who ultimately provides this capital.
Below is a chart that outlines how mortgages are funded in the
United States. 67 percent, or $7.1 trillion, of home mortgages are held
in a GSE portfolio or securitized (agency and non-agency). Secondary
markets, therefore, are responsible for funding two thirds of
residential mortgage lending. The securitized home mortgage market can
be split between agency MBS and non-agency MBS, with 81 percent of all
MBS in the form of an agency pass-through or agency CMO.

To put the size of the MBS markets in perspective, the chart below
places them in the context of the other fixed-income markets. They are
larger than all markets but for Treasuries.

Another important issue to understand is who holds these
securities. Banks, pension funds, mutual funds, and insurance companies
are key investors in MBS. Foreign sources of capital, including
investment companies, sovereign wealth funds, and other Government
entities are also critical sources of capital for U.S. mortgage
markets. Below are two charts which outline the holders of both agency
MBS and non-agency MBS.

The most critical point of this testimony to this point is this: in
any consideration of the future of housing markets, the future of the
GSEs, or the future of mortgage lending, it is critical to remember
that these markets will not work without the participation of
investors. The U.S. mortgage market, as shown above, is huge. It is a
key component of the economy and job creation, and is largely funded by
many different kinds of investors. Therefore, any housing reform or
changes to the current regime must be viewed through the lens of
investor needs, and what investors are willing to pay for a given
investment opportunity.

C. The Role of the Agencies
A. What They Do
The Agencies have long played a crucial role in the U.S. mortgage
finance market. Fannie Mae and Freddie Mac purchase loans, securitize
them, and guarantee the timely receipt of principal and interest on
their MBS. Ginnie Mae securitizes Government-insured FHA, USDA, and
other Government guaranteed loans, and places a similar guarantee of
timely payment of principal and interest on the mortgage-backed
securities. For the last 30 years, the Agencies have played a critical
role in mortgage finance, utilizing securitization to expand the supply
of capital available for mortgage lending.
Standardization has been a key benefit of the Agency model. Due to
their size and the scale of their operations, the Agencies have driven
standardization of mortgage loan documentation, underwriting,
servicing, and other items in ways that have created a more efficient
origination process. This standardization extends beyond the Agency
market, and has driven standardization of lending processes more
generally, across product types, markets, and across institutions.
Perhaps more importantly, the activities of the Agencies have
driven the standardization of loan maturities out to 30 years, creating
a mortgage product that is affordable to a greater proportion of
consumers. Most people take for granted that typical mortgage loans
have a 30-year term, but given the nature of bank funding, this is not
a natural outcome. Before the implementation of Government programs
such as the Homeowners Loan Corporation, FHA, and Fannie Mae in the
1930s, mortgages tended to be short term and require a balloon payment
at the end of the term. This was directly related to the short-term
nature of bank funding. Many institutions derive a majority of funding
for lending from customer deposits which are redeemable upon demand.
The development of secondary markets for loans and MBS through
Government initiatives allowed banks to extend loans with longer terms.
Banks were able to access a longer-term funding source to match the
terms of the mortgages, transfer risk, reduce balance sheet
utilization, and reduce demands upon limited capital through loan sales
into active secondary markets and ultimately securitization. Without
the initiatives undertaken by the Government in the 1930s and the
continuing support of the GSEs, it is not clear that today's ``normal''
mortgage loan would have a 30-year term. In a world without Government
guarantees, the 30-year mortgage would likely still exist, but with
lesser availability and presumably higher cost, due in part to issues
related to risk hedging.

B. Agency Market Share Trends and Performance During the Crisis
The chart below shows the ratio of agency MBS issuance to non-
agency MBS issuance over the last 30 years. This chart clearly shows
the reaction of the agency and non-agency markets to the financial
crisis. Throughout the 80s and into the 1990s, the Agency share of the
MBS markets was in the range of 80 percent. As the non-agency markets
expanded in the mid-2000s, during the housing boom, the Agency share
fell to approximately 50 percent. Therefore, even at the peak of the
housing and securitization boom, the Agencies remained a critical
participant in the MBS markets. As the non-Agency MBS markets collapsed
in 2007 through the present, the Agencies took on a more critical role
than ever, in terms of providing funding for mortgage lending to
consumers. The Agency market was a stable source of funding throughout
the crisis.

Another issue to review is the cost of a conforming loan (a loan
eligible for securitization by a Agency) versus a non-conforming loan.
The chart below compares the spread between conforming loan rates and
non-conforming loans with balances that exceed the conforming loan
limit. During the financial crisis of 2008, the spread between
conforming and non-conforming mortgage rates increased to approximately
five times its historic level, and pricing on non-Agency MBS relative
to Agency dropped precipitously. The spread between these rates spread
has yet to return to its historic trend.

From these charts, you can clearly see that we need to reduce the
share of lending funded through the Agencies. Over the long run, it is
not healthy for the Government, in one way or another, to support 95
percent of mortgage lending. SIFMA therefore agrees that housing
finance reform is critical, and supports its careful implementation.
At the same time, we believe that it is important to keep in mind
that the Agencies have conferred significant benefits on U.S. mortgage
markets. We believe that housing finance can and should be reformed and
made more robust without destroying the benefits that the Agencies have
conferred. We caution that the drive for reform should not cause
collateral damage that would eliminate or make impossible the
beneficial impacts and legacy of the old system that developed around
the Agencies.
One of the most important benefits of the system developed over the
previous decades, if not the most important, was the development of a
liquid forward market for mortgage-backed securities known as the TBA
market. The TBA market allows lenders to hedge risk, attracts massive
amounts of private capital, and reduces the cost of mortgage lending.
SIFMA believes the TBA market should be a key component of a
successful, liquid, affordable, and national mortgage market, as well
as ensuring a sufficient level of capital is available to banks to
lend. The historically huge and liquid global markets described above
for Agency MBS are initiated by the TBA mechanism.

D. The TBA Markets
A. History
The genesis of the TBA market began in the 1970s, when members of
the Government Securities Dealers Association began to discuss
standards for the trading and settlement of bonds issued by Ginnie Mae.
In 1981, the Public Securities Association \2\ published the ``Uniform
Practices for the Clearance and Settlement of Mortgage-Backed
Securities and Other Related Securities'', which is a manual that
contains numerous of market practices, standards, and generally
accepted calculation methodologies developed through consensus
discussions of market participants, that are widely accepted and used
in the MBS and asset-backed security markets. The GSDA and PSA were
predecessors of SIFMA.
---------------------------------------------------------------------------
\2\ The Government Securities Dealers Association and the Public
Securities Association are predecessor organizations of SIFMA.
---------------------------------------------------------------------------
Participants in the TBA market generally adhere to market-practice
standards commonly referred to as the ``Good-Delivery Guidelines'',
which comprise chapter eight of this manual.\3\ These guidelines cover
a number of areas surrounding the TBA trading of agency MBS, and are
promulgated by and maintained by SIFMA, through consultation with its
members. The purpose of the guidelines is to standardize various
settlement related issues to enhance and maintain the liquidity of the
TBA market. Many of the guidelines are operational in nature, dealing
with issues such as the number of bonds that may be delivered per one
million dollars of a trade, the allowable variance of the delivery
amount from the notional amount of the trade, and other similar
details.
---------------------------------------------------------------------------
\3\ The Good Delivery Guidelines are a part of SIFMA's Uniform
Practices for the Clearance and Settlement of Mortgage-Backed
Securities and Other Related Securities, which is available here:
http://www.sifma.org/research/bookstore.aspx.
---------------------------------------------------------------------------
B. Mechanics of a TBA Trade
The majority of trading volume in the agency MBS markets today is
in the form of TBA trading. For background, a TBA is a contract for the
purchase or sale of agency mortgage-backed securities to be delivered
at a future agreed-upon date; however, the actual pool identities or
the number of pools that will be delivered to fulfill the trade
obligation or terms of the contract are unknown at the time of the
trade. Actual mortgage pools guaranteed by one of the Agencies are
subsequently ``allocated'' to the TBA transactions to be delivered upon
settlement. Settlement dates of transactions are standardized by
product type (e.g., 30-year FNMA/Freddie Mac pools, 30-year Ginnie Mae
pools, 15-year pools) to occur on four specific days each month.
Monthly settlement date calendars for the TBA market are published 1
year in advance by a SIFMA committee on a rolling 12-month basis. This
is done to increase the efficiency of the settlement infrastructure,
and facilitate forward trading. Most trades are executed for settlement
within one to 3 months, although some trading may go further forward
from time to time.

For example, Investor A could call up Market Maker A on May 23, and
order $10 million FNMA 5.5 percent coupon 30-year MBS, for settlement
on July 14. The investor does not specify specific bonds or CUSIP
numbers. On July 12, according to market practice, Market Maker A would
notify Investor A of the specific identities of the pools that will be
delivered on July 14. Most likely, these will be MBS that were just
issued at the beginning of July.
On the other side of an investor or market maker often stands a
loan originator. Originators can enter into forward TBA sale contracts,
allowing them to hedge the risk of their loan origination pipelines.
This permits the lenders to lock in a price for the mortgages they are
in the process of originating, benefiting the borrower with the ability
to lock in mortgage rates earlier in the process. Pricing on loans
varies from day to day with fluctuations in the TBA markets, and
lenders will often re-price loans for their bankers and correspondent
partners on a daily basis. Thus mortgage bankers follow the market in
order to make decisions on when to lock in a rate for a borrower.

C . Key Benefits of the TBA Markets
1. Liquidity for U.S. Mortgage Lending
The TBA market is by far the most liquid, and consequently the most
important secondary market for mortgage loans. This liquidity is
derived from its vast size ($trillions), homogeneity of collateral, and
the forward nature of the trading. This liquidity is due to one factor:
homogeneity. TBA trading is based on the assumption that the specific
mortgage pools which will be delivered are fungible, and thus do not
need to be explicitly known at the time a trade is initiated. At a high
level, one pool is considered to be interchangeable with another pool.
The sources of this homogeneity are primarily threefold:

The Agencies each prescribe standard underwriting and
servicing guidelines (FHA plays this role in concert with
Ginnie Mae in those markets)

Standardized market practices and guidelines (the ``Good
Delivery Guidelines'', discussed more below) ensure that
securities eligible for the TBA market are homogeneous, which
allows buyers and sellers to transact with confidence that
knowing the specific identity of a security they will trade, at
the time of trade, is not necessary;

The explicit or implicit guarantee on the MBS eliminates
credit risk from the risk factors investors must deal with.
This guarantee also attracts classes of investors who would not
otherwise participate in these markets; investors who are
statutorily prohibited from, blocked by investment guidelines
from, or simply do not desire to take on mortgage credit risk.

Thus, investors can buy securities without knowing their exact
identity because they know that (1) the underwriting will be consistent
across pools, (2) the servicing will be consistent across pools, (3)
the MBS and operational mechanisms around their trading will be
consistent across pools, and (4) they do not need to perform a loan-
level dive to explore credit risk before they purchase the bonds.
There are currently over $4 trillion in bonds eligible for TBA
trading--it is a vast market. It is also extremely liquid. Federal
Reserve data shows average daily trading volumes of Agency MBS reported
by the Fed's primary dealers as exceeding $300 billion per day over
each of the last 3 years. Private estimates of daily TBA trading
volumes exceed $600 billion (these estimates take in to account trading
beyond that of the primary dealers). Liquidity in this market is second
only to the market for Treasuries. This liquidity allows investors to
buy and sell significant quantities of securities quickly and without
disrupting the market. This makes the market very attractive to these
investors who have substantial funds to be invested.

This liquidity draws trillions of dollars of investment capital to
U.S. mortgage markets, as discussed in detail in the previous section
of this testimony. Given the size and liquidity of the market, buyers
and sellers are able to trade large blocks of securities in a short
period of time without creating distortions.
2. Originator Hedging and Rate Locks
As mentioned, this market allows lenders to sell their loan
production on a forward basis, in some cases before MBS pools are
formed, and hedge risk inherent in mortgage lending. A benefit of this
ability to hedge risk is that the TBA market allows lenders to lock-in
rates for borrowers. Lenders can sell forward in the TBA market at the
then-current interest rate. Without TBA markets lenders would either
have to charge substantially more for (probably shorter-term) rate
locks, because hedging in derivatives or options markets is more
expensive and less efficient. It is possible that some lenders simply
would not offer rate locks at all. The liquidity of the TBA market
creates efficiencies and cost savings for lenders that are passed on to
borrowers in the form of lower rates and broad availability of mortgage
products, and helps to maintain a national mortgage market.

3. Benchmark Status of the TBA Market
For all of the reasons outlined above, the TBA market is a
benchmark for all mortgage markets--it is the reference by which other
mortgage markets and products are priced. In this manner it is similar
to the Treasury market. This is an issue that is often overlooked, but
one that we want to highlight. Non-agency mortgage product is priced
relative to TBA; TBA provides a sort of risk-free reference point for
those markets. Without the TBA market, we believe that non-TBA markets
would be somewhat more volatile as pricing would become more
challenging. We also note that predictions of the movement of mortgage
rates in a world without TBA generally do not take into account this
role. While the actual change in rates would be quite dependant on the
exact contours of a mortgage finance system without TBAs, we suspect
that the change may be greater than many currently believe.
It is difficult to exaggerate the consequences from a loss of
confidence or liquidity in this market if a suitable replacement were
not found. The effects would be directly and immediately felt by the
average mortgage borrower. The impact would include, at a minimum,
higher mortgage rates, as yields required by investors would rise as
liquidity falls. It is also likely that credit availability would be
constricted. This would occur because secondary market executions for
originators would be more expensive and take longer, requiring longer
warehousing periods for loans they originate. Balance sheet capacity is
a currently a scare commodity for most lenders, and is finite in any
case. Furthermore, the ability of borrowers to lock-in rates on
mortgage applications would likely be reduced, creating uncertainty for
them and likely depressing real estate activity which is an important
component of broader economic activity.

E. Looking Forward--Considerations for TBA Markets and the Future of
Mortgage Finance
There is no single ``right answer'' or any easy solution to the
question of how to resolve the conservatorships of Fannie Mae and
Freddie Mac and/or define the future infrastructure for mortgage
finance in the U.S. Policymakers are faced with a series of difficult
choices, each with its own costs and benefits, which will shape the
future of housing finance. Ultimately, this essential infrastructure is
both a creation of and a reaction to past public policy choices, and as
such the future of it will grow out of further determinations of what
is the appropriate public policy regarding mortgage finance. While
there are many important questions, we believe a special and near-term
focus needs to be placed on resolution of the current status of the
GSEs and the restoration of the private-label securitization markets
for mortgages.
Secondary mortgage markets will continue to function regardless of
what policymakers decide. As the saying goes, there is a price for
everything. This price, however, is not always desirable to everyone.
Accordingly, policymakers need to determine what they want from the
mortgage markets before they can address what to do with the GSEs or
the broader infrastructure of mortgage finance. Among the issues for
policymakers to consider are:

how liquid secondary markets for loans and MBS would be;

the breadth of products that would be offered to consumers;

the capacity of lenders to extend credit;

whether national lending markets could be sustained or if
regional pricing differentials would reappear;

the cost and affordability of mortgage credit to consumers.

SIFMA believes that the TBA markets are one of the keys to a
successful, liquid, affordable, and national mortgage market. TBA
markets also ensure that a sufficient level of capital is available to
banks to lend. We repeat our previous statement: the historically huge
and liquid global markets for Agency MBS are initiated by the TBA
mechanism.
1. Can the TBA Market Function without a Government Guarantee?
Ultimately, the answer to this question is unknown. We are not
aware of any meaningful, consistent TBA-style trading of any other non-
guaranteed mortgage product at this time. To the extent that guarantees
were completely removed, we believe that the best case outcome with
respect to TBAs is a much smaller, much less liquid market. The worst
case outcome would be the dissolution of the markets. But in the end,
we do not know at this time.
As we mentioned earlier, the key driver of the TBA market is
homogeneity. In the future, one can envision a recreation of ``Good
Delivery Guidelines'' for a non-guaranteed product. However, this is
only one piece of the puzzle. The Agencies play a critical role in the
TBA markets through their standardization of underwriting and
servicing, and their enforcement of that standardization through
automated underwriting systems and otherwise. It is unclear to SIFMA
how this could be recreated to the degree of detail at which it
currently exists, and be done so in a format that was efficient and
manageable enough to support liquid TBA markets.
The guarantee on MBS traded in TBA markets eliminates a key risk--
credit risk. Investors in TBA markets focus on prepayment risk, that
is, the risk that borrowers will repay their loans early, and on
interest rate and market risk, or the risk that interest rates or
market pricing will move against them. This allows what are called
``rates investors'' to invest in the Agency MBS markets. Rates
investors, put simply, are investors who do not wish to take on credit
risk. They include various investment funds, and importantly, many
foreign investors.
In the non-Agency markets, investors must also deal with credit
risk. This entails an examination of the credit risk factors of the
loans that collateralize the MBS. Going forward, we expect that
investors will perform this review at a loan level, as disclosure
practices and regulations for non-Agency MBS drive to this end. In and
of themselves, loan level reviews are not practical for TBA trading
(because one cannot review loan level detail on an unknown pool of
loans). Therefore, to create a level of comfort that would allow
investors and market makers to trade non-agency collateral on a TBA
basis, underwriting standards would need to be very strict because they
would need to eliminate as much credit risk as possible. As a result,
lenders would likely draw such a small circle around eligible mortgage
loans that the supply of loans would likely not be sufficient to
support large and liquid TBA trading. Additionally, to define the
underwriting standards for every bank that would deliver into this
market, and on top of that to outline servicing procedures, would
entail a massive expansion of market practice guidelines in terms of
breadth and length. This would complicate the ability of investors to
get comfortable that the loans that underlie the securities they will
be delivered next month, or the following month, will comply.
Importantly, there would be no clear enforcement mechanism for
compliance.
The expansion of the usage of mortgage insurance to provide comfort
to MBS has been put forth as one alternative. SIFMA's discussions with
its members have evidenced significant doubts that the investing
markets would take anything near the current level of comfort from
private mortgage insurance solutions. In any case, members generally
believe this solution would be inadequate to support liquid TBA
trading.
Given all of this, it is not clear what proportion of the current
rates investor base would shift into the proposed new non-guaranteed
TBA markets. If a significant proportion of the rates investor base did
not shift into the new market, the potential liquidity and potential
size of the new market would be severely compromised (if it functioned
at all). It is also not clear on the supply side whether or not a
sufficient quantity of loans would be produced that would comply with
the extremely strict underwriting guidelines that would be needed. It
is notable that no other mortgage market or funding system via
depositories has ever provided sustained liquidity to the extent that
the Agency MBS markets have. It is also notable that each secondary
mortgage market that was not the beneficiary of a guarantee collapsed
in 2008.
SIFMA's Housing Finance Reform Task Force has concluded that some
form of explicit Government support is needed to attract sufficient
investment capital to maintain liquidity and stability in the TBA
market at a level comparable to that created over the last 30 years.
Members believe that total privatization of mortgage finance will
likely result in greater volatility, decrease efficiency, and
ultimately make mortgage loans more expensive and less available. There
are a number of ways that an explicit guarantee on MBS could be
structured. The bottom line for a guarantee is that investors in TBA
markets must know that they will receive back at least their invested
principal. Without it, certain rates investors would completely drop
out of the market and others would have significantly smaller
allocations of investment capital available for the asset class, and we
expect that at best, the peak volume and liquidity of such a market
would be orders of magnitude smaller than the current TBA market.
Furthermore, as discussed above, Agency MBS currently provide a
safe, liquid investment product for many risk-averse 401k plans,
pension plans, and insurance companies. Without this asset class, these
investors would struggle to replicate the combination of liquidity and
return, and would either move toward lower yielding products such as
Treasuries, or into riskier products such as corporate or other
sovereign debt. Such shifts in asset allocation would not only reduce
the flow of capital to mortgage markets, but it could also have a
negative impact on the performance of those investment vehicles in
times of stress.
A related issue in many discussions of housing finance reform
regards the appropriate number of number of chartered GSE-like
entities, with or without a guarantee. These would be organized by the
Government or by the private sector as co-ops or otherwise. Regardless
of specific structural form, we note that an increase in the number of
entities will not necessarily reduce risk, as the performance of each
entity will be strongly correlated. They all will make the same bet on
U.S. housing, and to the extent we have another national downturn, they
all will suffer. Also, because of a lack of diversification, a given
entity would be more exposed to regional economic downturns.
Organizationally, we also see challenges in recruiting 10, 15, or 20
skilled management, and especially risk management, teams. Furthermore,
to the extent a TBA market would be viable (see our discussion above);
a larger number of issuers would serve to fracture liquidity into
multiple smaller markets. Put simply, a trader can only monitor so many
screens at one time, and a large part of the liquidity in a given
market is derived from its size. To the extent that a larger number of
entities is a desired policy choice, we think it will be critical to
(a) have only one security issuer that (b) issues diverse pools
collateralized by loans from all of the issuing entities (i.e., similar
to Ginnie Mae's multi-issuer pools). This would create a larger,
unified securities market to stand behind the more fractured front end
of the system. This would minimize any regional differentiation in
pricing, maximize liquidity, and maximize the benefit to consumers.
SIFMA believes that the current situation is undesirable and
unsustainable and must be changed. We also believe strongly that
private capital should stand in front of any backstop or guarantee on
MBS. We note that it is a policy choice to decide the appropriate size
of the TBA market. Our concern lies with the end result, and that the
end result is liquid and beneficial to lenders, investors, and
consumers.

2. The Importance of a Smooth Transition to the Future Housing
Finance System, and the Recovery of Non-Agency MBS Markets
The future of mortgage finance in the U.S. is a critical policy
decision facing Members of Congress. The impact of this decision will
reverberate across the nation's housing markets, across financial
markets, and across the economy. It is no exaggeration to say that the
future state of the housing finance system is central to the future of
our nation as a whole. Regardless of what Congress chooses, the
transition from our delicate current situation to the future must be
carefully considered.
We have discussed above SIFMA's view that the TBA market is central
to the functioning of our mortgage markets. To the extent that Congress
desires to create a new mortgage finance regime that makes this
possible, SIFMA would strongly support doing so. It will be important
to put in place the basic structures that are required, as we have
discussed, to allow for a transition from one TBA environment to the
next with minimal disruption to current securities or mortgage markets.
Such a regime would allow for the preservation of a homogeneous
mortgage market eligible for TBA trading.
To the extent that Congress decides to significantly pull back or
completely eliminate the Government support for mortgage lending and
thereby significantly shrink or make impossible TBA trading, it will be
important to create a smooth path from the current state, which is over
90 percent Government supported, to the future state. Ultimately, as
the Government role is pulled back, something or a combination of
things must fill in the hole in mortgage funding that will be left
behind.
In either case the role of the Agency MBS market should and will
shrink from where it is today. Likely the most critical of the
components that will allow this to happen will be the reinvigoration of
the non-Agency MBS markets. These markets, aside from a few small
transactions, have been dormant in terms of their funding of new
origination. The bottom line to get these markets going is that we must
get to a point where issuers of MBS and investors in MBS see eye to eye
on the value proposition. Investors must receive a return that meets
their needs, and issuers must pay a cost that works economically. There
are a number of obstacles in the path. For example, many investors
suffered significant losses on holdings of non-agency MBS in the latter
part of the last decade, and it will take time for confidence to be
fully restored in those products. Mortgage demand from consumers,
because of the depressed economy, has significantly dropped.
Importantly, both investors and issuers face significant regulatory
uncertainty in addition to and apart from of the issues presented by
resolution of the conservatorships of Fannie Mae and Freddie Mac.
For example, servicing is a key component of the value proposition
for non-agency MBS. At this time, the future regulatory regime for
servicing is up in the air. Investors have identified a number of
concerns with current and past practices, and the market expects that
the current paradigm may see dramatic changes. However, no one is
certain of the timing or scale of these changes, which creates
significant uncertainly. A precedent-setting settlement of major
servicers with the State Attorneys General is expected, but the scope
and timing are unknown. FHFA is leading an important industry
discussion of the potential for revisions to the compensation of
servicers in Agency and non-Agency markets, but again, the end of the
story is still being written. The SEC in 2010 proposed a major set of
revisions \4\ to rules that govern asset-backed securities, some of
which were re-proposed,\5\ and some of which have been finalized,\6\
but the most critical elements are not yet final.
---------------------------------------------------------------------------
\4\ SEC's April 2010 Asset-Backed Securities rule proposal here:
http://www.sec.gov/rules/proposed/2010/33-9117.pdf, SIFMA comment
letter in response here: http://www.sec.gov/comments/s7-08-10/s70810-
79.pdf.
\5\ E.g., rules related to asset-level disclosure, shelf
eligibility, and disclosure in non-registered transactions reissued for
comment on July 26: http://www.sec.gov/news/openmeetings/2011/
agenda072611.htm.
\6\ E.g., disclosure related to repurchase demands: http://
www.sec.gov/rules/final/2011/33-9175.pdf.
---------------------------------------------------------------------------
Another issue relates to recently proposed credit risk retention\7\
and mortgage underwriting rules.\8\ The three main issues here are risk
retention, the definition of a Qualified Residential Mortgage (QRM),
and the definition of a Qualified Mortgage (QM). Each of these items is
open, and is expected to be finalized by regulators in the future. Risk
retention rules by their nature will change the economics of many
securitization transactions. In part, this is expected to help restore
the confidence of investors in securitized products and therefore
stands to provide a benefit to the securitization markets. On the other
hand, this benefit must be balanced by the preservation of
securitization as an economical funding alternative for lenders.
However, certain proposed provisions found in the credit risk retention
proposal, such as so-called ``premium capture'', have raised concerns
among many market participants as to their potentially devastating
impact on the economics of, and therefore future of, many type of
securitization transactions.\9\
---------------------------------------------------------------------------
\7\ Credit Risk Retention NPR: http://www.sec.gov/rules/proposed/
2011/34-64148.pdf, SIFMA's sponsor/issuer response: http://www.sec.gov/
comments/s7-14-11/s71411-79.pdf, SIFMA's AMG response: http://
www.sec.gov/comments/s7-14-11/s71411-80.pdf.
\8\ See proposal from Federal Reserve Board under Regulation Z that
would require creditors to determine a consumer's ability to repay a
mortgage before making the loan and would establish minimum mortgage
underwriting standards: http://www.federalreserve.gov/newsevents/press/
bcreg/20110419a.htm.
\9\ For further discussion of premium capture and its potential
impact, see SIFMA letters referenced above.
---------------------------------------------------------------------------
We expect the final form of the QRM and QM definitions to
essentially define the shape of the mortgage market after they become
effective. We expect that there will be little or no lending that falls
outside of the QM standards, given that significant liability may
attach to such loans. SIFMA has advocated that the final rules
delineating QM include a true, bright line, legal safe harbor so that
lenders will be comfortable to originate, and secondary markets will be
comfortable to purchase QMs in steady volumes. Many expect that
mortgage rates on QRMs will be lower level than those of non-QRMs (to
an extent that is unknown). Regardless of any individual decision with
respect to QM and QRM, it is important that these regulations be
closely coordinated and finalized in a manner where it is explicit to
lenders and secondary markets what is, or is not, a QM or QRM.
There are also significant regulatory revisions being made to the
permitted activities of banks, to global and national capital
standards, to the activities of credit rating agencies, the process of
obtaining ratings, and the usage of their ratings. These changes
include the capital treatment of mortgage servicing rights, eligible
assets for various liquidity and capital buffers, and more generally
changes to the capital treatment of securitized products.
All of this contributes to a great uncertainty as to the size,
scope, and liquidity of securitization as a funding tool for consumer
credit. It is difficult for lenders and creditors to make long-term
plans for how they want to run their lending programs and how they will
fund them, and it is difficult for investors to know the terms on which
they will be expected to invest. Key principles must be followed to
resolve this uncertainty: (1) Regulatory changes must be coordinated
and sequenced properly; (2) changes must be based on robust data
collection and analysis; (3) changes must keep in mind the dual needs
of any financial markets: investors must receive adequate returns, and
issuers must be able to fund at affordable cost levels.
All of these changes that directly impact the non-Agency markets,
and the goal of promoting the responsible resurgence of those markets
must then be viewed in connection with the resolution of Fannie Mae and
Freddie Mac, as they cannot be separated. One cannot come before the
other--they must work together. The ultimate question, yet to be
addressed, is that of the capacity of other forms of funding of
mortgage finance, be they non-agency securitization, covered bonds, or
new measures, to replace the support for mortgage lending that the
Government currently provides.

F. Conclusion
SIFMA greatly appreciates the opportunity to present this testimony
today and we hope that it is useful and informative to members of the
Subcommittee. SIFMA believes that the TBA markets play a critical role
in the current housing markets, and have provided tremendous benefits
to mortgage markets and consumers of mortgage loans. SIFMA therefore
believes that TBA markets can and should play a role in the future
housing finance system in this country. Regardless of the path chosen
for mortgage finance, SIFMA believes it is critical to properly
transition from the current market structure to the future. We stand
ready to assist Congress in any way necessary.
______

PREPARED STATEMENT OF PAUL T. VAN VALKENBURG
Principal, Mortgage Industry Advisory Corporation
August 3, 2011
Members of the Subcommittee:

Thank you for the opportunity to testify before the Subcommittee
today. Our firm, the Mortgage Industry Advisory Corporation (``MIAC'')
was started in 1989 and we serve mortgage market participants from
smaller mortgage companies to the largest banks and MBS investors. In
one of our product offerings, we offer loan origination hedge advisory
services to mortgage companies who want to mitigate their market and
fallout risk of loans and commitments of loans to borrowers. In order
to offer this service, we need to capture the most current pricing
information from the TBA (To-Be-Announced) securities and from the pool
insurance providers (currently only Fannie and Freddie) so that
mortgage companies can reflect these terms to prospective borrowers.
The TBA market has served to attract trillions of dollars in additional
capital to our housing finance system and to enable borrowers to have
access to fixed-rate 30-year mortgages. Fixed rate 30-year mortgages
enables borrowers to finance more home with a certain cash-flow
liability. I hope to offer some additional clarity and detail around
how these processes function in today's mortgage market and how they
might work in prospective TBA and mortgage markets.
The Committee has prepared a set of questions. Rather than answer
each one individually, I chose to provide a high level description on
the process whereby the TBA markets interact with the loan origination
process. In doing so, I believe that I will address each of these
questions with more clarity. In financial markets so much of what makes
a market or a product successful is a marginally better solution, e.g.,
slightly lower financing costs, slightly better liquidity,
incrementally better servicing advance terms, etc. It is therefore
necessary to have some specific descriptions of the market mechanics in
order to fully access their current role of the TBA market and how a
prospective housing finance system could be structured successfully.

1. What is the purpose and function of the TBA market?

a. What is the role of the TBA market in providing liquidity and
facilitating mortgage-backed securities (MBS) trades?

b. How does the TBA market affect certain mortgage constructs,
including the provision of 30-year fixed rate loans and
interest rate locks?

c. Does the TBA market allow for greater investor participation
and diversity than would exist without the TBA market?

d. How does the TBA market impact mortgage rates?

How the TBA market impacts loan terms for borrowers
Mortgage companies want to be in the business of creating mortgages
for prospective borrowers and not in the business of speculating on
interest rates and credit risk. The mortgage industry has developed so
that the price information of mortgage assets in the capital markets
and the credit risk insurers can be utilized by mortgage companies to
offer loan terms to borrowers reflecting the most current risk pricing.
This process is enormously complex on several fronts--the scope of the
data collection, the timing of the data collection, the legal
underpinning, the market conventions, and how all these players
interact and compete.

The Tradeweb electronic trading screen with the current and
executable TBA market prices. See www.tradeweb.com/businesses/rates/
tba_mbs for more information.
The image above is a screen shot of the TBA market on July 26th,
2011 at 3:11 PM EDT. The bids and offers of individual broker dealers
are consolidated and displayed on the Tradeweb screens. Market
participants still execute with their chosen broker/dealer and they can
use Tradeweb screens to execute. Smaller mortgage companies can't get
approved as counterparties to large broker dealers, so the pricing that
they execute at is usually \1/64\ or \1/32\ behind these screen prices.
The market is a forward market where securities are created and settled
in the future. The mortgage companies are modeling and estimating when
their loan commitments will close and how long it will take and the
cost to get a MBS pool certificate from Fannie or Freddie.
The TBA pricing on the screen are the most liquid segment of the
MBS market. The screen shows the FannieMae 30-year with pass-through
rates from 3.5 percent to 6.0 percent for settlement in August,
September, or October. The prices are quoted in percentage of notional
amount with a somewhat usual convention. For example, the bid side of
the October delivery FN4's is 99 (``the handle'') and 31 32nds plus
one-64th or a plus ``+''. The number just to the right of ``99-31+'' is
the offer side of the market. The ``00+'' means that the offer side
price is 100-00+ or 100 percent plus \1/64\. The difference between the
bid side of a market and the offer side of a market is a common method
for measuring the liquidity of a market. Market participants would say
that bid/offer spread is \1/32\. However, the depth of the market is
measured by the difference between the highest bid and the next highest
bid. Because large pools of mortgage are created with very similar loan
terms, the cover bids/offers are usually only \1/64\ off highest bid or
lowest offer. The TBA market is both liquid and deep. However, only the
largest mortgage companies can execute directly with the largest
broker/dealers who offer the best pricing. Slightly smaller companies
will execute with regional broker/dealers with a \1/64\ or \1/32\ price
difference because of their incremental counterparty risk.
If a mortgage company anticipates creating a mortgage on July 26th,
they are estimating that they will have the loan closed and eligible
for delivery into a FN30-4 security for the October month settlement.
They could sell the majority of their risk as a FN30-4 TBA for 100
percent and \1/64\ of a percent (100-00+) in the TBA market with a
large broker dealer or (100-01) with a regional broker. They would also
have to pay for the guarantee fee and some loan level price adjustments
to FannieMae. And they would have to decide whether they would prefer
to retain the mortgage servicing rights or sell/release the mortgage
servicing rights to another mortgage company. Below is a diagram
illustrating the basic flow of the pricing information from the TBA
market to the borrower.
As part of the price discovery process for mortgage companies, the
GSEs will charge a guarantee fee to provide pool insurance for the
investors in the MBS pool. The guarantee fee for a typical qualifying
conventional mortgage is 0.0025 or 0.25 percent. However, for loans
that are still eligible for the GSE pool insurance but with a more
credit risk, the GSE will charge substantially higher guarantee fees.
These guarantee fees are called Loan Level Price Adjustors (``LLPA'')
and the industry refers to them as risk-based pricing. For example,
loans with higher LTVs, low FICOs, Investment Property, Condo/Coop's,
and Cash-Out Refi's have LLPA's. I've included FannieMae's Selling
Guide with the most current LLPA's with this testimony.

The Loan Creation Process from Application to TBA

Initial Application--Loan Broker or Retail Loan Officer
(subset of necessary data for full underwriting; data needs to be
verified)

Application Approved/Loan Commitment created
(complete set of collateral and credit data for full underwriting;
verification of employment; verification of income; appraisal (IRLC
could still be subject to subsequent appraisal))

IRLC (Interest Rate Lock Commitment)--Exposes the mortgage company to
fall out risk and market risk.

Loan Closes and is subsequently QC'ed and delivered to Fannie for the
creation of a TBA security. The mortgage company obtains a pool
certificate number from Fannie.

Mortgage Company can sell the IRLC (Best Efforts) to a larger mortgage
company and the larger mortgage company can measure and manage the fall
out and market risk. Or the mortgage company can retain the IRLC and
measure and manage the fall out and market risk with their balance
sheet. If they retain the IRLC, they will be selling the future loan
with the Mandatory delivery option, meaning they will be required to
deliver the loan or pool of loans. The party that sells securities
(after the loans have been approved as collateral in TBA pools by FN/
FH) into TBA pools is required (mandatory) to deliver the TBA
securities on the appropriate settlement dates. If they closed more or
less loans than they expected, they have to absorb the market
consequences.

The owner of the mortgage servicing rights must be a Fannie approved
servicer for Fannie securities. Approval is subject to minimum capital
requirements and regular business process performance audits.

e. Are there any alternatives that would accomplish the same
objective and what are the obstacles to the development of such
a market?

In the current mortgage origination process with a liquid TBA
market, mortgage companies can know with a sufficient clarity and near
certainty what the exit value of GSE eligible, fixed-rate loans will
be. They can even sell their IRLC's to a larger mortgage company that
will measure and manage the fall out risk, pay a competitive price for
the MSR component, and handle the data delivery requirements.
Over the past several years, bank balance sheets are constrained as
they grow their capital and operate under stricter underwriting
guidelines. Mortgages created for bank balance sheets (typically
adjusted rate mortgages or jumbos) have represented only a very small
percentage of loan origination. Mortgage products that are not eligible
for GSE guidelines do not have a liquid TBA market and as a result,
mortgage companies have not been originating non-agency eligible
mortgages. The vast preponderance has been 30-year fixed rate mortgages
that form the collateral of outstanding MBS and currently qualified for
new MBS pools. This new housing finance capital was distributed from
global investor to U.S. home borrower by means of the TBA market. Even
the Fed chose to provide their $1.2 trillion of capital through the TBA
market.

Homogeneity and Standardization
The ``Uniform Practices for the Clearance and Settlement of
Mortgage-Backed Securities'' created and maintained by SIFMA has
provided sufficient homogeneity and standardization of the individual
MBS pool attributes and the clearing and settlement practices to enable
a liquid market to be created. Investors in mortgage-backed securities
recognize that each pool is a population of individual mortgages and
therefore, their pricing and valuation metrics employ actuarial or
statistical assumptions about the payment and prepayment behavior of
the population. This valuation paradigm and the homogeneous pool
attributes enable market participants to trade in large pools of TBAs
as fungible securities. If a privately guaranteed TBA market were to be
developed, the market would surely coalesce to an AAA credit rating.
And SIFMA settlement practices ensuring market liquidity would only be
tweaked slightly. A slightly lower credit rating would make the terms
of the mortgage more expensive to borrowers, but the market liquidity
could be utilized in same manner as today by mortgage companies to
hedge their interest rate risk for creating long term fixed-rate
mortgages.
If a solely private TBA market was created today, I believe that
the size of this market would be substantially smaller than the current
Government supported market. However, the development of a solely
privately supported TBA would necessarily involve a longer term
transition. We have time to explore developing a solely private market
in parallel with the current Government guaranteed market. The largest
buyers of TBAs today are the Wall St dealers gathering collateral for
CMO issuance. They could take AAA rated TBA and credit enhance their
CMO bonds to distribute the bonds to a wider audience of investors
including typical CMO buyers such as mutual funds, insurance companies,
and other non-banking investors. The investor acceptance of these new
products would be a gradual process.
The principal challenge of a privately capitalized TBA market is
who would hold the first loss or mezzanine risk pieces in the privately
guaranteed TBAs. In the market right now, there is an enormous amount
of this risk that is available for risk takers to purchase. And the
dynamics are such that there is a very weak bid for holding this risk.
One of the structural problems is that nearly all regulated financial
institutions would not be permitted to underwrite this risk without
enormous capital charges and regulatory oversight. The current market
for holding this residential credit risk is saturated. At this point, a
Government guarantee is necessary. However, I believe that going
forward consumers, investors, and tax payers are best served with a
single TBA issuer, standardized market practices necessary to preserve
and enhance market liquidity and investor acceptance, and segregated
credit providers either solely private or a hybrid private/Government
solution.

Single Issuer--Separate ``Government-Backed'' Creditor
The GSEs currently offer their guarantee to a number of mortgage
products that have failed to become actively traded securities in
liquid TBA markets. Although many other GSE loan guarantee program
exist, they represent only a tiny minority of loans that are
originated. This implies that an explicit Government guarantee alone is
not sufficient to create a mechanism to attract capital market
participants to invest in federally guaranteed mortgage securities. The
standardization and homogeneity of the securities and their market
practices and Government guarantee of timely payments have led to
market liquidity. This liquidity has created marginally better pricing
to enable mortgage companies to offer marginally favorable terms to
borrowers for those loan products that become the collateral for liquid
securities. A single federally approved and regulated securities issuer
would serve as a gatekeeper of the investor acceptance of newer
guarantee products. Moreover, a single MBS securities issuer would act
as an implicit agent of the investors to ensure market acceptance of
the newer TBA products.

Pricing the Credit Risk Accurately
Currently the GSEs have migrated toward Loan Level Price Adjustors
as mechanism for accurately pricing the incremental credit risk
associated with incrementally higher credit risk in the underlying
mortgages. Are these LLPA's accurate in measuring the expected and
potential credit losses of these underlying loans? Frankly, no one can
accurate price this risk because no one can foresee all the future
economic scenarios that would result in loan losses. We've recently had
one economic scenarios where U.S. residential real estate declined
approximately 30 percent and certainly this one scenario would be
useful but by no means is it comprehensive in describing all the future
economic scenarios and what the direct impact will be on particular
mortgage borrowers and their future default behavior. Accurate pricing
involves estimating the probably that the future credit event happens.
Accurately estimating the probability of a 30-year credit event and its
impact on particular mortgages is nearly impossible.
And what's the value of this credit insurance to the investor or to
the housing finance system that benefits from this incremental flow of
capital? What's the price and value of this credit insurance next
month? Or years later? Quantifying the current market price and value
of a Government guarantee is daunting challenge without any ongoing
market that prices and transfers this risk between two parties. Can we
create a market place where this risk can have some price discovery?
The GSE Government guarantee is basically pool insurance. Pool
insurance replicates the economics of a credit default swap. The GSEs
are currently the largest writers of protection in illiquid CDS. If we
take these pool insurance contracts (CDS) and allow these contracts to
be bought and sold to other financial institutions, we can distribute
the credit risk more broadly throughout the financial system. And we
gain valuable price discovery on the cost of this insurance in an
actively traded market. This price discovery can be very useful for
creating innovation but also have the regulatory role in quantifying
the risk and for establishing appropriate capital charges.

Other issues with providing Government guarantees
Using the U.S. Government balance sheet to insure against
particular credit events has a long history of over extension and
mispricing of these guarantees. Ideally, if we could create a mortgage
origination system that allowed for the credit risk to be underwritten
and priced in the private capital markets, this would ensure a more
effective allocation of capital and a more effective economic system. I
believe we should start from a point of view of attempting to build a
solely privately capitalized system and then, if circumstances dictate,
migrate toward a solution that utilizes hybrid Government guarantees. I
believe a liquid TBA market could be created with solely private
capital. The question of whether this purely private market reduces the
size of the market materially and thus provides less capital to the
housing markets can be managed as the market evolves. And as the fall
of 2008 capital market events illustrated, a Government guarantee in
times of unparalleled market skepticism was necessary for the capital
markets to become restored and operational. The solution should include
a mechanism or structure for the Federal Government to intercede to
guarantee market liquidity in extreme conditions.

Government Guarantor Competition
Over the past 30 years, we've had two GSEs and GNMA competing for
originator acceptance and investor appetite. This competition has led
to innovation of the financial products and increased the capital flow
into the mortgage market. One of the MBS securities' terms that has
evolved most significantly over the years is the servicing remittance
terms and the interaction with the guarantee of timely payment of
principal and interest. In the current housing finance system servicing
remittance terms are again highly controversial and therefore an area
where market innovation could be adopted. (MIAC is not an advocate of
the alternative minimum servicing fee proposal. MIAC believes having
the servicer continue to have a financial incentive in the performance
of the loans is crucial.)
Also in the current market place, larger servicers are usually
charged a lower guarantee fee by the current GSEs for the same
underlying loans than smaller mortgage companies. If the market
acceptance of solely privately guaranteed TBAs does not prove
sufficient to providing adequate capital to our housing finance needs,
a system of multiple Government credit provider would be a preferred
approach. Having a larger number of GSE-like competitors to offer
Federal Government credit guarantees could allow for more product
innovation and help widen the product mix of mortgage products that
could have liquid markets. The appropriate location on the spectrum of
the credit quality of the underlying mortgages and the appropriate mix
and cost of the originator retaining risk versus the GSE-like credit
guarantor of timely payment of principal and interest could be
competitively determined. Federal housing policy could prescribe that
particular borrowers could be priced with more affordable credit
guarantee terms.
The FHLBs have successfully weathered the recent mortgage credit
crisis and have served as an enormous source of liquidity for mortgage
holders, particularly non-agency eligible and illiquid mortgages.
Although they have historically operated with a very different model of
offering liquidity, they would be natural potential credit providers
for newly created mortgages that would qualify for the single new
securities issuer's parameters. Having multiple Government credit
providers each constructing innovative solutions of creating hybrid
private/Government credit enhancements would enable innovation and
competition to create better solutions for investors and borrowers. All
Government credit providers would have the same regulator and capital
requirements.
Mortgage companies could seek outside credit providers such as the
current private mortgage insurers or work with broker/dealers to create
and price new products for participants interested in taking the credit
risk of residential real estate. Over the past 10 years, the corporate
bond market has evolved so that credit default swaps provide market
participants a liquid market for trading and pricing heterogeneous and
complex credit risk. Pool insurance is a particular example of a credit
default swap. However, liquidity in pool insurance is non-existence,
but credit default swaps can provide active price discovery and
liquidity. An active market for residential credit risk would create
some level of price discovery for the credit risk in residential real
estate and also allow the risk to be more widely shared throughout the
financial system. This price discovery would also assist each of the
Government guarantors and their regulators in measuring the current
price for particular residential credit risk. The GSE-like creditor
providers could trade their credit risk between GSE-like institutions
to manage their risk exposure. They could trade and manage their credit
risk with private market risk takers. Having reference pricing for a
long dated residential credit risk would help the regulatory role in
helping to determine the appropriate amount of capital required by the
small set of approved Government guarantors.
How would Congress distribute the Federal guarantee among the
various Federal credit providers each of which would have a different
mix of credit risk and amount guaranteed? Each GSE-like should be
separately capitalized with each institution's own capital acting as
the first line of defense against future claims. One guiding principle
should be that each GSE-like be capitalized to at least the level of
capital required for a private market insurance company. In addition,
Congress could establish global, system wide limits to the credit risk
exposure.
This proposal does not eliminate the moral hazard and distortion in
capital allocation that is created by a Government guarantee that has
been historically mispriced and credit providers undercapitalized. I
merely attempt to outline a structure that in many ways mitigates these
risks and addresses the fundamental challenges with pricing this risk.
I believe that going forward the flow of capital to housing finance
will be best served if we can create a system that continues to have
competition in the particular guarantee programs. However, in order to
ensure liquidity and market acceptance, a single issuer with uniform
and capital market accepted practices is necessary. I believe we should
segregate the guarantee function from the issuance function.

2. How will proposed housing finance reforms impact the TBA market?
Please consider:

a. Proposals that change (or impact) the Government-sponsored
enterprises (GSEs) role or any Government guarantee;

b. Proposals that increase the number of GSEs or other MBS
issuers; and

c. Other proposals that affect the structure of the housing
finance system.

Regarding recent proposals, I only have passing knowledge of other
proposals at this time. Constructing a successful solution involves a
detailed and careful analysis of the risks and these complex tradeoffs.

3. What, if any, changes or improvement can or should be made to
improve the functioning of this MBS marketplace?

One principal lesson from the recent failures of the GSEs is that
it isn't prudent to mitigate the risk of writing insurance on the U.S.
residential real estate industry by restricting their investment
portfolios to the purchase to only securities with risk exposure to
U.S. residential real estate. In my view, the Federal guarantee
providers of the future should be restricted from purchasing MBS. The
goal of the guarantee is to subsidize housing by creating more capital
to the residential housing market and the risk is absorbed as general
obligation of the United States. Having the mortgage insurance capital
pool invest in mortgage investments is a structural failure for the
current GSEs.
In the past the GSEs have also played a role in using a small
portion of their investment portfolios to support the prices and
liquidity of their MBS securities. Many market observers believe these
temporary price supports, in modest magnitude, have led to greater
private capital involvement in the TBA market. This may be the case but
I believe that this role can be served by the Fed in the future. The
Fed currently plays a similar role in the U.S. Treasury markets and
could play this role of supporting the price of mortgage assets in the
capital markets in the future.

A future TBA market
The current mortgage origination market has been providing new
capital to borrowers as the result of an active and robust TBA market.
This liquid TBA market allows mortgage companies an effective means to
price and sell/hedge their mortgage loans. The TBA market with a
Government guarantee enables 30-year fixed rate consumers to transfer
their interest rate risk to the investors and provide them with a fixed
liability stream for 30 years. This liquid market has been the source
of approximately 90 percent of mortgage capital in the past few years.
At a time of tremendous contraction in private capital into the housing
markets, the TBA market has been the sole beacon of success. However
going forward, I believe that we can create a TBA product that reduces
the risk exposure of the Government guarantee and provides a role for
private capital risk takers. This won't be a simple task, but
introducing mechanisms to enable risk to be priced and traded will
enrich our housing finance system. The future U.S. Housing finance
system must protect the strength's of the TBA market and the pocket
books of the taxpayer, because an effective and stable housing finance
system is essential to our collective prosperity.

PREPARED STATEMENT OF ANDREW DAVIDSON
President, Andrew Davidson & Co., Inc.
August 3, 2011
Mr. Chairman and Members of the Subcommittee:

I appreciate the opportunity to testify before you today about the
TBA or ``To-Be-Announced Market.'' Despite its prosaic name, the TBA
market is a crucial component of the housing finance system. I believe
it is one of the greatest financial innovations of the last 50 years;
another colleague has called it a ``national treasure.'' The TBA market
helps lower mortgage rates, facilitates rate locks for borrowers
seeking to buy homes and has helped the make mortgages available
through the financial crisis. Policy alternatives to the GSEs may
enhance or disrupt this market. It is my hope that my testimony today
might give you a better understanding of the functioning of this
market.
I have been involved with Mortgage-Backed Securities (MBS) since
1985. I was a managing director at Merrill Lynch responsible for MBS
research and risk management for their mortgage trading desk. In 1992,
I founded Andrew Davidson & Co., a New York based firm, specializing in
the development and application of analytical tools for the MBS market
that serves over 150 financial institutions. I have a broad view of
housing finance as our clients include originators, servicers, mortgage
insurers, GSEs, investors, dealers and regulators.
Today's TBA market represents a more than 40-year evolution of a
voluntary system of trading mortgage-backed securities that provides
for efficient, transparent risk-transfer and funding for most of the
fixed-rate mortgages originated in the United States. The market serves
two primary purposes: First it allows originators of fixed-rate
mortgages to hedge the price risk associated with changing interest
rates from the time that the originator makes a commitment to a
borrower to lend at certain rate, until the loan is sold to an investor
in the form of a mortgage-backed security (MBS). Second, it allows
investors to engage in extremely large transactions to buy or sell MBS
at very low costs of execution. In addition to its primary purposes,
the TBA market also provides a mechanism for investors to efficiently
finance their holdings of MBS and provide liquidity to the market,
through a mechanism called ``dollar rolls'' or ``rolls.''
I have divided my testimony into three parts. Part I is a general
discussion of the TBA Market. Part II is a discussion of how proposed
housing reforms might affect the TBA market. Part III contains some
recommendations.

Part I. The TBA Market
To better understand the value of the TBA market, it might be
instructive to see how the TBA market is used to reduce risk in the
mortgage origination process. This is in Section 1. In Section 2 we
discuss features of the TBA market that make it effective. Section 3,
is a discussion of why the TBA market is able to achieve these
benefits. Section 4 provides estimates of the cost benefit to borrowers
associated with the TBA market.

Section 1. The Role of the TBA Market in Origination: Hedging Interest
Rate Risk
In this section we examine the role of the TBA market in the
origination process via an example that shows how the interest rate
risk of originators looks with and without TBA hedging. Before
launching into the assumptions and looking at originator profit and
loss, we review the links between the origination of loans to borrowers
to the TBA market of MBS.

Process Summary
The largest broker-dealers maintain an actively traded TBA market
and the prices from this market can be seen by originators and
investors in real time on electronic screens. The prices of TBA
securities, together with the loan-level pricing adjustments (LLPAs),
base guarantee-fees of Fannie Mae (FNMA) and Freddie Mac (FHLMC), and
the required servicing fee help originators determine the mortgage rate
for any loan which is eligible for securitization through the GSEs. For
example if the par security (the one priced closest to 100) in the TBA
market has a 4.5 percent net coupon. The originator would add on a
servicing fee of 25 basis points and a guarantee fee of 20 basis points
to produce a mortgage coupon of 4.95 percent.
Suppose a borrower applies for a loan and locks in their rate in
August, and the originator hopes to complete underwriting and close the
loan in late September. Then the loan would be delivered into an
October TBA security (suppose FNMA for our example). Between August and
October, the LLPAs and servicing fee going into the borrower's rate
would not change; however, interest rates can change, and any change in
the level of prevailing mortgage rates affects the value of TBAs.
Because borrower rates are locked, the risk of a mismatch between the
rate given to the borrower and prevailing rates at closing is borne by
originators.
If the mortgage loan is a product that is eligible for delivery
into TBAs however, the originator can estimate how many loans they will
be delivering into the October security based on their application
pipeline, approval rate, and historical ``fall-out'' rates (the rate at
which approved borrowers fall out of the application process for
various reasons). The originator can then short (or sell) the
appropriate quantities of Fannie TBA securities to hedge their
interest-rate risk. When the settlement of that TBA security occurs,
they will deliver loans to FNMA and receive a FNMA pool which they can
deliver to cover their short position with the broker-dealer. The
originator is fully protected from changes in interest rates because
they have already locked in the price for sale of the pool.

A Hedging Example
For this example, we make the following assumptions: as above, the
net coupon in the TBA market is 4.5 percent and the coupon to the
borrower is 4.95 percent and originator profits are $0.50 for every
$100 of loan balance originated in the absence of any interest rate
changes (relative to forward rates). We examine three scenarios: (1)
where interest rates do not change from the time that the borrower
locks in their rate until the time that the loans are delivered into a
security, (2) interest rates fall 100 basis points, and (3) interest
rates rise 100 basis points. Since the time that elapses between a
borrower's lock to loan closing can range from 30 to 90 days, this
range of interest rate changes often occurs.
If interest rates fall the par net coupon in the TBA market will be
3.5 percent. The 4.95 percent loan to the borrower will be more
valuable since it carries an above market coupon. However if interest
rates rise, the new par net coupon in the TBA market will be 5.5
percent. The 4.95 percent loan to the borrower will be at a below
market rate and will have fallen in value.
Figure 1 shows originator net profit taking into account the $0.50
in fee income as well as the impact of interest rate changes. The three
lines demonstrate originator profit assuming three different actions by
originators: no interest rate hedging, hedging using TBA markets in the
same loan type as the originated loan, and hedging using a different
instrument (labeled ``Cross Hedging'').

In the middle scenario, where rates are flat, we can see that all
three types of behavior result in the same $0.50 net profit; that is,
interest rate risk has no role, and originators earn their desired
$0.50 of each $100 originated. However, we can see that without any
hedging, if interest rates fall, profit swings up to $4. If interest
rates rise, the originator takes a loss of $5. (The asymmetry results
from a characteristic of mortgages called negative convexity, caused by
the right of borrowers to prepay when rates fall.)
The red line, with TBA hedging, shows that these swings in
originator profit are completely flattened with the use of a TBA hedge
in the same category as the originated loan. The TBA market allowed the
originator to lock in a sales price.
The green line shows that using a different instrument to hedge
could result in losses of $0.40 and $0.60 in the changed rate
scenarios. An example of a cross-hedge would be originating 7/1 hybrid
ARMs and hedging them using 15-year fixed-rate TBA markets (because
ARMs do not have TBA markets). When it is time to sell the pool of
ARMs, the originator will not be able to deliver the pool to satisfy
the short position. Instead the originator will need to buy back the
short position in the TBA market and enter into a separate transaction
to sell the ARM pool. There is no assurance that the price change on
the TBA and the price change on the ARM pool will match. The difference
between these price changes is called ``basis risk.'' The losses in
this third case stem from having an imperfect hedge, and the numbers we
show reflect potential errors occurring in both directions. The amount
of gain or loss is uncertain because the hedge does not lock in the
sale price.
Cross hedging cannot reduce risk as effectively as a direct hedge
in the market of the product. The effectiveness of cross hedging
depends on the similarity between the product and the hedge and the
cost of execution in the hedge market. Because of its liquidity and
close relationship to other mortgage markets, the TBA market is widely
used to hedge many non-TBA eligible mortgage products including hybrid
ARMs and non-agency mortgages.

Alternatives
Originators prefer to focus on the underwriting and credit
decisionmaking process and would like predictable profits. If the TBA
market did not exist, some alternatives might be:

(a) Originators take the interest rate risk or use other, more
imperfect hedges. This would most likely result in an increase
in mortgage rates to compensate the originator for taking this
risk.

(b) Offer mortgages that do not have a rate-lock feature, leaving
the interest rate risk with borrowers. This makes the mortgage
qualification process somewhat difficult, especially for
borrowers that are near qualification limits. For example if
their debt-to-income ratio is near a lending limit and rates
rise, the borrower would no longer afford the subject property,
and the purchase could not go through. This option would raise
the overall volatility of the real estate transaction process.

It is possible that both would occur, with rate-lock loans offered
at higher rates, giving borrowers the option to take the risk in
exchange for a lower rate (which could end up higher or lower at
closing). Borrowers with economic flexibility might prefer to take
their chances, whereas first-time buyers who are near their purchase
price limits would likely be forced to pay the higher rates.

Section 2. Benefits of the TBA Market
The TBA market delivers benefits to its participants due to several
important features.
Physical Delivery. The TBA market allows mortgage originators to
complete the sale of newly originated mortgages directly into the TBA
market. As described above, the TBA market establishes the pricing for
mortgages, and originators are assured of achieving their expected
profitability if they successfully close and deliver the loan. Other
forms of hedging are generally cash settled or do not allow the
delivery of the mortgage loan. In these cases, the originating firm
must separate the hedging process from the ultimate sale of the
mortgage loan. This creates basis risk, as described above.
Limited Delivery Option. The TBA market operates with sufficient
clarity as to the nature of the loans that will be delivered to the TBA
market. This means that investors are confident of the value of the
securities they will receive, and therefore do not substantially
discount their purchase price to accommodate adverse delivery by the
sellers. (These markets operate under a principle called ``cheapest to
deliver,'' which means that the seller will find the lowest value
instruments to sell to the buyers.) Some loans do have greater value
than average, primarily due to desirable prepayment characteristics.
The TBA market has found a way to accommodate loans which might have
greater value through the use of pool specific and stipulated trades,
without overly degrading the value of the vanilla TBA trades.
Low bid-ask spread. The size and scope of the TBA market means that
it can deliver extremely low transaction costs to buyers and sellers
with typical bid-ask spreads in the range of \1/64\ of a percent to \1/
32\ of a percent. This is comparable to the bid-ask spread on the most
liquid Treasury securities. The TBA market allows investors to make
very large commitments of capital in very short periods of time, with
little or no impact on pricing in the market, making it one of the key
markets used by investors to express interest rate exposure objectives.
Even during the financial crisis bid ask spreads in the TBA market
remained in check. As shown in Figure 2, bid-ask spreads at the worst
of the crisis may have approached \1/2\ point, but during this period
many other markets were not trading at all. Even today, non-TBA trades
in the mortgage market might have a 1 point or greater bid-ask spread
while the TBA market has returned to a \1/32\ of a point or less bid-
ask spread.

Built in financing market/liquidity provision. Another important
component of the TBA market is that it provides a mechanism for
financing ownership of mortgage-backed securities, much like a repo
(repurchase) market. Owners of MBS can sell their positions into the
TBA market in a near delivery month (say August) and re-purchase that
position in the next month (say September). Since they receive cash for
their sale in August and then pay cash to reacquire the position in
September, the investors have effectively borrowed money from the TBA
market. The sale of mortgages for one period and purchase in a later
period is called a ``roll,'' as investors have ``rolled'' their
position to a later month.
This mechanism also provides a way for investors to get
compensation for providing liquidity to the market. If there is a great
demand for MBS in the current month relative to a future month the
price paid for current delivery rises relative to the forward price.
This serves to effectively lower the financing cost for holders of
mortgages. MBS holders can take advantage of this financing benefit and
provide additional liquidity to the market while maintaining their
investment in MBS.
Transparency. Participants in the TBA market have access to current
pricing information from a variety of sources. For example, Bloomberg
and Tradeweb post current prices.
Analytical Tools. Investors and Originators using the TBA market
have access to a wide range of historical data and analytical tools
that help them assess value and risk in this market. While there is
substantial volatility in MBS prices, there is no shortage of
information and tools to help market participants assess the risks of
TBA eligible MBS.
Cross hedging. The TBA market also serves as a primary hedging tool
for non-TBA eligible agency loans and for non-agency mortgages. That
is, even loan products that are not eligible for physical delivery into
the TBA market make use of the TBA market for price risk reduction, but
with greater risk to the originator. Without the TBA market it would be
more difficult to hedge and manage the risk of non-agency fixed-rate
mortgages as much of the hedging for non-agency fixed-rate mortgages,
as well as many hybrid ARM products, utilizes the TBA market. Much of
the efficiency of these products is derived from the TBA market for
fixed-rate mortgages.

Section 3. Sources of Value of the TBA Market
As policy options for the GSEs are considered, policymakers may
want to consider what structural features allow the TBA market to be so
successful and consider whether proposed changes might impact the
functioning of the TBA market. The functioning of the TBA market is the
result of a combination of features. The success of a market is largely
determined by the confidence of the market participants. That
confidence creates liquidity, which is further self reinforcing. A loss
of confidence can lead to a rapid decline in liquidity and the collapse
of a market. I believe that participant confidence in the TBA market
arises from four key features. Changes to these four features might not
destroy the TBA market, but would likely reduce investor confidence.
The Government guarantee is a central feature of the TBA market.
GNMA securities have traded with the explicit guarantee of the U.S.
Government, and GSE securities traded with an implicit guarantee until
the conservatorship of the GSEs and now are backed by funding from
Treasury. The Government guarantee serves to eliminate the need for
credit analysis when evaluating TBA transactions. Investors do not need
to consider the credit worthiness of the borrowers, the adequacy of
credit enhancement, or the financial strength of the issuer when
investing in TBA eligible mortgages. Furthermore, due to the Federal
guarantee, GSE mortgages are exempt from SEC registration. This
exemption facilitates the TBA market because firms can sell securities
prior to issuance. For registered securities there is a prohibition on
sale prior to registration so originators are not able to sell pools of
loans that have not yet closed.
The underwriting requirements of the GSEs and the limitations on
the type, nature, and documentation of mortgages allowed in various
types of mortgage pools provides investors with confidence that the
mortgages in the pools that they buy will be sufficiently similar so as
to make forecasts of cash-flows reasonably certain. The GSEs have
served to standardize the entire mortgage origination and servicing
process. Without such standardization, investors would be less willing
to engage in TBA trades or would demand a greater premium for the risk
that they would receive non-standard loans and pools in TBA delivery.
The Securities Industry and Financial Markets Association (SIFMA)
also serves an important function by setting the rules on good delivery
for TBA trades. SIFMA standardizes delivery dates and notification
rules and limits which agency pools qualify for delivery into TBA
pools. This role protects investors from sellers including pools which
might have adverse performance characteristics in TBA pools.
While these structural features and the roles played by the GSEs
and SIFMA have been crucial to the success of the TBA, another
significant, but more elusive feature of the market is also important.
As this market has developed over the past 40 years, it has adapted to
changing market conditions. The ability of this market to adapt
enhances participant confidence. Examples of adaptations are: the
growth of the dollar roll market (described above), the on-going
evolution of settlement and clearing operations through DTCC and its
predecessor organizations, updated and evolving good delivery
guidelines from SIFMA and its predecessor organizations, the evolution
of the stipulated trade market as the GSEs produced enhanced data
disclosures, and the development of electronic trading platforms such
as Tradeweb.
These features enabled the MBS TBA market to withstand the recent
financial crisis with virtually no disruption. For example, the DTCC
risk management and clearing operations were able to shield key market
participants (its clearing members) from the failure of Lehman
Brothers.
The success of this market is a reflection of the confidence of the
participants to engage in tens of trillions of dollars of transactions
each year. Market participants estimate monthly trading volumes of
about $5 trillion split between dealer-to-dealer and dealer-to-customer
business. Average daily volumes are estimated at $300 billion and are
substantially higher around settlement days.

Section 4. Impact on Mortgage Rates
The TBA market has a significant effect on the availability and
cost of fixed-rate mortgages. It contributes in three important ways.
First, the extremely low bid-ask spread and high liquidity lower
the transaction costs for originators to sell mortgages. During normal
markets, the TBA market has a bid-ask spread that is \1/4\ to \1/2\
points lower in price (5-10 basis points in mortgage rate) than
alternatives. During the crisis period there were times when it was
nearly impossible to execute in non-TBA markets. Even several years
after the crisis, the bid-ask spread for many senior non-agency
mortgages is more than one point.
Second, the ability to hedge origination risk with an instrument
that allows physical delivery of the loans lowers the cost of hedging.
During normal time periods this probably lowers the cost of mortgages
by about \1/2\ point in price (about 10 basis points in mortgage rate).
It is important to note that the TBA market also lowers the cost of
non-TBA mortgages as it provides a good vehicle for cross hedging. Due
to the unique characteristics of mortgages, particularly prepayment
risk, other instruments are generally not close substitutes.
Third, the liquidity of the TBA market, combined with the
Government guarantee on the MBS serves to lower the rate on agency MBS
by about 25-50 basis points in rate relative to non-agency alternatives
during normal markets. As shown in Figures 3 and 4, during the
financial crisis that spread rose to more than 400 basis points and
still remain at much higher levels.

Overall the TBA market lowers mortgage rates for both TBA eligible
and non-eligible mortgages by about 30-70 basis points in normal
markets and facilitates lending that might otherwise be prohibitively
expensive during crisis periods.

Part II. Housing Finance Reform and the TBA Market
The process of reforming the housing finance system following the
financial crisis is not yet complete. The GSEs, Fannie Mae and Freddie
Mac still operate under Federal conservatorship and the private label
securitization market has not yet recovered. Over the past 3 years,
there have been many proposals to reform the housing finance system.
Due to the importance of the TBA market to the functioning of the
housing finance system, it is important to assess the impact of these
proposals on this market. Rather than assess each individual proposal
for the GSEs and their possible successors, for the purpose of this
analysis common features of those proposals will be addressed.

1. Retain or eliminate a Federal guarantee on the MBS.

2. Reduce or increase the number of entities that can issue
guaranteed MBS.

3. Guarantee MBS, not GSE obligations.

4. Provide only a catastrophic guarantee.

5. Utilize covered bonds.

Analysis of these points is difficult in that there are no
objective criteria to say what is required to maintain a successful
market. The success or failure of a market is largely a result of
investor need and investor confidence. The TBA market has proven itself
to be resilient to many changes in the mortgage market over the past 40
years. This resilience derives at least in part from the efforts of the
participants in the market to address problems as they arise. The
fundamental (unanswerable) question about a proposed change in the
structure of the GSEs is whether the market can adapt to the change.

Eliminate the Federal Guarantee
Eliminating the Federal guarantee on conventional MBS would be a
major blow to the TBA market. The Federal guarantee serves to insulate
investors from credit risk. As a result, investors do not need to
consider the credit worthiness of the borrowers or the issuers. If the
underlying MBS had a wide range of credit risk exposure, only the
riskiest, least valuable mortgages would be delivered into the TBA
forwards. This would rapidly degrade the value and liquidity of the
market. Issuers producing high quality loans would be unable to use the
market as a hedging and delivery vehicle. Tradeweb, the electronic
trading platform which handles about 65 percent of all dealer-to-
customer trading in TBAs and is involved in many other markets, does
not currently handle any markets that do not have Government
guarantees. This provides a strong indication of the importance of a
sovereign guarantee in promoting liquidity.
Eliminating the Federal guarantee would also, presumably, eliminate
the SEC exemption for conventional MBS. Without this exemption, firms
would be unable to sell MBS prior to issuance, thus would lose the
ability to hedge as they do currently. All hedges would need to be
``paired off'' and would result in additional basis risk and cost.
Finally, eliminating the Federal guarantee would also remove a
substantial portion of the investor base from the TBA market. Many
large investors utilize the mortgage-backed securities market to
execute trades driven by macroeconomic views and would not utilize a
market which combines credit risk with interest rate risk. With a
smaller investor base, liquidity would be dramatically reduced. It is
likely that the proportion of fixed-rate loans would be substantially
and permanently reduced and mortgage rates would be higher without the
twin benefits of the Government guarantee and the TBA market.
While it is unlikely that the TBA market, in its current form,
could survive the loss of the guarantee, it is likely that other
mechanisms to hedge and trade mortgage-backed securities would be
created by market participants. Such mechanisms, including futures
contracts on treasuries, interest-rate swaps, and credit-default swaps,
already exist. These vehicles as well as new vehicles would likely pick
up market share if the TBA market was not viable. The GNMA TBA market
could also continue to function separately if the FHA continues to
guarantee loans. There would likely be a long adjustment period before
any of these could match the liquidity, cost effectiveness, operational
efficiency, and stability of the TBA market. The loss of the TBA market
would likely lead to further disruptions in the housing finance as the
market shifts away from fixed-rate mortgages and markets slowly
develop.

Increasing the number of GSEs
Many proposals require the creation of numerous GSEs to decrease
concentration risk and increase competition. My view is that this is a
slippery slope for the TBA market. The TBA market has benefited from
the close cooperation between SIFMA and the GSEs. While they do not
always agree, they recognize the importance of maintaining the TBA
market. With more issuers such cooperation would be more difficult. In
addition, it is unlikely that securities issued by different issuers
would all be accepted as good delivery for a single TBA by market
participants. Even now Ginnie Mae, Fannie Mae, and Freddie Mac
securities all trade in separate markets. Even with additional issuers,
market participants would likely concentrate their trading in one or
two issuers. These issuers would then have a competitive advantage over
the other issuers leading once again to a concentrated market.
A large number of competitive issuers is not necessarily a good
thing for the market. The non-agency mortgage market had a large number
of issuers. (See Figure 5.) This led to a race to the bottom and ever
more complex securitization structures as issuers initiated changes to
their securitization programs to boost profits through product
differentiation or to subtly shift value from investors to issuers.
Note that most of the top 10 issuers of non-agency mortgage-backed
securities are gone and the non-agency mortgage market has not
recovered from the crisis. In addition, a substantial difficulty in
resolving the housing crisis is a result of the wide range of
contractual features created by numerous issuers.

Some proposals attempt to address the problem of multiple issuers
by having all MBS issued by a single Government entity, essentially
like GNMA. This proposal would help the TBA market provided that the
loans from the different originators had substantially similar risk
characteristics. If each originator had different underwriting,
documentation, or servicing standards, the market would fragment even
if they shared a common issuer and guarantor. Maintaining sufficiently
similar programs across multiple issuers may not be possible.
A single Government issuer, enforcing strict guidelines, might also
be detrimental to the TBA market as one of the hallmarks of the TBA
market is that it has been able to balance the needs of many
conflicting parties. A Government-run guarantor might not have the
flexibility to adjust to changing conditions. The long, difficult road
to eliminate seller-financed down-payments for FHA loans is an example.
SIFMA has been successful because it has been able to recognize and
address investor concerns as they arose. If a problem that arose in the
Government securitization program was not addressed expeditiously, it
could lead to a loss of confidence in the market.

Only Guarantee MBS
Many proposals would only allow a Government guarantee on mortgage-
backed securities, and not on debt issued by the GSEs. Such proposals
would likely have a positive effect on the TBA market. Investors would
retain confidence in the guaranteed MBS, and without a guarantee on
their debt, the GSEs would be less willing to grow large portfolios of
mortgages and mortgage-backed securities. In fact, many proposals
explicitly prohibit the GSEs from retaining a mortgage portfolio. These
proposals would likely have a positive impact on the TBA market,
because they would remove a conflict within the GSEs. The large
portfolios at times might have motivated the GSEs to encourage higher
spreads and less liquidity in the market. In addition they were
competing with investors to purchase the best loans. That competition
at times led to lower values and less confidence in the TBA market. If
the competing portfolio incentive is reduced or eliminated, the GSEs
would focus more directly on enhancing the value of TBA eligible pools.
Reducing the GSE portfolios could have the effect of increasing
mortgage rates as a large investor is taken out of the market. It would
be important for this process to be gradual and transparent, so the
market had time to adjust to the change in investor base.

Catastrophic Guarantee
Some proposals have suggested that the Government only provide a
catastrophic guarantee to mortgage-backed securities. If structured
appropriately, such an approach could allow the TBA market to continue
substantially unchanged, while protecting tax payers from significant
risk. The important component of such a guarantee is that the investors
in the TBA-eligible MBS do not need to assess whether or not the
Government guarantee will protect them. That is, the investors want to
know that they have the full assurance of the Government that they will
be paid regardless of the credit performance of the borrowers, issuers,
or guarantors. Approaches where the investors retain risk for the
failure of the issuer or guarantor are less likely to be consistent
with the continuation of the TBA market.
A catastrophic guarantee can be created at the issuer level or at
the MBS level. At the issuer level the guarantee would require that the
issuer maintain sufficient capital to cover potential losses. The
Government guarantee would cover losses once the issuer failed.
Provided that the Government guarantee covered the full obligation of
the issuer whether or not the issuer was properly capitalized, much
like deposit insurance, investors would not need to focus on the credit
worthiness of the issuer. The Government would need to actively
regulate the issuers and guarantors to assure that they had sufficient
capital.
A catastrophic guarantee could also be provided at the MBS level.
In this case a portion of the MBS would be guaranteed by the
Government; the remainder would be subject to credit risk. In this
solution, the senior guaranteed bonds could trade in the TBA market,
while the non-guaranteed portion would trade in a separate market. I
favor such an approach as it can minimize taxpayer exposure while
maintaining the liquidity of the market. Freddie Mac has issued
securities with a guarantee only on senior bonds in the multi-family
market, demonstrating the viability of this approach.

Covered Bonds
Covered bonds generally are not a solution for fixed-rate mortgages
as they do not transfer interest rate risk and prepayment risk to
investors and would not be consistent with the TBA market. The Danish
covered bond market is an exception in that Danish Covered Bonds are
essentially mortgage pass-throughs with a guarantee from the
originator/issuer. A similar system in the United States could provide
alternative methods for hedging and funding of fixed-rate mortgage
loans but would probably take some time to develop sufficient liquidity
and institutional support to be a viable substitute to the TBA market.

Part III. Recommendation
Given the importance of the TBA market, the best strategy for
reforming the housing finance system may be to make a series of
transformational changes to the current structure of the GSEs rather
than scrapping the existing system and starting anew. By building off
the current structure of the GSEs, it may be possible to preserve or
even enhance the TBA market while addressing flaws in the existing
structure that contributed to the financial crisis. Given the weak
state of the housing market and the economy, completely eliminating the
GSEs or completely replacing them with an alternative structure would
likely be severely disruptive. Gradual transformation of the GSEs would
also allow the private MBS market and other alternatives to develop.
One such step that is possible without a complete dissolution of
the GSEs is to increase the amount of private capital ahead of the GSE
guarantee, thereby decreasing risk to the taxpayers. It is unlikely
that the GSEs could raise additional equity capital until their future
role and structure is determined. Moreover it will probably be better
not to reconstitute the GSEs as shareholder-owned companies which can
deliver a Federal guarantee. Thus capital would need to be provided in
a different form.
I believe that private capital can be put in front of the GSE
guarantee through the use of commonly used credit enhancement
structures. Mortgage insurance, either at the loan level or the pool
level, could be used to reduce risk to the GSEs. As some proposals
favor the use of this type of structure for the future of the housing
finance system, this would be a good opportunity to test these ideas.
Greater use of mortgage insurance would require that the insurers had
adequate capital to back up their obligations.
An even better approach, in my opinion, would be to encourage, or
require the GSEs to utilize a senior subordinated structure to attract
capital that would stand in the first lost position, either side by
side with the GSEs or ahead of the GSEs for some of their MBS issuance.
Such an approach would reduce risk to the taxpayers and help the GSEs
and regulators determine the cost and availability of private capital.
If properly executed, senior securities created under this structure
and guaranteed by the GSEs would remain eligible for TBA delivery,
while the junior classes, which would not have a GSE guarantee, would
trade separately. If such a program were successful, it could be
expanded, and if not, policymakers would better understand the
obstacles to replacing the GSEs.
Utilizing the current capabilities and infrastructure of the GSE to
implement housing finance reform offers the best chance to improve our
economy without the risk of severe disruptions.

Summary
The TBA market is an important component of the housing finance
system. It is currently central to the pricing and hedging of fixed-
rate mortgages. The TBA market helps lower mortgage rates, facilitates
rate locks for borrowers seeking to buy homes and has helped the make
mortgages available through the financial crisis. The TBA market has
evolved over a 40-year period and has proven to be resilient. It
functioned extremely well during the financial crisis.
The continued success of the TBA market depends upon the confidence
in traders, and confidence is difficult to measure or forecast.
Therefore, it is difficult to determine in advance what changes to the
market would be detrimental. It is likely that elimination of the
Government guarantee would severely disrupt the TBA market and
permanently reduce the availability of fixed-rate mortgages. Other
changes to the structure of the housing finance system may have
positive or detrimental impact on the TBA market, but those effects are
harder to predict. While it is likely that other mechanisms could
replace the TBA market over time, it is unlikely that new market
mechanisms would have the same efficiency as the TBA market.
While it is tempting to start from scratch, it is probably better
to preserve those aspects of the existing housing finance system that
have worked well and correct the flaws that contributed to the crisis.
Step-by-step transformation of the GSEs may be a less disruptive path
to reform. Adding private capital in front of the Government guarantee
through the use of subordinate bonds would be a good first step.