[House Hearing, 107 Congress]
[From the U.S. Government Printing Office]



                               before the

                      COMMITTEE ON SMALL BUSINESS
                        HOUSE OF REPRESENTATIVES

                      ONE HUNDRED SEVENTH CONGRESS

                             SECOND SESSION


                     WASHINGTON, DC, MARCH 13, 2002


                           Serial No. 107-47

         Printed for the use of the Committee on Small Business

78-663                       WASHINGTON : 2002
For Sale by the Superintendent of Documents, U.S. Government Printing Office
Internet: bookstore.gpo.gov  Phone: toll free (866) 512-1800; (202) 512-1800  
Fax: (202) 512-2250 Mail: Stop SSOP, Washington, DC 20402-0001

                      COMMITTEE ON SMALL BUSINESS

                  DONALD MANZULLO, Illinois, Chairman
LARRY COMBEST, Texas                 NYDIA M. VELAZQUEZ, New York
ROSCOE G. BARTLETT, Maryland             California
FRANK A. LoBIONDO, New Jersey        DANNY K. DAVIS, Illinois
SUE W. KELLY, New York               BILL PASCRELL, Jr., New Jersey
STEVE CHABOT, Ohio                   DONNA M. CHRISTENSEN, Virgin 
PATRICK J. TOOMEY, Pennsylvania          Islands
JIM DeMINT, South Carolina           ROBERT A. BRADY, Pennsylvania
JOHN R. THUNE, South Dakota          TOM UDALL, New Mexico
MIKE FERGUSON, New Jersey            CHARLES A. GONZALEZ, Texas
DARRELL E. ISSA, California          DAVID D. PHELPS, Illinois
SAM GRAVES, Missouri                 GRACE F. NAPOLITANO, California
EDWARD L. SCHROCK, Virginia          BRIAN BAIRD, Washington
FELIX J. GRUCCI, Jr., New York       MARK UDALL, Colorado
TODD W. AKIN, Missouri               JAMES R. LANGEVIN, Rhode Island
BILL SHUSTER, Pennsylvania           BRAD CARSON, Oklahoma
                                     ANIBAL ACEVEDO-VILA, Puerto Rico
                      Doug Thomas, Staff Director
                  Phil Eskeland, Deputy Staff Director
                  Michael Day, Minority Staff Director

                            C O N T E N T S

Hearing held on March 13, 2002...................................     1


Dorn, Hon. Nancy, Deputy Director, Office Of Management and 
  Budget.........................................................     3
Barreto, Hon. Hector, Administrator, U.S. Small Business 
  Administration.................................................     5
Crawford, Christopher, Executive Director, National Association 
  of Development Companies.......................................     8
Wilkinson, Anthony, President & CEO, The National Association of 
  Government Guaranteed Lenders..................................     9


Opening statements:
    Manzullo, Hon. Donald........................................    34
    Velazquez, Hon. Nydia........................................    36
Prepared statements:
    Dorn, Hon. Nancy.............................................    39
    Barreto, Hon. Hector.........................................    44
    Crawford, Christopher........................................    51
    Wilkinson, Anthony...........................................    61
Additional Information:
    Post hearing submission from Small Business Administration...    61



                       WEDNESDAY, MARCH 13, 2002

                          House of Representatives,
                               Committee on Small Business,
                                                    Washington, DC.
    The Committee met, pursuant to call, at 10:06 a.m. in room 
2360, Rayburn House Office Building, Hon. Donald Manzullo 
(Chairman of the Committee) presiding.
    Chairman Manzullo. Good morning and welcome to this hearing 
of the Committee on Small Business. A special welcome to those 
who have come some distance to participate and attend this 
hearing. A special welcome also to the Administrator. In case 
he leaves very, very quickly, which I have noticed, there is a 
third bambino on the way and that is where he belongs. If he 
does not leave quickly I will usher you out myself, Mr. 
Barreto, and take you to the hospital personally to get you 
    We are here today to discuss the subsidy rate calculations 
for the 7(a) and 504 loan programs. First of all I want to 
commend the Administration for recognizing the importance of 
this issue and beginning to correct the problem. We have seen 
more movement on this issue over the past eight months than 
over the past eight years. Thanks, in large part, to the 
personal involvement of Dr. Lloyd Blanchard who left OMB to 
join SBA as its Chief Operating Officer. Excellent choice and a 
good move. Good move for the SBA and for the OMB.
    The subsidy rate calculation has weighted loans originated 
by the preferred lenders program more heavily than other SBA 
loans which dropped the subsidy rate by 20 percent. I 
appreciate that change.
    The Administration is also committed to developing an 
econometric model for the next budget cycle which is it hoped 
will provide a more accurate forecast for the performance of 
the SBA loan portfolios.
    I also appreciate the willingness of the SBA to examine 
other alternatives to make sure that access to the 7(a) program 
is not cut in half next year.
    I look forward to working with the SBA as a constructive 
partner to see what we can do together.
    I wish we had an accurate subsidy rate calculation now so 
that there would be no question about the availability of the 
7(a) program in 2003. Fees could also be reduced in the 504 
program. Inaccurate subsidy costs results in overpayment of 
fees and eliminate flexibility of program delivery. I trust 
that we will not be in this predicament next year.
    Again, I commend the Administration for moving, I believe, 
very rapidly on this, staying in constant contact with our 
office. I look forward to this issue being resolved.
    I now yield for an opening statement by my good friend and 
colleague, the Ranking Member, Mrs. Velazquez of New York.
    [Chairman Manzullo's statement may be found in the 
    Ms. Velazquez. Thank you, Mr. Chairman.
    It should come as no surprise to anyone on this committee 
that access to capital is access to opportunity. It is the one 
tenet of business development that holds true regardless of 
time, place or industry. If entrepreneurs can't tap into 
credit, the difficult task of starting a business can become 
insurmountable. That is where the SBA loan program comes to the 
rescue. With competitive interest rates and a streamlined 
application process, these programs have been responsible for 
lending over $10 billion annually, providing 40 percent of the 
long term capital mix for small business.
    These lending programs bridge the gap between affordability 
and accessibility, creating a partnership with the private 
sector that is the government's best bargain investment for its 
    Unfortunately, a disconnect in both Democratic and 
Republican Administrations has limited the reach of these 
programs. We have the potential to help many more small 
businesses. This disconnect has translated into an over-charge 
for both lenders and borrowers of more than $1 billion to date. 
This is a fact, plain and simple, and it is the most blunt 
example of the arrogance of our government that I have seen in 
my service in Congress.
    Every year we raise this issue, yet all we hear are 
excuses. Either the model is not right or more data are needed. 
There is always an excuse and never a solution. What has been 
clear is that the OMB in both Democratic and Republican 
Administrations have been cooking the books in an attempt to 
hide this Stealth Tax from small businesses, and it taps our 
nation's entrepreneurs as a cash cow for the U.S. Treasury.
    Frustration with this complacency and negligence forced a 
coalition of lawmakers, including myself, to cut the fees for 
7(a) lending in half in order to open up more capital and 
reduce the cost to small businesses. This move presented the 
Administration with a golden opportunity to do the right thing 
and finally report an accurate subsidy rate. Unfortunately, the 
Administration again chose to play budget games with this 
year's budget proposal, jacking up the subsidy rate for the 
7(a) program to almost double the current level.
    The Bush Administration's excuse for increasing what they 
should have lowered this time was our fee reduction. But if the 
fee reduction is their excuse, what possible explanation can 
there be for the 504 program which also saw its fees increase 
without any substantive rate reduction? This is just one 
example of the inconsistencies that have plagued this debate.
    Another example is all the phony numbers that are floating 
around. In many cases the figures used to calculate the default 
rate so critical to the overall subsidy rates are laughable, 
and are even contradicted by the Administration's own budget. 
This junk math will taint almost any proposed solution.
    The latest excuse we will hear is that SBA has contracted a 
firm to help construct an econometric model to help more 
accurately predict true program performance. But given all the 
phony numbers flowing around, no equation will ever actually 
report the true cost of the program because if you put garbage 
in you most certainly will get garbage out.
    This is the time of reckoning. Either OMB and SBA report an 
accurate subsidy rate or we on the Committee working with other 
like-minded Members will do it ourselves, even if it means 
taking the money from OMB and SBA to make up the shortfall or 
making changes to credit reform or even if we have to legislate 
the subsidy rate every year.
    Yes, these are drastic actions, but this problem has gone 
on for too long to be left unresolved any longer. This change 
is especially critical at a time when our country is attempting 
to climb out of a recession. Small businesses make a crucial 
contribution to this effort. They hauled us out of the last 
recession into the greatest peacetime expansion on record. They 
will do it again with our help.
    So the question for this Committee and the Administration 
is, do we fix this problem and provide a $5 billion stimulus to 
this nation's small businesses to go out and create jobs? Or do 
we continue the draconian practice of taxing small businesses 
and keep them from providing the boost this economy so 
desperately needs?
    I know which way I am leaning, Mr. Chairman, and I am 
pretty sure about you too.
    Thank you very much.
    [Ms. Velazquez's statement may be found in the appendix.]
    Chairman Manzullo. Thank you.
    We will start with Nancy Dorn who is the Deputy Director of 
the Office of Management and Budget. Normally we would start 
with the Director, but this is the preference so that your 
testimony will sequence into the next, is that correct?
    Ms. Dorn. That's correct.
    Chairman Manzullo. I look forward to your testimony.
    All the statements of the witnesses and the Members of 
Congress will be made part of the record.

                    OF MANAGEMENT AND BUDGET

    Ms. Dorn. Thank you, Mr. Chairman, and good morning.
    Chairman Manzullo, Congresswoman Velazquez and Members of 
the Committee, I am Nancy Dorn, Deputy Director of the Office 
of Management and Budget. I am pleased to be here today to 
discuss OMB's role in implementing the Federal Credit Reform 
Act of 1990. It is in this capacity that OMB worked with SBA on 
cost estimations for loan programs.
    Before I begin let me say that it is unfortunate that the 
premature release of preliminary and incomplete estimates for 
SBA's 7(a) business loan program may have led to unrealized 
expectations. I would like to point out that our efforts to 
improve the accuracy of the subsidy rate calculation, the first 
such effort in over a decade, resulted in a 20 percent 
reduction in cost. The Administrator will explain in further 
detail the steps that OMB and SBA took to improve the 
methodology and the progress made in developing an econometric 
model for use in the fiscal 2004 budget.
    While the Administration recognizes the importance of SBA's 
lending programs, we feel that the best way to address the 
needs of small business is through sound tax and regulatory 
policy. Reductions in individual taxes benefit the 
approximately 20 million small businesses who pay taxes at 
individual income rates. In addition this Administration is 
committed to reducing the financial and administrative burden 
on small businesses, complying with tax, environment, health 
and labor regulations which again benefit the vast majority of 
the nation's 25 million small businesses.
    To address the lending needs of small businesses affected 
by the September 11th attacks, the Congress passed and the 
President signed legislation to temporarily lower fees for SBA 
lending programs and transfers from the rest of the government. 
While the fee reductions may help those businesses and lenders 
that participate, it also means that the subsidy rate and the 
cost of these programs rises. Given the additional costs, the 
budget supports a 7(a) loan volume that's lower than the 2002 
loan volume even with an increase in appropriations for the 
basic 7(a) program.
    With respect to the fiscal 2003 budget, the increase in the 
subsidy rate for the 7(a) program is not the result of any 
Administration proposal. However, we believe that there may be 
opportunities to leverage additional lending in other SBA 
programs to offset the decreased loan volume in the 7(a) 
    I would like to speak for just a minute about the Federal 
Credit Reform Act since that is one of the main reasons that 
OMB is involved in this process.
    As you know, the Federal Credit Reform Act became effective 
in fiscal year 1992 and it required federal agencies to 
accurately measure the true cost of lending programs by 
budgeting up front for the expected net loss to the government.
    Prior to that time the cost of federal loan programs was 
very uncertain and in many instances it took years for the 
costs to become apparent. Loan defaults and guarantee claim 
payments were reflected in the budget years after loan 
disbursements and guarantee commitments had been made.
    The cost of interest subsidies was clear only after several 
years of experience with market interest rates. The cost of 
direct loans was systematically over-estimated and the cost of 
loan guarantees systematically under-estimated. Credit reform, 
federal budgeting, and accounting for credit programs on an 
equal footing by requiring that the cost of loan programs be 
calculated and recorded up front in the years that the loan is 
made or guaranteed. It takes the best information available at 
a given point in time to measure the budget impact of federal 
loan programs. It uses the actual historical cash transactions 
of loan programs to compare the net present value of payments 
by the government with the net present value of receipts. This 
allows policymakers to make more informed decisions about 
credit programs and to compare more accurately the budget 
impact of credit programs with other federal expenditures.
    Consistent with the Budget Enforcement Act agencies must 
also use the technical assumptions derived at the time of the 
budget's release for budget execution which means that they 
must use the default and recovery rates, the discount rate, and 
other relevant assumptions as provided in the budget for that 
fiscal year. This protects the integrity of the analysis and 
protects the loan volumes provided by Congress by removing or 
changing the subsidy rate estimate during or after the 
appropriations process.
    Credit reform also requires that legislative changes to the 
subsidy rate be scored accordingly. Under credit reform 
agencies must update their subsidy estimates each year to 
reflect the accurate performance and experience in loan 
programs. Programs that experience higher costs than projected 
generate upward cost re-estimates and therefore require an 
additional mandatory appropriation from the Treasury to cover 
those losses.
    Programs experiencing lower costs and projections generate 
downward cost re-estimates and therefore return excess funds to 
the Treasury. These re-estimates do not require further action 
in the annual appropriations process.
    The Federal Credit Reform Act gives OMB the authority to 
make subsidy estimates. We have in most cases delegated that 
authority to the agencies but have retained the right to review 
and approve all estimates submitted by the agencies as part of 
their budget formulation and execution. We take this role very 
seriously at OMB. The staff works to ensure that estimates have 
been calculated in accordance with the requirements of the 
Federal Credit Reform Act and the applicable guidance issued by 
OMB, Treasury, the Financial Accounting Standards Advisory 
Board and other entities.
    A large part of this role is ensuring that consistent 
standards are applied. To help agencies meet these requirements 
OMB's developed a number of tools to estimate and re-estimate 
the cost of credit programs as well as spreadsheets to assist 
in the end-of-year calculations. We have alsoprovided detailed 
training on credit budgeting concepts and applications.
    Chairman Manzullo. How are you doing on time?
    Ms. Dorn. I am about out.
    Chairman Manzullo. Can you conclude there, or can you get 
    Ms. Dorn. Yes, sir. Let me just say that from our 
perspective we have a good working relationship with SBA. We 
have worked very closely historically on these subsidy 
estimates. We understand that the Congress is concerned about 
the current situation and we look forward to working both with 
the SBA and with the Congress to look for ways to alleviate the 
stress on the system.
    I look forward to answering any questions you may have.
    [Ms. Dorn's statement may be found in the appendix.]
    Chairman Manzullo. Thank you.
    Mr. Barreto.
    Mr. Barreto. Thank you.


    Mr. Barreto. Good morning, Mr. Chairman, Ranking Member 
Velazquez and Members of the Committee. Thank you for this 
opportunity to discuss the subsidy rates for the SBA's credit 
programs. I hope that my testimony and our discussion that will 
follow will clear up some of the misconceptions regarding the 
models and assumptions we use to calculate these rates.
    First of all, please allow me to explain----
    Chairman Manzullo. Administrator, excuse me. We just heard 
the call for the vote. My preference would be for us to go vote 
and then come back right away. Would that be okay with you?
    Mr. Barreto. That would be fine, sir.
    Chairman Manzullo. We will be back in about ten minutes. 
Thank you.
    [Off the record for a vote from 10:26 to 10:34 a.m.]
    Chairman Manzullo. Administrator Barreto, do you just want 
to start over?
    Mr. Barreto. Whatever you would like, sir.
    Chairman Manzullo. We got your name. Why do you not proceed 
from there. Thank you.
    Mr. Barreto. Thank you, Mr. Chairman, Ranking Member 
Velazquez, and Members of the Committee. Thank you for this 
opportunity to discuss the subsidy rates for the SBA's credit 
programs. I hope that my testimony and our discussion that will 
follow will clear up some of the misconceptions regarding the 
models and assumptions that we use to calculate these rates.
    First of all, please allow me to explain what a subsidy 
rate is. The subsidy rate represents the amount of 
appropriations necessary as a percent of total loan volume to 
cover the projected loan default for a cohort of loans made in 
a single fiscal year. It is calculated by finding the net 
present value of cash flow to and from government over the life 
of the loans made in that year.
    A simple equation that defines the subsidy rate would set 
the subsidy rate equal to cash outflows minus cash inflows. The 
loan defaults purchased by SBA represents the cash outflows. 
The cost is offset by the cash inflows which include up front 
fees from the borrower, ongoing fees from the lender, and 
whatever we can recover from the purchase default which has 
amounted to about half of all defaults.
    In general, higher fees and lower defaults reduce the 
subsidy rates and lower fees and higher defaults raise the 
subsidy fees.
    I understand your concern why the SBA's calculation of the 
subsidy rates for the 7(a) and 504 programs has seemed 
inconsistent with actual performance. Nevertheless the subsidy 
rates reflect SBA's average historical performance.
    The challenge we face, however, is looking into an 
uncertain future and predicting the average loan performance 
for the next cohort of loans for each year of life for that 
cohort. Doing so is akin to predicting the average test score 
performance for the next class of kindergartners for each year 
that they are in school all the way through high school, 12 
years into the future.
    One approach is to predict the 13 average annual scores by 
using the data from the past year's average scores in grades K-
12. This approach is essentially the same one we have been 
using to estimate loan performance. We take past performance 
from the last 16 years to predict how the next cohort of loans 
going forward for 15 to 20 years will perform.
    To focus more on the 7(a) program, Mr. Chairman, I realize 
that for nine of the past ten years we have experienced net 
downward re-estimates in the 7(a) program during what proved to 
be the most extraordinary period of economic expansion in our 
country's history. Just as the stock market grew faster than 
the underlying fundamentals warranted, this same phenomenon 
caused the SBA to similarly underestimate the performance of 
its loan portfolio.
    The issue in retrospect becomes whether the SBA should have 
followed the path of irrational exuberance in forecasting over 
that which Chairman Greenspan showed such concern or whether we 
should have stayed the course and continued to rely on 
fundamentals. I believe that the SBA followed the more prudent 
course by relying on fundamentals.
    From 1993 to the present fiscal year the subsidy rate for 
the 7(a) program was successfully reduced from 5.21 percent to 
1.07 percent based on historical experience.
    For fiscal year 2003, SBA executed a new interim 
calculation method which weights preferred lender loans more 
favorably than non-preferred lender loans with improved 
methodologies based on SBA's data to reliably delineate the 
default experience of the two programs.
    The subsidy rate was consequently lowered by one-fifth, all 
the way to .88 percent which at the President's $85 million 
request for 7(a) budget authority would have funded $9.7 
billion in lending.
    The President's plan was to provide record levels of 
lending to small businesses. However, P.L. 107-100 subsequently 
reduced the fees paid by borrowers and lenders for a two year 
period beginning in October 2002 causing the subsidy rate to 
double to 1.76 percent.
    Since this legislation doubled the cost of providing 7(a) 
loans the President's request of $85 million can only fund half 
as much in 7(a) lending. Nevertheless the SBA is continuing to 
create a more accurate method of calculating the subsidy rate.
    We work closely with the General Accounting Office and the 
Office of Management and Budget to determine the best way to 
incorporate historical data in calculating subsidy rates. While 
GAO ultimately disagreed with the interim method chosen by SBA 
and OMB, both GAO and OMB agreed with SBA that the econometric 
approach is the best method to use.
    We all agreed that a sound methodology adequately addresses 
economic fluctuations and programmatic changes while remaining 
unbiased, producing subsidy estimates that are just as likely 
to be revised upward as they are downward over a long term 
period of time. An econometric model is a more sophisticated 
weighting method than the PLP weighting model as it creates a 
way for each relevant factor and estimates the subsidy rate 
based on the relative strength of these weights.
    I am confident that our implementation of the econometric 
model in fiscal year 2004 willresolve many of the concerns you 
have. It is a proven model. As I stated, OMB and GAO both agree that it 
is the most appropriate method to use and other federal agencies, 
including the Federal Housing Administration and the U.S. Department of 
Education already use it to calculate the subsidy rates for their 
credit programs.
    To implement the model the SBA has contracted with the 
Office of Federal Housing Enterprise Oversight or OFHEO, an 
agency which uses such a model to monitor the housing credit 
markets and the performance of mortgage loans.
    Mr. Chairman, Ranking Member and Members of the Committee, 
thank you for the opportunity to appear here today. I will be 
happy to answer any of your questions.
    Thank you.
    [Mr. Barreto's statement may be found in the appendix.]
    Chairman Manzullo. Thank you.
    Our next witness is Christopher Crawford who is the 
Executive Director of the National Association of Development 
Companies. We look forward to your testimony, Mr. Crawford.


    Mr. Crawford. Good morning. I am pleased to again come 
before the Committee to comment on the SBA 2003 budget model 
and subsidy assumptions. I would like to thank you Chairman 
Manzullo, Ranking Member Velazquez and the Committee for 
convening this hearing.
    The 2003 budget adds to the injustices of the past budgets 
by again forcing our borrowers to pay millions of dollars in 
fees to the U.S. Treasury. It appears it will take this 
Committee's oversight to motivate the Administration to stop 
using these loan programs to cover deficits.
    This budget increases the annual fee charged each 504 small 
business borrowers from .410 to .425 percent. The program is 
supposed to pay for itself through these fees, and it does this 
and far more.
    Since we went to zero subsidy in 1997 borrowers have paid 
the Treasury $400 million in excess fees and interest according 
to SBA's own re-estimates. Even paying fees over and above 
these inflated cost estimates the Administration demands still 
more from small business by increasing the fee for 2003. This 
is truly an unwarranted tax on our borrowers.
    This fee increase is caused by errors we believe SBA and 
OMB are making in their assumptions and their forecasts.
    First, the estimate of loan defaults is 8.3 percent. 
Attached to my statement that you have been provided are charts 
that show loan defaults for the past 12 years are nowhere near 
eight percent.
    The President's own budget supports our estimates, not 
SBA's forecast.
    Page 49 of the budget request and performance plan admits 
the true defaults are running $60 to $70 million annually, at 
an annual volume of $2 billion or more defaults are averaging 
under five percent.
    Second, SBA loan recovery forecasts do not match actual 
performance through the asset sales or our CDC liquidation 
program. They forecast collection of 58 cents of every dollar 
on defaulted loans. However, SBA also forecasts they will spend 
38 cents to make that recovery leaving a net recovery of only 
20 cents of each dollar.
    Our CDC liquidation program has averaged 55 percent 
recovery since it began some three years ago. At the same time 
SBA states they are recovering 50 percent for loans sold 
through their assets sales.
    I have to ask you, where did the money go? It did not go 
into the recovery expenses for either the liquidation program 
or the asset sales since neither program has substantial 
operating costs.
    Third, we believe SBA is using erroneous forecasts of 
future loan prepayments. We are funding 504 loans today at a 
record low market rate, often below six percent. Even with 
servicing and SBA fees added on, borrowers will obtain long 
term rates that they are not likely to want to pre-pay in 
future years. I believe pre-payments will in fact go down, not 
increase for this cohort.
    Finally, even the SBA recognizes that they are not doing a 
good job of forecasting our portfolio performance yet their 
solution is to change the subsidy models rather than improve 
their data analysis. They plan to spend the next five years 
building and debugging a new econometric model that was 
actually created to reflect housing costs, not commercial 
lending costs. This process will consume time, money and 
resources to get it right and give the Administration further 
excuses for continued subsidy errors.
    Our subsidy problems have led to inflated fees and has made 
504 truly a Treasury cash cow as Ms. Velazquez pointed out. 
Borrowers are paying hundreds of millions of dollars in excess 
fees and are now told that they are going to pay even more for 
the 2003 cohorts.
    We strongly object to this budget proposal and we need your 
help. I ask this Committee to get to the bottom of the 
Administration's questionable assumptions. Without your 
intervention I fear that our small business borrowers will 
continue to pay excessive fees for years to come.
    Thank you for this opportunity to provide comments.
    [Mr. Crawford's statement may be found in the appendix.]
    Chairman Manzullo. Thank you.
    Our last witness is Anthony Wilkinson, President and CEO of 
The National Association of Government Guaranteed Lenders.
    Mr. Wilkinson.

                    GUARANTEED LENDERS, INC.

    Mr. Wilkinson. Thank you, Mr. Chairman, Ms. Velazquez, and 
thank you for holding today's hearing.
    As we all know, users of SBA loan programs have been 
substantially overcharged the last decade. In the case of the 
7(a) program OMB admits through the budget process that the 
overcharge or the downward re-estimate in the 7(a) program 
totals $1.429 billion. I have listed these downed re-estimate 
amounts on page one of my written testimony.
    On page two of my written testimony is another page that 
discloses that OMB has not been forthcoming in the re-estimate 
process. This shows that OMB requires the use of rates in the 
re-estimate process that are higher than what is statistically 
expected. Had OMB used the expected default rates in the re-
estimate process they would recognize an additional $400 to 
$600 million in downward re-estimates in the 7(a) program. This 
would bring the total overcharge to approximately $2 billion 
and it was $2 billion that OMB's representative admitted to at 
the Senate Roundtable discussion that was held last September. 
That is $2 billion since the start of credit reform in 1992, or 
on average a $200 million per year overcharge.
    Now let us take a look at the current year. OMB is 
estimating defaults at 12.73 percent, a number that has not 
been seen over the last decade. We know from the Senate 
Roundtable lastyear that SBA had recommended using a five year 
look-back period, a method that SBA believes would have been more 
predictive of near term loan performance. OMB rejected this, and OMB 
rejected each of the other six methods that were proposed, deciding to 
stick with the method that provided the highest default estimate and 
hence the highest costs to 7(a) program users.
    In a previous hearing before this Committee SBA testified 
that the default rate for the 7(a) program was being managed in 
the eight to ten percent range. If OMB were to use a more 
realistic default assumption for fiscal 2003 we believe the 
subsidy rate would fall by approximately 100 basis points or 
one percent. That would leave a 7(a) subsidy rate of 
approximately .76 for fiscal 2003 and we would need 
appropriations of about $91 million to meet the anticipated 
demand for next year.
    Now considering the $2 billion that has been overcharged, 
users of the 7(a) program have already sent enough money to the 
Treasury to fund the next 20 years worth of 7(a) loans. That is 
just not reasonable.
    Mr. Chairman, NAGGL has repeatedly said that we do not care 
what model OMB uses in calculating subsidy rates but we do 
believe the resulting calculation should be fair and 
reasonable. We have never demanded that it be totally accurate, 
just reasonable.
    We believe that there should be an equal likelihood that 
the re-estimates could be revised upward as well as downward.
    It is obvious that the calculation to date has not been 
fair or reasonable. OMB has been levying taxes on 7(a) program 
users. We start fiscal 2003 with a default estimate of 12.73 
and we know we are going to have a down re-estimate as soon as 
the year ends. The same thing is going to happen for this 
current fiscal year because they used a default estimate of 
13.87, so we will have a downward re-estimate.
    I direct your attention to the subsidy rate re-estimate for 
loans that were originated just in fiscal year 2001. OMB says 
that the subsidy rate has already fallen by 40 percent, from 
1.17 to .71 during the fiscal year. Many of the loans approved 
in fiscal year 2001 had not even been disbursed by the time 
they figured out they have over-estimated the cost of loans 
made in fiscal 2001.
    In Mr. Barreto's testimony today he cited that all but one 
of the years there had been downward re-estimates in the 7(a) 
program. The very first re-estimate in 1995 was an upward one. 
OMB used this upward re-estimate to pitch the need for higher 
fees in 1996. Unfortunately, we had to go along so in 1996 the 
7(a) program incurred a huge fee increase, one that we know now 
was not needed. As a result of the 1996 fee increases many 
lenders over time exited the program. Fortunately Congress saw 
how OMB was taxing 7(a) program users and you passed fee 
reduction legislation last year. That legislation eliminated 
only about half of the 1996 fee increase that was not needed. 
Hopefully we can consider further cost reductions and hopefully 
get the ratio of downward re-estimates to appropriations down 
to ten years rather than the current 20.
    Lastly I would like to share our concerns over the move to 
the so-called econometric model. The current model used by SBA 
is a simple net present value cash flow model. It is fairly 
easy to understand. But the 7(a) subsidy calculation is not 
driven by the model, it is driven by the assumptions plugged 
into the model. OMB could have easily had a subsidy calculation 
that was fair and reasonable simply by adjusting assumptions. 
They had recommendations from SBA and GAO and they chose to 
ignore them.
    We believe an econometric model will be much harder to 
understand, one; and two, will have many more factors that will 
be driven by OMB assumptions.
    To put it in simpler terms, we have cracked the old black 
box and figured out what OMB was doing. But now they plan to 
build a new black box so they can continue to overcharge users 
of the program.
    Mr. Chairman, OMB has not been held accountable for the 
decisions they make in the subsidy calculation process and 
until they are we fear the overcharges will continue. OMB had 
every reason to adjust its assumption in the fiscal year 2003 
budget. There were mandates from both the Senate and House 
Small Business Committees, Republicans and Democrats alike. 
There was report language in the fiscal 2002 Treasury Postal 
Appropriation Bill and I can go on and on. Let me just say they 
chose to ignore it all.
    Instead OMB tries to shift the focus of the discussion away 
from them and talk about how Congress is to blame for the 
fiscal year 2003 budget shortfall and how we should all wait 
another year while they build their new black box.
    Mr. Chairman and Members of the Committee, SBA program 
users need your help. It is clear they have been taxed by OMB. 
It is clear that OMB has not had any desire to fix this 
problem. They have made mockery of the Federal Credit Reform 
Act and as a result have made the appropriation process very 
    It is time, this fiscal year, right now, for a solution. We 
believe OMB accountability has got to be a big part of that 
    Thank you, Mr. Chairman.
    [Mr. Wilkinson's statement may be found in the appendix.]
    Chairman Manzullo. Thank you all for your testimony.
    I have just a couple of questions of Nancy Dorn.
    I do not know if I heard this right, but maybe it is 
reflected on page two of your statement under, see where it 
says federal credit reform about a third of the way down, the 
    Ms. Dorn. Yes, sir.
    Chairman Manzullo. It says the Federal Reform Act--Wait a 
second. I'm sorry. I am on the first page.
    The first page, second full paragraph. `Before I begin'.
    Ms. Dorn. Right.
    Chairman Manzullo. `Before I begin let me say it is 
unfortunate that the premature release of preliminary and 
incomplete estimates for SBA's 7(a) business loan program may 
have led to realized expectations.' Unrealized expectations.
    Can you explain what you mean by that? I guess first I have 
to read it right, but can you explain what you mean by that?
    Ms. Dorn. Mr. Chairman, I think there was some confusion in 
the last year. This was certainly before I came to OMB, but I 
think there was some confusion last year about what the 
calculations that SBA and OMB were going to lead to in the 
subsidy rate. I think there was some thought that we were going 
to reduce the subsidy rates but that the reduction would be 
greater than what it turned out to be.
    Chairman Manzullo. What caused that? Mr. Blanchard, were 
you familiar with what happened there? Did you want to comment 
on that?
    Mr. Blanchard. Yes, sir.
    Chairman Manzullo. That was before Mrs. Dorn's tenure.
    Mr. Blanchard. Yes, sir, Mr. Chairman. I am familiar with 
what happened.
    The staff members of the House and the Senate Small 
Business Committees were eager to learn what the subsidy rate 
would be using this new model for 2003 before we 
actuallycompleted the budget and the estimation of this rate.
    At the time, sir, I was at OMB working on this particular 
issue and----
    Chairman Manzullo. You briefed us in our office on this, 
    Mr. Blanchard. Yes, sir.
    I resisted providing a point estimate of what that would be 
because we simply did not know. We had preliminary analysis. We 
had done that to try to see whether this model, this new 
interim model would actually perform as well as we think it is 
performing. But at the time I was being pushed by staff members 
to provide some kind of estimate just so they could have a 
sense of where it would be.
    In my haste I gave them a range of estimates just to 
satisfy them, really trying to be forthcoming with information 
and to work closely with our legislative partners in providing 
that range, sir. That was then taken as fact. They sort of 
banked on that, so to speak. I think that banking led to some 
preliminary calculations for the Senate Bill 1196 that----
    Chairman Manzullo. Did that have anything to do with the 
estimated amount of so-called budget surplus?
    Mr. Blanchard. I am not sure what you are referring to with 
regard to the budget surplus.
    Chairman Manzullo. In other words, when the subsidy rate is 
set, or calculated, does that have anything to do with the 
general gross revenues that come into the government?
    Mr. Blanchard. The subsidy rate does indeed have something 
to do with that--As Administrator Barreto mentioned earlier in 
his testimony, the subsidy rate is essentially a prediction of 
the appropriations that are necessary to cover defaults. 
Without knowing what those defaults would be, we simply make a 
prediction. As time goes on we learn more about that default 
behavior and then we adjust what is held in the account to 
cover those defaults. So it adjusts upwards and downwards. 
Unfortunately the adjustments have been such that it has 
created this surplus that has gone back to the Treasury over--
    Chairman Manzullo. You answered about 80 percent of the 
question. Let me follow up and maybe get more particular.
    Does the estimated amount of surplus in general revenue of 
the government have anything to do with calculating the subsidy 
    Mr. Blanchard. No, sir. Not at all.
    Chairman Manzullo. Totally separate.
    Mr. Blanchard. Yes, sir.
    Chairman Manzullo. The reference that Ms. Dorn made to the 
premature release of preliminary and incomplete estimates for 
SBA's 7(a) business loan programs, that was reference, Dr. 
Blanchard, to a September 5th Roundtable that you had had with 
the Senate?
    Mr. Blanchard. No, sir. Actually that was subsequent to 
that Roundtable.
    Chairman Manzullo. Because at that Roundtable, that is 
where you had said the history of this program is one that has 
had an unfortunate one. `The Administration has worked in its 
first year to correct this problem and it is one that we 
inherited that as you have all mentioned is a serious problem. 
WE recognize that over the past 10 to 12 years there is a 
cumulative $2 billion that has gone back to the Treasury.'
    The last part of that is correct, is it not? The 
    Mr. Blanchard. I actually do not think that is correct. I 
believe it is $1.4 billion as Mr. Wilkinson referred to in his 
    Chairman Manzullo. $1.4 as opposed to $2 billion.
    Mr. Blanchard. Correct.
    Chairman Manzullo. Because you did not have the figures in 
front of you at the time and I can understand that. That must 
have been the case.
    Mr. Wilkinson. I would add on to that, there was a sheet in 
my testimony called Unweighted Purchase Rates that goes through 
the rate they use in the re-estimate process, yet the rate that 
they expect in terms of defaults are much lower than they are 
using in the re-estimate process.
    So while $1.429 billion has been recognized, it is 
anticipated there is going to be another $400 to $600 million 
that will be downward estimated in the future.
    Chairman Manzullo. Thank you.
    Let me give you a very simplistic situation here. I am not 
into macroeconomics, I am into microeconomics. If you are the 
one that loses your job you could care less about whether or 
not the rest of the neighborhood is working.
    Let us take a look at a loan that goes to A Corporation, 
and let us say this loan is $100,000. Figure for me, and you do 
not have to give the exact figure, but work through with me how 
you would figure the subsidy rate just on this one loan. 
However you want to do it. Explain the elements and kind of 
give me a rough idea.
    Mr. Blanchard. Okay.
    The subsidy rate is calculated before loans are made. What 
we are trying to do is estimate for fiscal year 2003, when we 
make that $100,000 loan and many more of those loans, what will 
be the default rate?
    Chairman Manzullo. You want to set aside a pocket of money 
to cover what could be a loss.
    Mr. Blanchard. That is right.
    Chairman Manzullo. That is really what the subsidy rate is. 
Okay, go ahead.
    Mr. Blanchard. That is right.
    Based on the historical performance of loans of similar 
sorts, we estimate in advance what that default would be. Not 
only in the first year of that loan, but in the first, second, 
third, and throughout the life of the loan.
    Chairman Manzullo. With regard to the 7(a) and the 504 
programs, are there different categories of default? Is it 
based upon the nature of a business or is everything lumped 
together in the 7(a) and the 504?
    Mr. Blanchard. As it has been calculated in the past up to 
FY 03, all 7(a) loans have been treated the same in that 
calculation and all the 504 loans have been treated the same in 
that calculation.
    Chairman Manzullo. Then on the first element, the default 
rate, I am trying to find out where there are areas of 
agreement and where there are areas of disagreement, and maybe 
that is where we can hone in.
    Is there disagreement as to the default rate that occurs 
with these loans?
    Mr. Blanchard. Yes.
    Chairman Manzullo. Not a very good start, is there.
    Dr. Blanchard or Administrator Barreto, however you want to 
do it. Let us take it element by element and let us see----
    Mr. Blanchard. We disagree in what the default assumption 
should be, sir.
    Chairman Manzullo. Okay. Walk me through that.
    Mr. Blanchard. Our estimate of the default rate is based on 
the average performance of all of the loans that have been made 
from 1986 to the present date, so data up through 2001.
    Chairman Manzullo. Mr. Wilkinson, what is yours based on?
    Mr. Wilkinson. The performance of the program since----
    Chairman Manzullo. What do you think it should be based on, 
    Mr. Wilkinson. Since the start of credit reform, when 
credit reform started there were substantial programmatic 
changes. The program that existed in 1986 is not the program 
that exists today.
    Chairman Manzullo. Why is that?
    Mr. Wilkinson. Lenders have a much greater investment in 
the loans today, the guaranteed percentages have declined, the 
lenders have to pay an ongoing fee that was not there in 1986. 
So there is a much greater financial interest of the lender in 
the loan program today.
    Chairman Manzullo. Do you agree with that, Dr. Blanchard?
    Mr. Blanchard. I do agree that the program has changed, 
which is precisely why we are weighting the PLP loans 
differently than the non-PLP loans for fiscal year 2003.
    Mr. Wilkinson. The default rate that they attach to PLP 
loans is higher than any default rate we have seen since 1991. 
They are using 11 point, it is over 11 percent default for PLP 
in the weightings they are using this year and anticipated 
defaults in the 2000 cohort is 8.1.
    Chairman Manzullo. So Dr. Blanchard you are going back to 
1986 and you are averaging all the loans since that time to 
come up with the----
    Mr. Blanchard. That is correct.
    Chairman Manzullo. You are saying that about 11 percent of 
the loans that default or 11 percent of the amount of money 
loaned has been in default?
    Mr. Blanchard. We are saying 11 percent of the SBA's 
portion of the outstanding----
    Chairman Manzullo. Okay. It is based on money, not the 
numbers of the loans.
    Mr. Blanchard. That is correct.
    Chairman Manzullo. Mr. Wilkinson, what is wrong with that 
    Mr. Wilkinson. First of all they do a simple average. There 
were some----
    Chairman Manzullo. I prefer to keep things simple. That is 
why I am going through this step by step. But what is wrong 
with simple average?
    Mr. Wilkinson. First of all, it is an arbitrary point in 
time. They picked in 1986 because that is where they start with 
their data.
    Chairman Manzullo. Dr. Blanchard, he said that 1986 is an 
arbitrary date.
    Mr. Blanchard. Sir, we have a chart that can help 
illuminate it a little more easily.
    Chairman Manzullo. That would be fine.
    Mr. Blanchard. This takes us down into averaging. Let me 
set up the chart.
    Chairman Manzullo. Mr. Barreto, I did not mean to ignore 
you, I just left it up to you as to who would answer the 
    Mr. Barreto. I brought my expert this time.
    Chairman Manzullo. That is fine.
    Mr. Blanchard. Mr. Chairman, we thought that you might want 
to know precisely how this model works, so we built a graphic 
that walks you through the cash flow model.
    Now it is important to recognize, sir, that the projection 
of the default rate is something that we do outside of this 
model. As our industry partners mentioned earlier, the main 
assumption is the default rate.
    Without knowing what the default rate will be in the 
future, the best evidence we have is that of the past. What we 
are going to do in the future with regard to our calculation 
method, the econometric model will calculate the default rate. 
What you do is separate the estimation of the default from the 
estimation of the subsidy rate. The default is an assumption 
that drives the subsidy rate.
    Chairman Manzullo. The reason I went through that is I 
thought I would be a nice guy and try to find certain areas 
where people agree on getting to this.
    Before you do that, Dr. Blanchard, is it the default rate 
that is the most contentious of the four items in here?
    Mr. Blanchard. Yes, sir.
    Chairman Manzullo. Then I guess I started with the right 
    Mr. Blanchard. You sure did.
    Chairman Manzullo. How do you want to do that?
    Mr. Blanchard. As you see in front of you there are 
basically three components to that. Obviously taxpayers pay 
taxes into the Treasury so that is an important component on 
the back end. But the outflows, remember that the basic 
equation here is that the subsidy rate equals the cash that 
goes out minus the cash that comes back in. The cash that goes 
out, the only cash that goes out in these loan programs are our 
purchase of the defaulted loans.
    This number says that our expected default rate is 8.2 
percent. This is of all outstanding loans. This number is 
equivalent to the 12 percent assumption that Tony mentioned 
that is a proportion of SBA's guaranteed portion. I do not want 
to confuse anybody. But that sets the outflows.
    Then we collect fees from the borrowers, we collect fees 
from the lenders, and then we recover some of those defaulted 
loans. You see the arrows point to that outflow and inflow. 
That then tells us how much money we have to set aside in the 
form of appropriations to pay for the rest of the program. That 
is all this model shows you, sir.
    You see that the arrows and the numbers that are attached 
to them are also maps to the equation at the bottom of that 
    Chairman Manzullo. Mr. Wilkinson, do you have a problem 
with this chart? Or just with the assumption of the default 
    Mr. Wilkinson. It is the assumption of the default rate. 
The arrows are right. That is how the cash flows in and how the 
cash flows out.
    Chairman Manzullo. Okay.
    Tell me again, why you think that? What Dr. Blanchard is 
telling us is that they go back to 1986 and take the average 
default rate of the amount of money SBA puts out.
    Mr. Wilkinson. It is a simple average going back to 1986. 
It ignores the program changes that have occurred since 1986.
    They have testified that this program is being managed in--
    Chairman Manzullo. What are those program changes since 
1986 that you think should do away with simple averaging?
    Mr. Wilkinson. Fees have gone up, lenders have seen the 
guarantee percentage decline. Remember, there was a point in 
time there was 90 percent guarantee. The average guarantee 
since 1986 has gone down.
    So again the lender has more at risk in each of these 
    Chairman Manzullo. Let me stop you right there.
    Dr. Blanchard, those two elements he just put in, why 
should they not be figured into the simple averaging?
    Mr. Blanchard. Before the calculation for the model in 2003 
there was no incorporation of the programmatic changes. He is 
correct in that regard. That is why we have gone to this method 
of weighting the PLP loans, which is the programmatic change 
that he is referring to, itis different than the way we weigh 
the non-PLP loans.
    Chairman Manzullo. So are you agreeing on the subsidy rate 
for what OMB wants to do for 2004?
    Mr. Wilkinson. I do not have a clue what they are doing for 
2004. There has been no----
    Chairman Manzullo. 2003, I am sorry.
    Mr. Wilkinson. Am I agreeing to that subsidy rate? 
Absolutely not.
    Chairman Manzullo. Dr. Blanchard just said that before 2003 
they did not take into consideration programmatic changes, but 
after that date they do take that into consideration.
    Mr. Wilkinson. In my opinion the change they made for 
fiscal 2003 is simple window dressing.
    It made a minor change but did not address----
    Chairman Manzullo. Yes, but he said that they take it into 
consideration in 2003. So there has been apparently a major 
change in the manner of the calculations. Would you disagree 
with that?
    Mr. Wilkinson. I disagree. There has not been a major 
    Chairman Manzullo. Dr. Blanchard, is there a document that 
has been generated that shows that those elements that Mr. 
Wilkinson says are the program changes? Is there a document 
that shows that those have been taken into consideration for 
    Mr. Blanchard. Yes, sir. That document is the President's 
    Chairman Manzullo. Okay.
    Do you agree with that, Mr. Wilkinson?
    Mr. Wilkinson. I will agree that they made some changes in 
2003 as it was from 2002, but the changes are very minor.
    Chairman Manzullo. Can you show us that section from the 
budget? Does anybody have it handy there?
    Mr. Blanchard. What we did, I do not have it handy, sir, 
but I can tell you that in fiscal year 2002 the subsidy rate 
was 1.07 percent. Our calculation in implementing this new 
method that weights the PLP loans differently from the non-PLP, 
reduced that subsidy rate to .88 percent. That is a one-fifth 
reduction in the subsidy rate, sir. That is about a 19 basis 
point reduction in the subsidy rate. That is how significant 
this model reduced that subsidy rate.
    Chairman Manzullo. Do you agree that it is a reduction, Mr. 
    Mr. Wilkinson. First of all, the 1.07 was too high. A 19 
basis point shift when there has been $1.4 billion----
    Chairman Manzullo. He is saying it is going in the right 
    Mr. Wilkinson. It has been going in the right direction for 
a decade. Very slowly. And we know that if they start the year 
with a 12.73 percent default rate we are going to have a 
downward re- estimate at the end of the year. WE are going to 
have one this year.
    Chairman Manzullo. So you are saying that even though it is 
a reduction, that the reduction still is not----
    Mr. Wilkinson. WE would have to have----
    Chairman Manzullo [continuing]. Default range.
    Mr. Wilkinson [continuing] A major downward turn in the 
performance of 7(a) loans to come close to the default estimate 
in the model. That to me is the real issue.
    Chairman Manzullo. Would you agree with that, Dr. 
Blanchard? This is sort of interesting, going back and forth 
here, but----
    Mr. Blanchard. A hyperbole of that statement I can accept, 
but history has shown that we will not have another year like 
all of the extraordinary years that we had in the late 1990s.
    Chairman Manzullo. Okay.
    Ms. Velazquez. And even with all those extraordinary years 
that you had in the late 1990s you did not do the right thing 
in terms of fixing the subsidy rate.
    So Mr. Barreto, I believe we need to get one thing 
straight, and that is the notion which the Administration keeps 
saying that somehow this mess we are in is due to congressional 
actions aimed at giving small businesses some relief from being 
overcharged by program fees.
    So that we are clear, it is our job, Congress' job, to 
    Based on the past record of the programs that participants 
were being grossly overcharged, we legislated a program change.
    It is the Administration's job to implement that change. So 
you knew what the subsidy rate will be and you choose to fund 
the program at half this level.
    So if you are looking to place the blame somewhere, look no 
further than the table where you and Ms. Dorn sit, because it 
lies right there.
    So my question to you is, you have this legislated change 
for three months before the budget was released. Why do you 
choose not to go back and fix it?
    Mr. Barreto. Thank you, Mrs. Velazquez for the question. I 
appreciate it.
    At no time are we trying to affix blame or point the 
finger. I think one of the things we are trying to do is 
    There is this misperception out there that the President 
and the Administration chose to cut the 7(a) program in half 
and as we----
    Ms. Velazquez. But that is exactly what you did.
    Mr. Barreto. But as we indicated before, with the subsidy 
rate going down to the .88, which we were very happy about 
being able to get down to .88, with the current President's 
budget, that would have been able to provide for a $9.7 
    Ms. Velazquez. Excuse me. We changed the fee. You did not. 
We did it.
    Mr. Barreto. Yes.
    Ms. Velazquez. And then you chose.
    Mr. Barreto. The budget had been completed before the time 
this legislation actually went into effect. But again, what we 
are trying to do is clarify the record to reflect that our 
intention was to have $9.7 billion available for 7(a) lending.
    Ms. Velazquez. Mr. Barreto, you could spend the whole 
morning here giving me a nice explanation but you do not answer 
my question.
    My question is why you, if you had three months to fix the 
subsidy rate, you did not.
    Ms. Dorn, why did you not?
    Ms. Dorn. I think the point that we would make is three are 
sort of three components to this. One is the establishment of 
the subsidy rate which is based on criteria. There has been 
some discussion about the kind of criteria that goes into the 
calculation. You look at the history of the program, you look 
at the actual loans that have been defaulted on, and you, as 
was pointed out, we have actually taken into account the fact 
that the preferred lenders' default rate has been less. That 
was cranked into the formulation.
    So for over a year we established a subsidy rate, and that 
is a separate exercise from either the establishment of the 
fee, which you point out Congress changed last year. And then 
the issue of the appropriation that we put against the program.
    Ms. Velazquez. Let us move to the next question.
    I am sure you have heard the accusations that OMB is using 
phony numbers to inflate the loan program subsidy rate. How do 
you respond to that?
    Ms. Dorn. Well Congresswoman, I understand the frustration 
that people feel when they see the history of this program, and 
I looked at the numbers of the loan subsidy rate and the re-
estimates over the last years going back to fiscal year 1992 
and I can say the program is not perfect.
    You can see by looking at the re-estimates that over the 
last several years the margins are falling and they are falling 
    For example in 1992 they missed it by $200 million. In 
fiscal year 2000 we missed it by $2 million. So there is some 
refinement going on in how we calculate the subsidy rate, and I 
appreciate the frustration that people fee.
    I think we are making some progress.
    Ms. Velazquez. Let me ask this question. Since the late 
1990s SBA has engaged in the practice of selling defaulted 504 
loans through asset sales. The track record for this has been a 
net recovery rate of approximately 50 percent. However in the 
Administration's recent budget you include in that recovery an 
assumption of just 20 percent. Where did the 30 percent go and 
how do you justify it?
    Ms. Dorn. I will have to turn to Hector for that.
    Mr. Barreto. On the asset sales that we do, it includes 
    Ms. Velazquez. Excuse me, Ms. Dorn. It was not SBA, it was 
OMB that established the 20 percent. The 30 percent. It is not 
SBA, it is OMB. Is not OMB the one saying it?
    Ms. Dorn. Well, OMB is indeed the place where the budget 
comes from, but we work very closely with----
    Ms. Velazquez. Mr. Crawford, could you please tell me if 
this does not sound like a phony number to you?
    Mr. Crawford. Yes, ma'am. I would have to agree with you 
that the numbers are certainly strange.
    If you look at the growth collection rate it is running 
downward from 67 to 58 percent. The expense rates are running 
nearly 40 percent. They are recovering a net of 20 percent.
    If you look at the asset sales, there have been five asset 
sales to my understanding. They have sold a total of about 1300 
loans that have gone bad. In our entire program since 1986 they 
sold nearly 900 notes related to those 1300 loans and their own 
numbers are a net recovery of 50 percent.
    So I do not understand how you get from 50 percent to 20 
percent. What happened to the other 30 percent?
    Ms. Velazquez. Ms. Dorn, if we were to assume a 50 percent 
recovery rate, maybe he would be able to answer this question 
when she finishes.
    If we were to assume a 50 percent recovery rate that has 
been the historical trend, what would the subsidy rate be?
    Ms. Dorn. I am sorry, Congresswoman, I do not know the 
answer to that question.
    Ms. Velazquez. Do you have the answer to that question?
    Mr. Barreto. What I would like to do is give Dr. Blanchard 
another opportunity to explain that. It is a very important 
question in front of us right now about how that is determined 
and what happened to that other 30 percent as you mentioned.
    Mr. Blanchard. Ms. Velazquez, Mr. Crawford's point about 
asset sales, and incorporating our 50 percent return on asset 
sales into recovery for the 504 program, we are comparing 
apples to oranges.
    He is talking about the recovery for one loan program. The 
asset sales program engages in sales of products or assets from 
many of our loan programs. So that 50 percent recovery rate is 
a combined recovery rate from multiple loans of which 504 
loans, defaulted 504 loans are only a small percentage.
    Ms. Velazquez. Can you tell me what the rate would be if 
you have a recall rate of 50 percent?
    Mr. Blanchard. No, I cannot tell you that right off the----
    Ms. Velazquez. Mr. Crawford.
    Mr. Crawford. It looks to me like it would probably go, it 
would probably cut the borrower fee by probably close to 60 
percent which would lead to a borrower fee probably in the .25 
down from .42. So it would be a pretty substantial cut in the 
actual fee.
    I would like to make a statement regarding Lloyd's comments 
about the asset sales.
    It is my understanding that each of the asset sales 
consisted of a significant number of pools, each of which had 
homogeneous assets in those pools. In other words they did not 
want to sell short term notes and long term notes because Wall 
Street is not going to pay a lot of money for that kind of a 
    Wall Street is going to look for a pool of loans or notes 
or collateral that is fairly consistent, and it is my 
understanding that our real estate notes as well as the real 
estate notes of 7(a) were put into similar pools and a fairly 
discreet number of pools.
    Now the number that I use, the 50 percent, I was told by an 
SBA manager that that was in fact the recovery rate, the net 
recovery rate for 504 specifically, not for 7(a), not for 
disaster loans but 504.
    Ms. Velazquez. Thank you.
    Ms. Dorn, last year when Mr. Blanchard testified before the 
Senate, he stated that once the President's budget proposal is 
submitted to Congress the assumptions contained in the budget 
cannot be changed for subsidy rate purposes. Is that your 
position as well?
    Ms. Dorn. Congresswoman, the Federal Credit Reform Act 
specifically directs that that is the case.
    Ms. Velazquez. What is the 7(a) default rate assumption 
used in determining the program fiscal year 2003 subsidy rate?
    Ms. Dorn. Congresswoman, as I think we previously 
testified, we used the subsidy rate of .88 for the 2003 budget.
    Ms. Velazquez. It wasn't 12.73?
    Mr. Barreto. For the default rate?
    Ms. Velazquez. The default rate.
    What is the subsidy rate?
    Ms. Dorn. The subsidy rate is .88. I'm sorry, 
    Ms. Velazquez. So the default rate for the----
    Ms. Dorn. I am sorry, I have been corrected. The subsidy 
rate is 1.76 in 2003 I am told.
    Ms. Velazquez. So the default rate is 12.73. Is that 
    Ms. Dorn. That is correct.
    Ms. Velazquez. That is why I am so puzzled. Because when 
you look at the default rate assumption for the 7(a) program 
containing Table Six of the credit supplement to the 
President's fiscal year 2003 budget it is a default rate of 
    So which is it? 12.73 or 9.38?
    Mr. Barreto. Congresswoman, what I have just been told is 
that there was a typo in----
    Ms. Velazquez. Oh, a typo.
    Mr. Barreto. There was a typo in that budget.
    Ms. Velazquez. So what is it?
    Mr. Barreto. It is 12.73.
    Mr. Wilkinson. I would direct your attention to the chart 
on the wall that says 8.19.
    Chairman Manzullo. I have 9.38 on the budget.
    Ms. Velazquez. Tony, would you please explain this type of 
phony number here?
    Mr. Wilkinson. It is real interesting. The credit 
supplement does contain the 9.38 default assumption which we 
would believe would be in the correct range of default that 
this program has been managed to. But it has on occasion been 
very difficult to get a straight answer on some of the 
    Again you look at a 9.38 in the credit supplement. They 
have a chart up here that says 8.19. The default estimate in 
the subsidy models show the 12.73.
    Ms. Velazquez. Ms. Dorn, let us see if we find some more 
    Let us look at Table Six of the credit supplement. If you 
look at the default rate assumptions for 504 it is 8.32 
percent. Is that correct?
    Ms. Dorn. I do not have that table in front of me.
    Mr. Barreto. That is a correct number.
    Ms. Dorn. 8.32 is correct according to the Administrator.
    Ms. Velazquez. So we do not have a typo there.
    Ms. Dorn. No.
    Ms. Velazquez. Okay, good.
    But if you look at page 49 of SBA's budget request you 
state for the 504 program, and quote, ``Default amounts to 
about $60 to $70 million annually.'' With 504 loan volume 
averaging approximately $2 billion for the last several years 
this $60 to $70 million in annual defaults would equate to a 
default rate of approximately three to four percent, not 8.32 
    That cannot be right.
    Ms. Dorn, how does OMB get 8.3 percent for a default 
forecast given the $60 to $70 million forecast in the budget?
    Ms. Dorn. Congresswoman, I would refer back to Dr. 
Blanchard's testimony about how these loans are calculated 
using historical data and historical default rates.
    Ms. Velazquez. Mr. Crawford, maybe you will be able to 
explain that.
    Mr. Crawford. Yes, ma'am.
    I do not have a fancy chart like that because I cannot 
afford it, but I have a little chart that I put in the back of 
my testimony that has the real historical default rates for our 
program. These numbers are not projections. They are taken from 
our trustee, the Bank of New York, so they are real, honest to 
goodness numbers.
    I decided to plot a little curve against those default 
rates, and that is that heavy black line. It seems to indicate 
that the default rate for the last 12 years has been running 
anywhere from three, 3.5, four, 4.5 percent.
    Ms. Velazquez. Not eight percent.
    Mr. Crawford. Not 8.3 percent. So I have no explanation for 
where 8.3 comes from.
    Chairman Manzullo. Can I go to Ms. Napolitano before we----
    Ms. Velazquez. Let me, just one more question.
    Chairman Manzullo. Okay.
    Ms. Velazquez. Ms. Dorn, I guess Mr. Crawford that these 
again sound like phony numbers.
    But Ms. Dorn, you stated earlier that OMB was not using 
phony numbers, but how can you say that when you are using 
incorrect recovery, incorrect loss levels, and incorrect 
    The last company that kept their books like this was a 
little company in Texas, Enron. And you know what happened to 
    So how can you sit before this Committee and tell us this 
garbage with a straight face? Because it does not matter what 
model you use. If you put garbage in you will get garbage out, 
and that is what we have seen here today.
    Thank you, Mr. Chairman.
    Chairman Manzullo. I would like to go to Ms. Napolitano.
    Somewhere during the course of this I would like to know if 
the SBA ever advised this Committee that there was a typo error 
in this budget. Was that ever done?
    Mr. Blanchard. I am embarrassed to say, Mr. Chairman, that 
I just realized it today.
    Ms. Velazquez. Really?
    Chairman Manzullo. Okay.
    Ms. Napolitano?
    Ms. Napolitano. Thank you, Mr. Chair. I am enjoying the 
questioning that our Ranking Member is doing. I hate to 
interrupt her as she is on a roll.
    But I would like to ask one of the questions that has kind 
of puzzled me, and this is Ms. Dorn's testimony, on her page, 
probably about page four in the OMB's ruling implementing FCRA. 
And I direct you to paragraph two. The second sentence starts 
off, `Staff worked to ensure the estimates have been calculated 
in accordance with the requirements of FCRA and applicable 
guidance issued by OMB, Treasury, Financial Accounting 
Standards Advisory Board and other entities. A large part of 
its role is ensuring that consistent standards are applied to 
similar programs,' and it goes on.
    How do you use the same standards for the SBIC program as 
you do for the 7(a) and/or the 504? Is that correct? Do you use 
the same standards?
    Ms. Dorn. We use similar inputs, Congresswoman.
    I think one of the issues here is how we have gotten to the 
subsidy rate. We had a long discussion about that. One of the 
issues that we talked about is the historical performance of 
loans and the historical performance of defaults.
    Ms. Napolitano. That is not the information we are getting, 
is that you do not use the same standards.
    Ms. Dorn. Well clearly there are some differences between 
SBIC and----
    Ms. Napolitano. What differences are there? In other words, 
what is it that separates one from the other and why are not 
the same standards applied to both? SBIC is what we are getting 
to, the small business investment program.
    Ms. Dorn. Congresswoman, I think we have tailored some 
criteria to deal with the different kinds of--Clearly 7(a) is a 
different kind of program than 504 which is a little different 
than the program that you cite.
    I think in general we try to use historical data. We have 
tried to pump in more current, changes in current law, changes 
in current econometrics----
    Ms. Napolitano. You are not using the same standards?
    Tony, can you explain this?
    Mr. Wilkinson. I do not think they do consistently apply 
the standards to similar programs. We are active in with the 
B&I program over in Department of Ag, it gets a little 
different treatment.
    I am most concerned about the statement that they calculate 
the rates in accordance with the requirements of the Federal 
Credit Reform Act. There are no requirements in the Federal 
Credit Reform Act of how they calculate the subsidy rate.
    Ms. Dorn says they were required to use all historical cash 
flows. Nowhere in the Federal Credit Reform Act does it say 
they have to use all historical cash flows. Again, we go back 
to the Senate Roundtable where six or seven different types of 
approaches that would have been more reflective of the 
performance of the program were presented and all of them were 
rejected, and yet OMB decided to keep the one that gave the 
highest default estimates. Even though they know it is not in a 
reasonable range.
    Ms. Napolitano. Mr. Wilkinson, Tony, would you comment more 
about the Ag? You mentioned it and you just skimmed over it. 
That is important.
    Mr. Wilkinson. We have been working on the Business and 
Industry loan program and have not has as much luck in learning 
all about their subsidy model as we have in the 7(a) program, 
so I will not profess to be an expert on their model and should 
save that one for another day.
    Ms. Napolitano. Okay.
    Then Ms. Dorn, given that there is apparently a discrepancy 
in how people are looking at how you apply the standards, are 
you willing to correct this inequity?
    Ms. Dorn. Congresswoman, we are working hard on trying to 
refine the data to go into the econometric model, but----
    Ms. Napolitano. It is not the refining of the data I am 
concerned with. It is the application of the data that concerns 
    Ms. Dorn. We do this on an annual basis, so clearly every 
year we review what we have done in the past and we are trying 
to refine the data which, as I pointed out before you came in, 
has resulted in a significant improvement on our estimates. In 
the early 1990s we were overestimating in the $200 million 
range; in fiscal 2000 we missed it by $2 million.
    So there is improvement in this data. It is not a perfect 
science, and we welcome your input and the Committee's input 
into the different factors that need to be considered. This is 
not a static calculation that does not change from year to 
year. We have continued to try and refine the model.
    Ms. Napolitano. I must say that I have a lot of small 
business and Mr. Barreto knows that, who really count on these 
loan programs, and I am finding out that a lot of them are not 
qualifying simply because of the different standards utilized.
    So Mr. Chair, I yield back the balance of my time to Ms. 
Velazquez if Ms. Velazquez wants to finish up.
    Chairman Manzullo. Let me go to Mr. Issa.
    Ms. Velazquez. Okay.
    Mr. Issa. Thank you, Mr. Chairman.
    I apologize for not being here for the whole time. 
Unfortunately I have an amendment on the Floor which is the 
usual excuse for rudeness we have here in this body.
    But I am very concerned about this and I guess directing to 
the left side of the panel, especially to the Administrator and 
Ms. Dorn, I came out of the business world. I was both a CEO 
and on many boards. What I hear you saying, and for that matter 
said very eloquently in writing, is that there is a historic 
difference between what you say is going to happen and what 
    In business we talk about a quarter--maybe, being an 
anomaly. Two quarters is a trend. Three quarters is a question 
of whether the CEO is still going to be around.
    Now maybe we act too quickly in business, but a decade 
seems to be too slow even for government.
    What I am sharing the sentiments of my colleagues is, why 
is it that this year, not next year, not the year after, can we 
not have a substantial, based on historic, and you certainly 
have the evidence, the data now to do it, why can we not have a 
halving of the difference? Why is it that while you are 
calculating we cannot narrow this from eight to seven? Maybe 
not to four? Maybe not to what we would like to believe the 
past performance shows for the future. But what is it, if there 
is nothing in the body of law that we have delivered you that 
stops you, what stops you from doing what you can to improve 
your ability to deliver services? And clearly, based on what I 
have seen, you have the ability to do it. This is a decision 
you are making.
    When a CEO makes a decision to continue miscalculating 
something historically, I have to ask how long do I want to 
have people who miscalculate that way?
    I will take either of your answers.
    Ms. Dorn. Well Congressman, we do not intentionally 
miscalculate. I think there is a process of refining that we 
have made some significant progress in the last couple of 
years. I cannot speak very well to the program over the last 
ten years in terms of how the program was run, but I think we 
are making a significant effort to try and narrow the 
differences here.
    The other thing that I would point out is something that 
you as a businessman know very well, and that is that the 
period of time between 1992 and 2000 was a period of 
extraordinary growth and economic prosperity in this country. 
We have seen over the last year that that is not necessarily 
the template for every year for the rest of all time. We have 
to take that into account.
    There have been experiences where the federal government 
has provided loans in response to outside influences and other 
things that turned out not to be a great idea. The S&L 
situation of the early 1990s is a great example of where the 
government overextended itself.
    Mr. Issa. And I am not asking you to change your model for 
who you loan to. I am not saying we should change the risk 
    What I am saying is you have a past performance, a success 
if you will, of having a given default rate and a given cost 
for it, and yet we are asked here, with the reluctance of the 
appropriators, to over-appropriate if you will based on a 
historic misjudgment.
    I do not agree with you that you have made substantial 
reforms because if the facts continue to show this big a 
disparity, this is like a four percent loss at the bottom line 
in your corporation every year, and I am not even giving you 
credit for the multiplier effect to the benefit of your making 
greater amounts of loans, being able to do more. Forget about 
all of that. Forget about what the multiplier is.
    The bottom line is you are miscalculating by an amount 
greater than we tolerate losses in business for a quarter or 
two and you have done it and your predecessors have done it for 
a decade.
    My question succinctly back to you is, why is it that you 
cannot make a substantive change this year to reduce that 
disparity? You do understand that this body, and we are the 
people, at least are part of the body that appropriates if 
    If we suddenly see a change it is not a huge amount of 
dollars, it is not something that Congress is not going to go 
okay, for one year we can do a supplemental to make up for some 
anomaly. The fact is, you are not exercising, as I see it, the 
power you have by every year miscalculating by this huge amount 
and then coming back and saying effectively look how well we 
did. I do not want you to do that well. I want you to be 
accurate, or at least much closer.
    Chairman Manzullo. Okay.
    Mr. Wilkinson. Can I add on just one statement?
    Mr. Issa. Please.
    Mr. Wilkinson. That is based on the default assumptions 
that are plugged into the model we are going to have to go ask 
for appropriation dollars for fiscal 2003 in an amount double 
what we are going to need. And it is exactly right, this makes 
the appropriation process very, very inefficient.
    Chairman Manzullo. Before I go on to Congresswoman 
Millender-McDonald, let me ask a question, Dr. Blanchard.
    The typo that appears on page 20 of the budget on Table Six 
at 9.38, you said that is a typographical error. It should be 
the 8.1 figure, Doctor?
    Mr. Blanchard. Mr. Chairman, this is a number that I just 
saw this morning before we sat down in front of this committee.
    Chairman Manzullo. You mean the 9.3?
    Mr. Blanchard. That is right. That number should be 12.73, 
    Chairman Manzullo. It should be 12.73?
    Mr. Blanchard. That is correct.
    Chairman Manzullo. But it is 8.1 on the chart.
    Ms. Velazquez. Mr. Chairman?
    Chairman Manzullo. Just a second.
    Mr. Blanchard. That number should be 12.73 which is the 
default rate that SBA has projected for fiscal year 2003.
    Chairman Manzullo. But that is 8.195 up there.
    Mr. Blanchard. The 8.19 is equivalent to the 12.73 and let 
me explain how.
    You are dividing the default in this chart by total loan 
volume. But as you know, SBA only guarantees 85 percent of that 
total loan volume.
    In the budget, the way we produce that number in our 12 
percent assumption is defaults divided by the proportion of the 
outstanding loans that SBA guarantees.
    In a sense, sir, the denominator is smaller driving that 
number up. They are equivalent to each other.
    Chairman Manzullo. So where it says 9.38 it should be 12 
    Mr. Blanchard. Seven-three.
    Chairman Manzullo. The fact that there is a typographical 
error that is in this table on Table Six on page 20, does that 
have anything to do with the final calculation of the subsidy 
    Mr. Blanchard. No, sir. It does not.
    Chairman Manzullo. Okay.
    Ms. Millender-McDonald?
    Ms. Millender-McDonald. Thank you, Mr. Chairman, and than, 
you and the Ranking Member for providing us this opportunity to 
speak about a very critical issue.
    Again, I think the basic question that has been raised here 
is whether or not OMB should rely on ten years of data versus 
more recent data to produce a default rate that appropriate 
current reality?
    I am sorry that I was not here, but I need to kind of get 
that back in front of me, why is it that we are relying on ten 
years of data as opposed to something that is more current?
    Perhaps you have answered that, but I need to hear it 
    Ms. Dorn. No, I am happy to explain our rationale on that 
    Our rationale is simply that for this program and for the 
other federal loan programs we rely on the historical data that 
we have. In this case it goes back I think to 1986 because 
economic conditions do play heavily into historical defaults 
and credits issued. So we do use the data that we have going 
back but it is not the only data that we have plugged into the 
calculation. And as we have gone forward we have sought to 
refine that.
    As we pointed out in the fiscal 2003 calculation we did use 
the information that default rates are lower for those loans 
that go through preferred lenders and we did crank that into 
the data which did reduce the subsidy rate fairly 
significantly, by about 40 percent?
    Mr. Barreto. Twenty percent.
    Ms. Dorn. Twenty percent for fiscal year 2003.
    So we are continuing to work to make it more reflective of 
not only the historical past but the more recent changes in law 
and performance.
    Ms. Millender-McDonald. Then how many 7(a) loans that have 
been made since 1986 are still on the books of the SBA and what 
percentage of these loans represent the loans made in those 
years and of the loans outstanding today?
    Ms. Dorn. Can I turn that to SBA? They have a better----
    Mr. Barreto. We have to provide you with an accurate count 
on that, Congresswoman, in terms of how many loans are still in 
the portfolio. We would be happy to do that. We did not bring 
that specific number with us today but we can get it to you 
very quickly.
    Ms. Millender-McDonald. How soon can we expect that, Mr. 
    Mr. Barreto. I think we should have it to you in the next 
48 hours.
    Ms. Millender-McDonald. That would be great because I 
really do need to review that. This question is for either Mr. 
Wilkinson or Mr. Crawford.
    Do you have any confidence small businesses and your 
members are going to get a better shake from this econometric 
model than they do from the present model?
    Mr. Wilkinson. I said in my verbal testimony that----
    Ms. Millender-McDonald. And I am sorry, I had three other 
    Mr. Wilkinson. That's fine.
    We are concerned about a new model. The model that we have 
today is really fairly simple to understand. We have issues 
with some of the assumptions that are plugged into the model. 
The econometric model is going to use many more `relevant 
factors' all of which are going to have a weighting that's 
going to be determined by somebody at OMB and we likened it to 
the fact that we have broken open the old black box----
    Ms. Millender-McDonald. Is that to your benefit or what?
    Mr. Wilkinson. I doubt it.
    Ms. Millender-McDonald. I see the gentleman in the blue 
shirt shaking his head in an adverse way. Dr. Blanchard.
    Dr. Blanchard, why are you disagreeing with Mr. Wilkinson?
    Mr. Blanchard. I do agree with Mr. Wilkinson that the 
econometric model will in effect weight multiple factors in 
accordance with their impact on the subsidy rates. However, 
those weights are not set arbitrarily by OMB or SBA. Those 
weights are determined by the data itself. The historical data 
    Mr. Barreto. Can I also take a stab at it?
    Ms. Millender-McDonald. Yes.
    Mr. Barreto. We----
    Chairman Manzullo. What does that mean?
    Mr. Barreto. What I would like to comment on is that the 
econometric model is not going to determine the subsidy rate, 
but one of the things the Congresswoman said is that we have to 
be very careful of what we are putting in.
    We believe the econometric model is going to allow us to 
come up with more accurate measurements so that what we are 
putting in provides us a much more realistic idea of what is 
actually occurring.
    We did that this last year, and I also want to say that we 
realized that this was a big issue before we came on board. I 
got on board in August and we began working on this situation 
    It is difficult for us to be able to look back and say all 
the factors that people were considering at the time they were 
developing subsidy rates in the past. But what we do understand 
and what we are very clear on is that we need to do a better 
job, that we need to be more accurate.
    That is why we felt very gratified that we were able to 
take the subsidy rate from 1.07 down to .88. In a very short 
period of time we were able to make some progress. We are going 
in the right direction.
    Obviously the fees went down because of P.L. 107-100, that 
caused the subsidy rate to go back up. But that was never our 
intention. That was never something that we had anticipated.
    If that subsidy rate would have been held at that .88, we 
would be looking at a budget authority right now of $9.7 
billion for our 7(a) loan program. That was our intention.
    We will continue to work with you, we will continue to work 
with the organizations that are here because the work is not 
done. We think that the econometric model is going to get us to 
a place where we are going to have more accurate reflections 
and we are not going to have this continued problem because we 
are concerned about it as well.
    We have to be competitive. We have to think and act like a 
small business just as the Congressman said. If our loans 
become too expensive, we are going to cease to have the ability 
to do what our mission is, which is to touch more small 
    At the end of the day we want to do more small business 
loans. WE think that our average loan size right now is too 
large. Inc. Magazine says that most small businesses are 
financed with $50,000. Our average loan size is $239,000.
    Ms. Millender-McDonald. Mr. Administrator, I want to yield 
to the Ranking Member----
    Chairman Manzullo. Ms. Millender-McDonald, the time has 
expired. Let us let Mrs. Kelly----
    Ms. Millender-McDonald. Oh, my time has expired? I just 
wanted to answer the--I wanted to convey----
    Chairman Manzullo. Let me get a question in because the 
bell is going to go off and I want to make sure everybody has 
the opportunity to ask questions.
    Mrs. Kelly?
    Ms. Kelly. Thank you very much.
    Actually, there are two things.
    Mr. Barreto, I know you have not been at the SBA for a very 
long period of time, but I have been sitting here year after 
year after year listening to the SBA come in and say we want to 
do a better job, and yet I know there are people in my area who 
have come in and asked for 7(a) loans and they are told we do 
not want to process those loans, they are too small. They want 
to process loans that are larger because they have to do the 
same amount of paperwork.
    We need some real change. We need some people who are 
dedicated to coming here in front of this Small Business 
Committee and saying we have changed it. We have reduced the 
paperwork and we are responsive to the smaller loan needs of 
these people who are trying to get into the program.
    We cannot grow small business in this nation unless you 
address these problems.
    I really hope that this coming fall, perhaps, you will come 
back to this Committee and you will come back and say here is 
what we have actually done, and you will have done something.
    The small businesses of this nation I do not think can any 
longer wait.
    I want to go also to Mr. Wilkinson.
    Mr. Wilkinson, I think you had a response to Ms. Millender-
McDonald's question. I would like to give you time to respond.
    Mr. Wilkinson. Yes, I did have one more comment.
    At our annual convention last fall in San Francisco our 
lobbyist stood up and made a statement that he believed that 
OMB created assumptions so that it would force an answer in the 
model that was going to achieve a desired result. That they 
knew what they wanted the answer to be so they would plug in 
assumptions to match it. This was his speech from the 
    Two former OMB budget examiners came up after his speech 
and said you are exactly correct.
    Exactly correct.
    Hence why we are a little excited about getting this 
problem fixed, and why we would have no more confidence in an 
econometric model that has even more assumptions that are going 
to be plugged into the model.
    I hope things work out, but hence our concern.
    Thank you for allowing me to respond.
    Ms. Kelly. Thank you, Mr. Chairman. I yield back.
    Chairman Manzullo. I have a question here.
    Two weeks ago we had a hearing with the OMB and the 
Administrator on why six months had passed and the size 
standards had not been changed. What is apparent at this point 
is that OMB may be doing its job too well. Their job is to sit 
there and to make sure that as little money is spent as 
possible as they sit there at the throttle of how the 
Administration wants to spend money.
    But it reaches a point where I think what OMB is doing is 
not correct because it is so conservative in the estimates that 
it is actually doing a disservice to small business people.
    What we are going to do is sit down and we are going to 
examine the OMB and how they go about this because I think the 
starting point of OMB is to say we are going to take the most 
conservative position possible. But in all fairness, if you 
take a look at what happened in 1996, OMB underestimated the 
loss by $257 million so they were short almost a quarter of a 
million dollars and that will cause anybody to be very 
conservative coming out of the box for the next anticipated 
    The next year, for 1997 they were long by $16 million; for 
1998 by $279 million; 1999 by $545 million; 2000 by $235 
million; 2001 by $528 million; 2002 by $183 million.
    I think what is happening is that OMB is still reacting to 
what was an incorrect estimate for the loss during 1996.
    There were more questions that people wanted to ask. 
However you want to divide it, that is okay with me.
    Ms. Napolitano. I want to defer to the Ranking Member.
    Ms. Velazquez. Thank you, Mr. Chairman.
    Ms. Dorn, is not the outcome of econometric modeling 
dependent upon the assumptions used in the model, just as the 
present model's results depend on the assumptions used in it?
    Ms. Dorn. Yes, maam. We are----
    Ms. Velazquez. I just want a yes or no.
    Is it a fact, Mr. Barreto you made reference to it, that 
you are moving to an econometric model, an admission that the 
current system is broken?
    Ms. Dorn. I think we believe an econometric model would 
provide more accurate data.
    Ms. Velazquez. So you believe that the present system is 
broken is it not? Do you not?
    Ms. Dorn. Congresswoman, I would----
    Ms. Velazquez. Just say yes or no.
    Ms. Dorn. I would prefer to use my own words, which is that 
this is a process of refinement. I would agree with you that 
the system is not close to perfect. We are making some changes 
in it even as we speak that are going to make it more accurate.
    Ms. Velazquez. Mr. Wilkinson?
    Mr. Wilkinson. Can I go back to the Chairman's comment? You 
were reading from some sheet that talked about expected default 
rates. From the numbers I have in the budget they showed an 
expected default rate in 1996 of 17.26 and the expected default 
rate of 9.97. They more than doubled the estimate in 1996. The 
estimate in 1997 was 17.25, they were on pace to do 9.13. I am 
not following----
    Chairman Manzullo. This is the actual amount of loss or 
overcharge. In 1996 OMB underestimated the loss by $257 
million. Then in subsequent years they overestimated.
    I just brought that up to show that the numbers that they 
are working with, the reason they are perhaps so conservative 
is because they had in 1995, they underestimated the loss by a 
quarter of a billion dollars.
    Mr. Wilkinson. In 1996----
    Chairman Manzullo. 1996 rather.
    Mr. Wilkinson. In 1996 there was a downward re-estimate of 
about $100 million.
    Chairman Manzullo. These are the re-estimates on it from 
    Mr. Wilkinson. I guess I would point you to page one of my 
testimony that lists the re-estimates that have been reflected 
in the annual budgets coming forward.
    Chairman Manzullo. I am on page 205 of the budget.
    If somebody else has another question here, please go 
    Ms. Velazquez. I have a question, Mr. Chairman.
    Ms. Dorn, how is the calculation of the assumptions for the 
econometric model going to correct the existing inaccuracies of 
today's method of forecasting defaults, prepayments and 
    Ms. Dorn. Congresswoman, we are looking to incorporate not 
only the historical data which we have used and will continue 
to use, but also to factor in more current conditions.
    Lloyd, do you want to talk about that in any more detail?
    Mr. Blanchard. Sure.
    Ms. Dorn. SBA is the agency that does this. OMB has 
responsibilities under the Federal Credit Reform Act, but OMB 
does not, contrary to popular belief, just make up these 
    Mr. Blanchard. Ms. Dorn is correct, Ms. Velazquez. The 
econometric model is simply an estimation tool. It estimates 
the default rate.
    The assumptions that underlie the econometric model are 
technical assumptions. They are not the assumptions that Mr. 
Wilkinson refers to. That model itself helps us, it takes 
business factors, program factors, economic factors and says 
how do these factors predict performance of these loans? 
Performance is the same thing as the default rate. Then that 
default rate is plugged into the cash flow model to predict a 
subsidy rate.
    Ms. Velazquez. Who is going to decide the assumptions used? 
    Mr. Blanchard. The assumptions used, ma'am, are, those that 
come from data. Those come from what the unemployment rate is, 
for instance. What the President's budget publishes as the 
unemployment rate. If that is a relevant factor, which I 
believe we all think it is.
    Ms. Velazquez. Mr. Barreto, you mentioned that you are 
going to use OFHEO to implement the model. I have serious 
concerns about OFHEO because home loans fit very exact criteria 
and perform in very predictable manners. All loans are 
underwritten using maximum guidelines that are established for 
total installment of debt including new home loans versus gross 
monthly income as a percentage. How statements of gross monthly 
income as a percentage will not amount to a price value.
    So due to the very precise underwriting criteria that is 
followed by every lender granting home loans, they perform in a 
very predictable fashion and are easily securitized.
    It is also very easy to predict loss due to the very large 
volumes of loans that are guaranteed on a regular basis, making 
subsidy rate models very simple to implement.
    These loans all fit in the round hole or square hole, 
depending on the market they are serving.
    Business loan underwriting is not done using that criteria 
but lenders apply different standards as it relates to cash 
flow coverage, debt to wealth and so forth. Therefore these 
loans do not fit square holes and/or round holes. This makes it 
much more difficult to predict how these loans will behave or 
perform over their life.
    At liquidation time, values of equipment will vary greatly 
due to present economy and the health or sickness of a certain 
type of business.
    Based on that, how can an oversight agency who's primary 
function is to oversee a very predictable industry now adapt a 
similar model to an industry that has no hard and fast 
underwriting guidelines?
    Mr. Barreto. Great question. Thank you, Congresswoman.
    We chose OFHEO for a couple of different reasons. First of 
all as we stated earlier, GAO and OMB when we were talking 
about what an effective approach would be, believed that the 
econometric modeling approach would help us in the short term 
and would definitely help us in the long term.
    One of the reasons that we chose OFHEO is because they have 
a lot of expertise in statistical analysis and working with 
econometric models. I agree with you though, and that is that 
you cannot take another industry that is dissimilar and try to 
overlay a system that works there over here. We are very 
mindful of that. Yet they do have the oversight responsibility 
for Fannie Mae and Freddie Mac, but also we were very intrigued 
by the fact that they hired Dr. Robert Dunsky.
    Dr. Robert Dunsky was formerly with PriceWaterhouseCoopers 
and he is very very familiar with the SBA asset sales program, 
and with our loan programs. We are very excited about is his 
understanding of the agency, his expertise, the fact that he 
has this expertise coming from the private sector. And I 
believe that working together with him, since he is already 
familiar with us, we are going to be able to come up with a 
system that is customized to the SBA.
    At the end of the day, a lot of larger agencies are able to 
do this. They have teams of economists that work for them, et 
cetera, but we needed to find a solution that was definitely 
going to be cost-effective and definitely be efficient.
    We think that OFHEO offers----
    Ms. Velazquez. Mr. Barreto, do you know why you do not, 
Congress allow Freddie Mac and Fannie Mae do business loans? 
This is like compare day and night. And we sit on the Financial 
Services Committee. We know the competence of OFHEO in dealing 
with Freddie Mac and Fannie Mae.
    Mr. Barreto. OFHEO is not going to be doing our lending 
program but they will be helping us to develop an econometric 
    Chairman Manzullo. Why do we not have these two gentlemen 
help you develop the econometric model?
    Mr. Barreto. We share a lot of information already and we 
are always open to getting more information and helping.
    I know when we worked with NADCO in the past, we actually 
developed a pilot program with them, provided them with a lot 
of our information because we know that NADCO has said to us in 
the past,'' we think there is a better way of doing this.''
    We are currently working with them and we are hopeful to 
receive some of that feedback from the discussions that we had 
on ways that we can do it better.
    Chairman Manzullo. We have to wind this up.
    A very quick question.
    Mr. Issa. The question can be responded to in writing, if 
you don't mind. Both from OMB and SBA if possible.
    Making an assumption that you have a time horizon that of 
at least five years for your reserves, not using a single year, 
not loading in the up or the down market, but using at least 
five years. If you were taking, if you will, the budget and 
multiplying it times five, would you arrive at a different 
number? If so, what would that reserve number be?
    Now you can use ten years, I don't mind, because I 
personally believe that if you look at any horizon five years 
or longer what you are going to say is our problem is a one 
year calculation.
    Because we are looking at ten years, but then we are always 
loading in a higher number for one year because of the things 
that might happen.
    If you come back to us with a five year or greater, and I 
prefer at least five, but you can do give or ten, and give us 
the number you believe if you plugged it in without all those 
hypotheticals for any one year, but knowing that over the years 
those hypotheticals fade, and give us that number, it may 
empower this Committee to recommend to the full Congress some 
additional discretion for you to make these calculations to 
narrow what we perceive as an unwarranted delta between reality 
and predictions.
    Thank you.
    Mr. Wilkinson. Congressman, attached to my testimony is a 
GAO report that did a five year look-back----
    Chairman Manzullo. We are running out of time here. I want 
to thank everybody. This has been great.
    Will Rogers talked about what the country needs is a one-
armed economist because the economists will say on the one hand 
you get this, on the other hand you get the other. Everybody 
gets an A+ here for sincerity and honesty and workmanship and 
craftsmanship and everything. It is interesting because the 
problem here is that the members of the Small Business 
Committee are accusing the Administration of being too 
conservative in their estimates. You might want to think about 
that. But that is your job as a watchdog.
    I look forward to working with you very closely. Mr. 
Barreto, as soon as he was confirmed came in my office and he 
said let's talk about this subsidy rate. OMB came in and the 
first thing out of the box was that this is very difficult, we 
want to work with you on it. And we look forward again to a 
very close working relationship because I know your heart is--
    I would assure these two gentlemen over here that these men 
are very sincere and want to work on this and continue the 
dialogue because they are listening and I know I am.
    Thank you, and good luck on that new baby.
    Mr. Barreto. Thank you very much, Mr. Chairman and Ranking 
Member Velazquez.
    Chairman Manzullo. Thank you.
    [Whereupon, at 12:02 p.m. the Committee was adjourned.]