Credit Reform: Improving Rural Development's Credit Program Cost
Estimates (Correspondence, 08/22/2000, GAO/AIMD-00-286R).

Pursuant to a legislative requirement, GAO provided information on the
Department of Agriculture's (USDA) Rural Development (RD) mission area's
progress in improving its credit program cost estimates.

GAO noted that: (1) the task force has made some progress in addressing
the deficiencies related to reasonably estimating the cost of RD's
credit programs; (2) progress made since GAO's last briefing to the
Steering Committee for credit Reform Implementation on December 16,
1999, includes updating sensitivity analyses and completing new cash
flow models for the non-housing direct and guaranteed loan programs; (3)
however, several milestone dates from the plan were missed because of a
shortage in both the staff and funding resources USDA devoted to
resolving long-standing credit reform weaknesses; (4) inadequate
resources hampered progress in critical areas, such as assessing the
quality of the data used to estimate future loan performance and
developing a new cash flow model based on available data to estimate the
cost of the housing direct loan programs; (5) as a result, RD continues
to have limited assurance that credit program costs estimates are
reasonable; and (6) while oversight for credit reform implementation
efforts falls under the guidance of the USDA Executive Steering
Committee for Credit Reform Implementation, the Steering Committee as a
whole has had little direct involvement in overseeing progress to
improve RD's credit program cost estimates.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  AIMD-00-286R
     TITLE:  Credit Reform: Improving Rural Development's Credit
	     Program Cost Estimates
      DATE:  08/22/2000
   SUBJECT:  Data integrity
	     Cost analysis
	     Credit
	     Projections
	     Government guaranteed loans
	     Direct loans
	     Financial management systems

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Testimony.                                               **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************

GAO/AIMD-00-286R

RD's Credit Reform Implementation

United States General Accounting Office Washington, DC 20548

Accounting and Information Management Division

B- 286010 August 22, 2000 The Honorable Dan Glickman The Secretary of
Agriculture

Subject: Credit Reform: Improving Rural Development's Credit Program Cost
Estimates

Dear Mr. Secretary: This letter and the attached briefing slides are the
third in a series of status reports on the Rural Development (RD) mission
area's progress in improving its credit program cost estimates under credit
reform. Until RD makes significant progress in developing well supported,
reasonable estimates of its credit program costs, an unqualified audit
opinion on the Department of Agriculture's (USDA) consolidated financial
statements will not be possible. Furthermore, since USDA is the largest
direct lender in the federal government and the credit program amounts are
material, USDA's inability to properly implement credit reform will continue
to contribute to our inability to give an unqualified audit opinion on the
consolidated financial statements of the U. S. government.

Therefore, since April 1999, we have assessed RD's credit reform
implementation efforts in such areas as (1) identifying key cash flow
assumptions, (2) improving cash flow models, (3) assessing cash flow model
data, and (4) implementing other procedures to enhance the credit subsidy
estimation process. As part of this effort, we have provided regular
briefings to USDA's Executive Steering Committee for Credit Reform
Implementation. Shortly after our first briefing in June 1999, the USDA
credit reform task force developed a detailed implementation plan to guide
the agency through its efforts to improve its credit program cost estimates.
Today, we briefed the Steering Committee on (1) the status of RD's efforts
to implement procedures from the plan and (2) recommendations designed to
make the necessary improvements to RD's cost estimates in a more timely
manner.

Results in Brief

Overall, the task force has made some progress in addressing the
deficiencies related to reasonably estimating the cost of RD's credit
programs. Progress made since our last briefing to the Steering Committee on
December 16, 1999, includes updating

B- 286010 GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 2
sensitivity analyses 1 and completing new cash flow models for the non-
housing direct

and guaranteed loan programs. However, several milestone dates from the plan
were missed because of a shortage in both the staff and funding resources
USDA devoted to resolving long- standing credit reform weaknesses.
Inadequate resources hampered progress in critical areas, such as assessing
the quality of the data used to estimate future loan performance and
developing a new cash flow model based on available data to estimate the
cost of the housing direct loan programs. As a result, RD continues to have
limited assurance that credit program costs estimates are reasonable. While
oversight for credit reform implementation efforts falls under the guidance
of the USDA Executive Steering Committee for Credit Reform Implementation,
the Steering Committee as a whole has had little direct involvement in
overseeing progress to improve RD's credit program cost estimates.

Background

Prior to the implementation of the Federal Credit Reform Act (FCRA) of 1990,
credit programs, like other federal programs, were reported in the budget on
a cash basis. Thus, loan guarantees appeared to be free while direct loans
appeared to be as expensive as grants in a given budget year. FCRA and the
related accounting standard, Statement of Federal Financial Accounting
Standards (SFFAS) No. 2, 2 Accounting for Direct Loans and Loan Guarantees,
together known as credit reform, were established to more accurately measure
the government's costs of federal credit programs and to permit better
comparisons both among credit programs and between credit and noncredit
programs. When the Federal Accounting Standards Advisory Board (FASAB) 3
developed SFFAS No. 2, it recognized that financial accounting should
support the budget and that accounting standards for credit reform should be
consistent with the budgeting under credit reform. This mirroring provides
more assurance through the financial statement audit that credit program
budget estimates are reasonable since the estimates should be calculated
with the same data and same process. However, without consistency between
the data and processes used to prepare credit program cost estimates for the
financial statements and budget, this assurance cannot be fully achieved.

As part of implementing credit reform, agencies are required to estimate the
net cost of extending or guaranteeing credit- generally referred to as the
subsidy cost- based

1 Sensitivity analysis is a process used to identify the assumptions that,
when adjusted, have the greatest impact on the credit subsidy estimate. 2
SFFAS No. 2 was amended by SFFAS No. 18, Amendments to Accounting Standards
For Direct Loans and Loan Guarantees. The objective of the amendments was to
improve financial reporting for subsidy costs and performance of federal
credit programs.

3 FASAB was created by the Office of Management and Budget, Treasury, and
GAO to develop and recommend accounting principles for the federal
government. These three agencies approved SFFAS No. 2, Accounting for Direct
Loans and Loan Guarantees, in July 1993.

B- 286010 GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 3 on
the present value 4 of estimated net cash flows excluding administrative
costs.

Because credit program cost estimates are based on estimated cash flows,
agencies have to be able to predict borrower behavior– how many
borrowers will pay early, pay late, or default on their loans and at what
point in time. Generally, the best predictor of borrower behavior is prior
historical data adjusted for expected changes in future economic events.
However, alternative data sources including informed opinion 5 or proxy data
6 may be used in certain limited instances on a temporary basis to estimate
future cash flows in the absence of adequate historical data. Agencies use
historical information, or alternative data sources if necessary, and
sophisticated computer models, known as cash flow models, to estimate the
cost of a credit program.

As we previously reported to your office, 7 RD has received a qualified
opinion on its financial statements since 1994, due in part to its inability
to reasonably estimate its credit programs' costs. RD's problems
implementing credit reform, among other things, have also prevented USDA
from achieving an unqualified audit opinion on its consolidated financial
statements. This long- standing problem is attributable, in part, to a lack
of adequate historical data, which resulted largely from system
inadequacies. For example, prior to the implementation of credit reform,
USDA systems did not track certain key cash flow data critical to estimating
the cost of a credit program. These problems also raise concerns over the
integrity of RD's credit program budgetary data because the same cost
estimates are generally used for both budget preparation and financial
reporting.

Previous efforts to effectively implement credit reform at RD have had
limited success. In response to the Inspector General's (IG) audit findings,
USDA organized a task force and developed a detailed implementation plan. 8
The plan includes procedures to (1) identify key cash flow assumptions, (2)
improve cash flow models, (3) assess cash flow model data, and (4) implement
other procedures to enhance RD's credit subsidy estimation process. The plan
also includes points of contact for many procedures as well as some
milestone dates.

Beginning in April 1999, we have assessed RD's efforts to improve its credit
program cost estimates under credit reform. Since our December 1999,
briefing to the Steering Committee, the task force has primarily focused on
RD's non- housing direct

4 Present value is the worth of a future stream of returns or costs in terms
of money paid immediately. In calculating present value, prevailing interest
rates provide the basis for converting future amounts into their
“money now” equivalents.

5 Informed opinion refers to the judgment of agency staff or others who make
subsidy estimates based on their programmatic knowledge and/ or experience.
6 Proxy data refer to data from other credit programs that have borrowers
with similar characteristics, such as socioeconomic status, education, or
geographic distinctions. 7 Correspondence dated June 25, 1999, to the Deputy
Secretary, Department of Agriculture.

8 The detailed credit reform implementation plan included most of the
procedures outlined in our June 16, 1999, briefing to the USDA Executive
Steering Committee for Credit Reform Implementation.

B- 286010 GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 4 loan
programs and the guaranteed loan programs, which RD reported at $40.9
billion

and $11.6 billion, respectively, as of September 30, 1999. The task force is
focusing secondarily on the direct housing loan programs, which RD reported
at $28.5 billion as of September 30, 1999. The task force has focused on the
largest programs within each loan program type that are significant to both
the agency's financial statements and budget to ensure the new procedures'
maximum effectiveness in improving the reasonableness of the overall cost
estimates of RD's credit programs.

Scope and Methodology

We evaluated RD's progress since our last briefing to the Steering Committee
on December 16, 1999, in completing procedures by the related milestone
dates included in the credit reform implementation plan. We participated in
weekly status meetings of the task force and discussed RD's progress with
numerous agency representatives throughout the process.

We assessed RD's implementation efforts based on procedures in the FASAB
Accounting and Auditing Policy Committee's Technical Release, Preparing and
Auditing Direct Loan and Loan Guarantee Subsidies Under the Federal Credit
Reform Act, as well as procedures designed to assist other credit agencies
in successfully

implementing credit reform. These efforts were in four areas: (1)
identifying key cash flow assumptions, (2) improving cash flow models, (3)
assessing cash flow model data, and (4) implementing other procedures to
enhance credit subsidy estimates.

Specifically, to identify key cash flow assumptions, we assessed RD's
approach to sensitivity analysis to determine whether RD used a systematic,
nonjudgmental and well documented process. We reviewed cash flow models as
they were developed, identified and communicated formula errors, and
subsequently reviewed RD's revisions to the cash flow models. We also
reviewed cash flow models to determine whether they reasonably represented
the cash flows of the loan programs based on the laws and regulations that
govern them. To assess cash flow model data, we worked with the USDA IG to
develop a statistically valid sampling approach that the USDA IG and we will
use to test the reliability of data supporting the key cash flow assumptions
of USDA's credit programs. Lastly, as RD implemented other procedures to
enhance credit subsidy estimates, we reviewed and conveyed potential
improvements to the documentation of cash flow assumptions and cash flow
models.

We conducted our work in Washington, D. C., from January 2000 through August
2000 in accordance with generally accepted government auditing standards. We
requested comments on a draft of our briefing slides and letter from
cognizant RD budget and finance office officials and officials from USDA's
Office of the Chief Financial Officer (OCFO). On August 17, 2000, RD and
OCFO officials provided us with oral comments, which are discussed in the
“Agency Comments” section. The enclosed briefing slides
highlight the results of our work and the information we provided during the
briefing.

B- 286010 GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 5

Further Improvements Are Needed to Ensure That Credit Subsidy Estimates Are
Reasonable

Since December 1999, some procedures included in the implementation plan
have been completed or are currently in process to address deficiencies in
estimating the costs of RD's credit programs. However, a shortage in both
the staff and funding resources USDA devoted to resolving credit reform
weaknesses has continued to hamper timely progress in properly addressing
credit reform implementation problems. As a result, numerous key procedures
included in the plan remain incomplete, and several milestone dates were
missed. Most of the procedures not yet completed relate to (1) improving the
cost estimates for the guaranteed and housing direct loan programs and (2)
developing written policies and procedures for calculating credit program
cost estimates for all three program types: non- housing direct, guaranteed,
and housing direct loan programs.

Until RD makes significant progress in implementing the remaining key
procedures in the plan, RD will continue to lack an adequate process to
develop well supported, reasonable estimates of credit program costs. As a
result, financial statements and budget estimates could be misstated. Due to
the magnitude of RD's credit programs, an unqualified audit opinion on
USDA's consolidated financial statements will not be possible without the
successful implementation of credit reform. Because USDA is the largest
direct lender in the federal government, proper implementation of credit
reform is material to the consolidated financial statements of the U. S.
government. USDA's deficiencies in this area will also continue to
contribute to our inability to give an unqualified audit opinion on the
consolidated financial statements of the U. S. government.

The following sections summarize the status of key procedures outlined in
USDA's detailed implementation plan related to RD's non- housing direct,
guaranteed, and housing direct loan programs. The status of procedures that
relate to all RD's loan program types is also presented.

Non- Housing Direct Loan Programs As of December 1999, the implementation
plan included six key procedures not yet completed for the non- housing
direct loan programs. Table 1 summarizes the established milestone dates
from the plan and the current status of these six procedures. Since
December, RD has completed three of these procedures, two are in process,
and one procedure was not yet started. Delays in completing the new cash
flow model significantly hampered progress in other areas.

B- 286010 GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 6

Table 1: Status of Non- housing Direct Loan Program Procedures Procedure
Original due

date (if any) Revised due date (if any) Current

status Finalize new cash flow model Completed

May 2000

Perform sensitivity analyses to identify key cash flow assumptions Completed

June 2000

Document cash flow assumptions Sept. 1999 Completed July 2000

Assess cash flow model data reliability April 2000 Aug. 2000 In process

Document cash flow model Sept. 1999 In process

Compare estimated to actual cash flows Aug. 1999 Not started

The three procedures that were completed relate to developing a new cash
flow model, performing sensitivity analyses, and documenting cash flow
assumptions. The new cash flow model to estimate the cost of RD's non-
housing direct loan programs was completed and reviewed. During the review
process, several refinements were implemented. After the cash flow model was
revised and completed, sensitivity analyses were updated and documented by
RD staff to identify the key cash flow assumptions. These assumptions
include the average borrower interest rate and loan term. Finally, the cash
flow assumptions were documented, including data sources and calculation
methods.

The two procedures currently in process are assessing cash flow model data
and documenting the cash flow model. To ensure that cost estimates are based
on reliable data, planning for data reliability tests was completed to
determine if accounting system data used as support for key cash flow
assumptions are reliable. The data reliability tests for the key cash flow
assumptions are currently in process. If these data are determined to be
unreliable, it may be necessary to (1) adjust the values of cash flow
assumptions, (2) identify alternative data sources, and/ or (3) revise the
cash flow model. In addition, documentation of the logic flow and mechanics
of the cash flow model is in process. This documentation was reviewed and is
currently being revised.

The procedure in the plan that was not started involves comparing estimated
to actual cash flows. This comparison allows an agency to identify and
determine if significant differences exist and whether assumptions related
to future loan performance need to be revised.

Guaranteed Loan Programs As of December 1999, the implementation plan
included six key procedures not yet completed for the guaranteed loan
programs. Table 2 summarizes the established milestone dates from the plan
and the current status of these six procedures. Since

B- 286010 GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 7
December, RD has completed one of these procedures, four are in process, and
one

procedure has not yet been started.

Table 2: Status of Guaranteed Loan Program Procedures Procedure Original due

date (if any) Revised due

date (if any) Current

status Finalize new cash flow model July 2000 Completed

Aug. 2000

Document cash flow model Sept. 2000 In process

Document cash flow assumptions Sept. 2000 In process

Perform sensitivity analyses to identify key cash flow assumptions Aug. 2000
In process

Assess cash flow model data reliability Aug. 2000 In process

Compare estimated to actual cash flows Not started

The one completed procedure relates to developing a new cash flow model to
estimate the cost of RD's guaranteed loan programs based on available data
from the accounting system. This new cash flow model was reviewed and tested
to ensure that calculations worked as intended. During the model
development, regulatory and legislative requirements were considered to
ensure that it reasonably represents the cash flows of the loan programs
based on the laws and regulations that govern them.

The four procedures currently in process relate to (1) documenting the new
cash flow model, (2) documenting the cash flow assumptions, (3) performing
sensitivity analyses, and (4) assessing the reliability of cash flow model
data. As the new cash flow model was developed, its logic flow and mechanics
were documented and are currently being reviewed and revised for model
refinements. RD staff are documenting the cash flow assumptions, including
the data sources and calculation methods. After the cash flow model was
finalized, RD staff performed sensitivity analyses to identify the key cash
flow assumptions and are reviewing the results. Once the key cash flow
assumptions are identified, tests may be performed to determine if
accounting system data used as support for these assumptions are reliable.
These tests are needed to ensure that cost estimates are based on reliable
data. If these data are determined to be unreliable, it may be necessary to
(1) adjust the values of cash flow assumptions, (2) identify alternative
data sources, and/ or (3) revise the cash flow model. Planning for data
reliability tests was completed, and RD staff are scheduled to begin
conducting these tests this fiscal year.

The procedure that was not started involves comparing estimated to actual
cash flows. This comparison allows an agency to identify and determine if
significant differences exist and whether assumptions related to future loan
performance need to be revised. A milestone date was not established for
this procedure.

B- 286010 GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 8
Housing Direct Loan Programs

As of December 1999, the implementation plan included eight key procedures
not yet completed for the housing direct loan programs. Table 3 summarizes
the established milestone dates from the plan and the current status of
these eight procedures. RD currently has one of these procedures in process,
and seven procedures were not yet started. For the seven procedures for
which no progress was made, completing six of them depends on developing a
new cash flow model 9 based on available data to estimate the cost of these
loan programs. However, development of the new cash flow model was not
started. Due to the magnitude of these loan programs, it would be unlikely
that RD or USDA could receive an unqualified audit opinion unless and until
a new cash flow model is developed. Therefore, the development of the new
cash flow model is critical for RD.

Table 3: Status of Housing Direct Loan Program Procedures Procedure Original
due

date (if any) Revised due date (if any) Current

status Assess existing cash flow data July 2000 In process

Develop new cash flow model May 2001 Not started

Ensure cash flow model represents regulatory requirements May 2001 10 Not
started

Document cash flow model May 2001 10 Not started

Perform sensitivity analyses to identify key cash flow assumptions May 2001
10 Not started

Document cash flow assumptions May 2001 10 Not started

Assess cash flow model data reliability Not started

Compare estimated to actual cash flows Not started

The procedure currently in process is assessing the cash flow data in the
accounting systems. An outside contractor was hired to analyze existing cash
flow data for the housing direct loan programs to determine if the systems
contain the appropriate data to reasonably estimate future loan performance
and identify alternative data sources, if necessary. This analysis is
necessary prior to developing the new cash flow model for RD's housing
direct loan programs to ensure that the cash flow model is based on
available data. Once the results of the outside contractor's analysis are
known, it may be necessary to temporarily use alternative data sources for
cost

9 The cash flow model currently used to estimate the cost of the housing
direct loan programs includes several cash flow assumptions that RD was
unable to adequately support. 10 The due date for this procedure is
presented in the plan as being completed with the development of the new
housing direct cash flow model, which is presented as being completed by
contractors. However, there currently is no contract for this work and no
funds were provided for these purposes.

B- 286010 GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 9
estimates and develop and prioritize system change requests to ensure that
in the

future, accounting systems provide all essential data to estimate future
loan performance.

As previously mentioned, the new cash flow model has yet to be developed.
The plan proposes developing the cash flow model and completing several
other procedures, including performing sensitivity analyses and documenting
the cash flow model and assumptions, with the assistance of outside
contractors. However, there currently is no contract for this work and no
funds were provided for these purposes. While developing the cash flow
model, it will be necessary to ensure that it reasonably represents the cash
flows of the loan programs based on the laws and regulations that govern
them. Once the cash flow model is completed, it will then be possible to
complete other key procedures- including documenting the cash flow model's
logic flow and mechanics; performing sensitivity analyses to identify the
key cash flow assumptions; and documenting the cash flow assumptions,
including data sources and calculation methods. Since the plan presents
these procedures as being completed under the contract to develop the new
cash flow model, the milestone date for all these procedures is May 2001.

The plan also calls for data reliability tests to determine if accounting
system data used as support for key cash flow assumptions are reliable. If
these data are determined to be unreliable, it may be necessary to (1)
adjust the values of cash flow assumptions, (2) identify alternative data
sources, and/ or (3) revise the cash flow model. Finally, the plan calls for
comparing estimated to actual cash flows. This comparison allows an agency
to identify and determine if significant differences exist and whether
assumptions related to future loan performance need to be revised.
Completing these procedures also depends on completing the new cash flow
model. To date, no progress was made in these areas and no milestone dates
were established.

Procedures Related to All Loan Programs As of December 1999, the
implementation plan included five key procedures not yet completed that
relate to improving RD's overall credit subsidy estimation process. Table 4
summarizes the established milestone dates from the plan and the current
status of these five procedures. Since December, RD has completed one of
these procedures, one is in process, and three procedures have not yet been
started.

B- 286010 GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 10

Table 4: Status of Procedures Related to All Loan Programs Procedure
Original due

date (if any) Revised due date (if any) Current

status Hire additional staff Completed

May 2000

Resolve reestimate issues with OMB March 2000 In process

Develop policies and procedures for calculating estimates of credit program
costs Feb. 2000 Not started

Implement procedures to automate the estimation process Not started

Conduct “lessons learned” sessions with HUD regarding the
balances approach Sept. 1999 Jan. 2000 Not started 11

The procedure that RD completed in this area is the hiring of additional
staff to assist with efforts to implement credit reform. The procedure
currently in process is resolving the timing and frequency of reestimate
issues with the Office of Management and Budget (OMB).

Three other procedures related to all loan program types have not yet been
started. The first is developing written policies and procedures for
calculating credit program cost estimates, including a formal supervisory
review process and a process to ensure that key cash flow assumptions are
reviewed and approved by budget, accounting, and program offices. The second
is developing and implementing procedures to automate the credit subsidy
estimation process to reduce workloads and the chance of data entry errors.
These procedures could include (1) centralizing input sections by program
instead of by cohort 12 and (2) linking data files to cash flow model input
sections. Finally, RD has yet to hold “lessons learned” sessions
with the Department of Housing and Urban Development (HUD) staff to
understand how they implemented the balances approach 13 and automated their
estimation processes. However, the task force has put this procedure on hold
until OMB issues guidance on the balances approach.

Implementation Plan Oversight Could Be Improved

The implementation plan was developed to help guide the agency through its
credit reform implementation efforts extending into fiscal year 2001. This
plan includes points of contact for many procedures and some milestone
dates. Thus far, RD has experienced many delays in meeting milestone dates,
and several procedures did not

11 The task force has put this procedure on hold until OMB issues guidance
on the balances approach. 12 A cohort includes those direct loans or loan
guarantees of a program for which a subsidy appropriation is provided for a
given fiscal year even if disbursements occur in subsequent years. 13 The
Office of Management and Budget has characterized the balances approach as a
simplified, more direct method of reestimating credit subsidies. Detailed
implementation guidance has not yet been developed.

B- 286010 GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 11 have
due dates established. For example, for the 20 key procedures not yet

completed, 6 have missed the due dates, 4 lack due dates, and 10 have due
dates in the future. Further, for the 5 recently completed procedures, 2
were completed after the due dates and 3 lacked due dates. Until RD makes
significant progress in implementing the remaining key procedures in the
plan, RD will continue to lack an adequate process to develop well
supported, reasonable estimates of credit program costs. As a result,
financial statements and budget estimates could be misstated.

While oversight for credit reform implementation efforts falls under the
guidance of the USDA Executive Steering Committee for Credit Reform
Implementation, the Steering Committee as a whole has had little direct
involvement in overseeing progress to improve RD's credit subsidy estimates.
The Steering Committee does not meet on a regular basis to monitor the task
force's progress and was not made aware of missed or changed milestone
dates.

However, there were individual efforts by task force and Steering Committee
members to eliminate barriers impeding implementation efforts. An OCFO task
force representative has provided general progress updates approximately
every 2 weeks to the USDA Chief Financial Officer and has met separately on
various occasions with individual members of the Steering Committee to
discuss specific issues, such as funding requests. While these efforts are
important, Steering Committee oversight and monitoring of the task force's
activities are critical to timely and successful implementation of credit
reform at the agency.

RD Has Not Committed Sufficient Resources to Correct Credit Reform Issues

In our March 2000 testimony, 14 we stated that limited progress in
addressing the deficiencies related to estimating the cost of USDA credit
programs was primarily attributable to a shortage of resources, both staff
and funds, to properly address the problem. Although the agency recognized
the need for additional qualified staff, between April 1999 and April 2000,
RD had internally reassigned only two staff that worked part time on credit
reform implementation issues. Recently, RD has hired two additional staff
that management plans to dedicate full time to solving these longstanding
problems.

Delays in completing the non- housing direct cash flow model due to a
shortage of both staff and funding resources USDA devoted to resolving
credit reform weaknesses have significantly hampered progress in performing
sensitivity analyses and assessing the quality of the data RD uses to
predict future loan performance. Additionally, delays in identifying the
necessary funding for outside contractors have significantly impeded the
development and implementation of a new cash flow model for RD's housing
direct loan programs. In April and June 1999, we met with the Steering
Committee and discussed how other agencies have successfully used outside
contractors to help gather adequate historical data, establish a reliable
basis

14 Financial Management: USDA Faces Major Financial Management Challenges(
GAO/ T- AIMD- 00- 115, March 21, 2000).

B- 286010 GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 12 for
cash flow estimates, and improve cash flow models in a timely manner. In our

March 2000 testimony we again discussed this issue. Because the single
family housing program's interest calculations are very complex, it is
unlikely that the housing direct cash flow model will be developed in a
timely manner with the resources that USDA has made available to the task
force thus far. While funds were made available to assess the existing cash
flow data for the housing direct loan programs, to date, no funding was
identified to complete the critical procedure of developing a new cash flow
model. Until this new cash flow model is developed, most of the other key
procedures related to the housing direct loan program cannot be completed.

Conclusions

RD's progress in improving credit program cost estimates is slow, and
several procedures are not completed because USDA did not allocate to the
task force the necessary dedicated full- time staff or adequate funding for
outside contractors. Without sustained top level management commitment,
guidance from the Steering Committee, and the necessary dedicated staff and
funding resources, RD will not be able to improve the quality of its credit
program cost estimates in a timely manner. Until RD makes significant
progress in implementing the remaining key procedures, it will continue to
lack an adequate process to develop well supported, reasonable estimates of
credit program costs. As a result, financial statements and budget estimates
could be misstated. Due to the magnitude of RD's credit programs, an
unqualified audit opinion on USDA's consolidated financial statements will
not be possible without the successful implementation of credit reform.
Since USDA is the largest direct lender in the federal government, proper
implementation of credit reform is material to the consolidated financial
statements of the U. S. government. USDA's deficiencies in this area will
also continue to contribute to our inability to give an unqualified audit
opinion on the consolidated financial statements of the U. S. government.

Recommendations

In order to improve Rural Development's estimates of credit program costs in
a timely manner, we recommend that the Secretary of Agriculture or his
designee take the following actions:

ï¿½ Commit the necessary staff and/ or funding resources to promptly implement
the procedures in the detailed implementation plan.

ï¿½ Set milestone dates for all the procedures included in the detailed
implementation plan.

ï¿½ Establish a process whereby the Executive Steering Committee for Credit
Reform Implementation routinely monitors the progress of credit reform
implementation efforts. With respect to the detailed implementation plan,
establish a process for the Steering Committee to (1) be notified when
scheduled due dates will not be

B- 286010 GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 13 met
so that corrective actions can be taken and (2) approve changes to scheduled

due dates.

Agency Comments

On August 17, 2000, we obtained oral comments on a draft of our briefing
slides and letter from cognizant RD budget and finance office officials and
USDA OCFO officials, who generally agreed with our findings, conclusions,
and recommendations. Their comments were incorporated where appropriate.

---- This report contains recommendations to you. The head of a federal
agency is required by 31 U. S. C 720 to submit a written statement on
actions taken on these recommendations. You should submit your statement to
the Senate Committee on Governmental Affairs and the House Committee on
Government Reform no later than 60 days after the date of this letter. A
written statement must also be sent to the House and Senate Committees on
Appropriations with the agency's first request for appropriations made more
than 60 days after the date of this letter.

We are sending copies of this letter to the Honorable Richard E. Rominger,
Deputy Secretary for Agriculture; the Honorable Sally Thompson, Chief
Financial Officer of Agriculture; and the Honorable Jill Long Thompson,
Under Secretary for Rural Development. Copies of this letter will be made
available to others upon request.

If you have any questions about this letter or the briefing, please contact
me at (202) 512- 9508 or Dan Blair, Assistant Director, at (202) 512- 9401.
Key contributors to this assignment were McCoy Williams, Marcia Carlsen and
Mary Papadopulos.

Sincerely yours, Linda M. Calbom Director, Resources, Community,

and Economic Development, Accounting and Financial Management Issues

Enclosure

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 14

Briefing to the USDA Executive Steering Committee for Credit Reform
Implementation

1 Accounting and Information

Management Division

Status of Rural Development Credit Reform Implementation Efforts

Briefing to the USDA Executive Steering Committee for Credit Reform
Implementation

August 22, 2000

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 15

2 Contents

Purpose Background Status of Task Force Efforts Implementation Plan Issues
Oversight Issues Resource Issues Conclusions Recommendations

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 16

3 Purpose

The purpose of this briefing is to provide the Steering Committee an update
on the status of USDA's credit reform task force's efforts to improve Rural
Development's (RD) subsidy estimation process since our December 16, 1999,
briefing. We will provide an update on the status of procedures related to

non- housing direct loan programs, guaranteed loan programs, and housing
direct loan programs.

Additionally, the status of procedures that relate to all loan program types
will be presented.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 17

4 Purpose

With this briefing we will provide recommendations designed to help RD
complete the remaining procedures in the implementation plan to improve
estimates of credit program costs.

We will formalize our recommendations in a letter to the Secretary.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 18

5 Background

On June 16, 1999, we briefed this Committee on procedures outlined in the
Technical Release,

Preparing and Auditing Direct Loan and Loan Guarantee Subsidies Under the
Federal Credit Reform Act, 1 which includes procedures that were
successfully implemented by other agencies, and

additional procedures that RD should implement. In response, USDA developed
a detailed

implementation plan, which included most of these procedures.

1 This Technical Release was approved by the Federal Accounting Standards
Advisory Board and is expected to be issued by the Office of Management and
Budget.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 19

6 Background

The detailed implementation plan was developed to help guide the agency
through its credit reform implementation efforts extending into fiscal year
2001.

This plan includes procedures to identify key cash flow assumptions, improve
cash flow models, assess cash flow model data, and other procedures to
enhance RD's credit subsidy estimation process.

The plan also includes staff responsible for most tasks as well as some
milestone dates.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 20

7 Background

On December 16, 1999, we briefed the Executive Steering Committee for Credit
Reform Implementation on the status of the task force's efforts to improve
RD's credit subsidy estimation process.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 21

8 Background

Between our June 1999 and December 1999 briefings, RD accomplished the
following:

RD budget and finance staff were trained to conduct sensitivity analyses 2
and performed these analyses for the large non- housing direct loan programs
with a preliminary version of the non- housing direct cash flow model.

Summaries of the regulatory and legislative requirements for all programs
were completed.

2 Sensitivity analysis is a process used to identify the assumptions that,
when adjusted, have the greatest impact on the credit subsidy estimate.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 22

9 Background

The non- housing direct cash flow model was compared to the summarized
program and legislative requirements to ensure that it reasonably
represented the cash flows of the loan programs based on the laws and
regulations that govern them.

A template was developed to document the cash flow assumptions, data
sources, and calculation methods.

An overview briefing was provided by the Office of Management and Budget
(OMB) on implementing the balances approach to reestimates. 3 3 OMB has
characterized the balances approach as a simplified, more direct method of
reestimating credit subsidies. Detailed implementation guidance is not yet
developed.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 23

10 Background

Since our last briefing, the task force has primarily focused on procedures
related to RD's non- housing direct loan programs and guaranteed loan
programs, which RD reported as $40. 9 billion and $11. 6 billion,
respectively, as of September 30, 1999.

The task force is focusing secondarily on procedures related to the housing
direct loan programs, which RD reported as $28.5 billion as of September 30,
1999.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 24

11 Background

Due to the magnitude of RD's credit programs, an unqualified audit opinion
on USDA's consolidated financial statements will not be possible without the
successful implementation of credit reform.

Because USDA is the largest direct lender in the federal government, proper
implementation of credit reform is material to the consolidated financial
statements of the U. S. government.

USDA's deficiencies in this area will continue to contribute to our
inability to give an unqualified audit opinion on the consolidated financial
statements of the U. S. government.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 25

12 Background

The programs most significant financially to both RD's financial statements
and budget were identified to ensure the new procedures' maximum
effectiveness in improving the reasonableness of the overall cost estimates
of RD's credit programs.

Subsequently, these procedures will be implemented for the smaller programs.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 26

13 Status of Task Force Efforts

The following slides present the status of key procedures outlined in USDA's
detailed implementation plan related to RD's

non- housing direct loan programs, guaranteed loan programs, and housing
direct loan programs.

Additionally, we will update the status of procedures that relate to all
RD's loan program types.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 27

14 Status of Procedures Related to NonHousing

Direct Loan Programs The following table summarizes key procedures that were
not

completed as of December 1999, their scheduled due dates from the
implementation plan, and the current status.

Procedure Original due date (if any)

Revised due date (if any) Current status

Finalize new cash flow model Completed May 2000

Document cash flow model September 1999 In process

Perform sensitivity analyses to identify key cash flow assumptions

Completed June 2000

Document cash flow assumptions September

1999 Completed July 2000

Assess cash flow model data reliability April 2000 August 2000 In process
Compare estimated to actual cash flows August 1999 Not started

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 28

15 Procedures Related to Non- Housing

Direct Loan Programs Procedures completed or in process for the nonhousing
direct loan programs:

The cash flow model was completed and reviewed.

Several refinements were implemented. The logic flow and mechanics of the
cash flow

model were documented and reviewed and are currently being revised.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 29

16 Procedures Related to Non- Housing

Direct Loan Programs Sensitivity analyses were updated and

documented by RD staff to reflect revisions made to the non- housing direct
cash flow model.

Key cash flow assumptions were identified. RD staff have primarily focused
their efforts on

documenting and supporting these assumptions.

The programs' cash flow assumptions were documented, including data sources
and calculation methods.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 30

17 Procedures Related to Non- Housing

Direct Loan Programs Data reliability tests were planned and are

currently in process for key cash flow assumptions. If it is determined that
accounting system data supporting key cash flow assumptions are not
reliable, it may be necessary to

adjust the values of cash flow assumptions, identify alternative data
sources, and/ or revise the cash flow model.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 31

18 Procedures Related to Non- Housing

Direct Loan Programs The procedure that remains to be started for the non-
housing direct loan programs:

Estimated loan performance needs to be compared to actual cash flows to
determine if significant differences exist and whether assumptions related
to future loan performance need to be revised.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 32

19 Status of Procedures Related to

Guaranteed Loan Programs The following table summarizes key procedures that
were not

completed as of December 1999, their scheduled due dates from the
implementation plan, and the current status.

Procedure Original due date (if any) Revised due

date (if any) Current status Finalize new cash flow model July 2000
Completed

August 2000 Document cash flow model September

2000 In process Perform sensitivity analyses to identify key cash flow
assumptions

August 2000 In process Document cash flow assumptions September

2000 In process Assess cash flow model data reliability August 2000 In
process Compare estimated to actual cash flows Not started

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 33

20 Procedures Related to Guaranteed

Loan Programs Procedures completed or in process for the guaranteed loan
programs:

A new cash flow model was developed based on readily available data from the
accounting system and concepts 4 from other agencies' cash flow models. This
cash flow model was reviewed and several refinements were implemented.

Regulatory and legislative requirements were considered during the
development.

The logic flow and mechanics of the new cash flow model were documented and
are currently being reviewed. 4 To facilitate model development, sections of
this model were patterned after other agency cash flow models that were
audited and determined to provide a reasonable basis to estimate future loan
performance.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 34

21 Procedures Related to Guaranteed

Loan Programs Sensitivity analyses were performed to identify the

key cash flow assumptions and RD staff are currently reviewing the results.

RD staff are preparing documentation for the programs' cash flow
assumptions. This documentation will include data sources and calculation
methods.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 35

22 Procedures Related to Guaranteed

Loan Programs Data reliability tests were planned for key cash

flow assumptions. If it is determined that accounting system data supporting
key cash flow assumptions are not reliable, it may be necessary to

adjust the values of cash flow assumptions, identify alternative data
sources, and/ or revise cash flow models.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 36

23 Procedures Related to Guaranteed

Loan Programs The procedure that remains to be started for the guaranteed
loan programs:

Estimated loan performance needs to be compared to actual cash flows to
determine if significant differences exist and whether assumptions related
to future loan performance need to be revised.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 37

24 Status of Procedures Related to

Housing Direct Loan Programs The following table summarizes key procedures
that were not

completed as of December 1999, their scheduled due dates from the
implementation plan, and the current status.

Procedure Original due date (if any) Revised due

date (if any) Current status Assess existing cash flow data July 2000 In
process Develop new cash flow model May 2001 Not started Ensure cash flow
model represents regulatory requirements

May 2001 5 Not started Document cash flow model May 2001 5 Not started
Perform sensitivity analyses to identify key cash flow assumptions May 2001
5 Not started Document cash flow assumptions May 2001 5 Not started Assess
cash flow model data reliability Not started Compare estimated to actual
cash flows Not started

5 The due date for this procedure is presented in the plan as being
completed with the development of the new housing direct cash flow model,
which is presented as being completed by contractors. However, there
currently is no contract for this work and no funds were provided for these
purposes.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 38

25 Procedures Related to Housing Direct

Loan Programs The procedure in process for the housing direct loan programs:

An outside contractor was hired to analyze existing cash flow data to
determine if the accounting systems contain the appropriate data to
reasonably estimate future loan performance and identify alternative data
sources if necessary.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 39

26 Procedures Related to Housing Direct

Loan Programs Procedures that remain to be started for the housing direct
loan programs:

Once the results of the contractor's analysis of existing cash flow data are
known, it may be necessary to

temporarily use alternative data sources and develop and prioritize system
change requests

to ensure systems provide essential data to model future loan performance.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 40

27 Procedures Related to Housing Direct

Loan Programs A new cash flow model needs to be developed 6

that will reasonably estimate the cost of the loan programs based on readily
available data or alternative data sources, if necessary.

During the development, ensure that the cash flow model reasonably
represents the cash flows of the loan programs based on the laws and
regulations that govern them.

The logic flow and mechanics of the cash flow model need to be documented.

6 The cash flow model currently used to estimate the cost of the housing
direct loan programs is based on several cash flow assumptions that RD is
unable to adequately support.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 41

28 Procedures Related to Housing Direct

Loan Programs Sensitivity analyses need to be performed to

identify key cash flow assumptions. The programs' cash flow assumptions need
to be

documented, including data sources and calculation methods.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 42

29 Procedures Related to Housing Direct

Loan Programs Data reliability tests need to be performed for key

cash flow assumptions. If it is determined that accounting system data
supporting key cash flow assumptions are not reliable, it may be necessary
to

adjust the values of cash flow assumptions, identify alternative data
sources, and/ or revise cash flow models.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 43

30 Procedures Related to Housing Direct

Loan Programs Estimated loan performance needs to be

compared to actual cash flows to determine if significant differences exist
and whether assumptions related to future loan performance need to be
revised.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 44

31 Status of Procedures Related to All

Loan Programs The following table summarizes key procedures that were not

completed as of December 1999, their scheduled due dates from the
implementation plan, and the current status.

Procedure Original due date (if any)

Revised due date (if any) Current status

Hire additional staff Completed May 2000

Resolve reestimate issues with OMB March 2000 In process Develop policies
and procedures for calculating estimates of credit program costs

February 2000 Not started

Implement procedures to automate the estimation process Not started Conduct
“lessons learned” sessions with HUD regarding the balances
approach

September 1999 January 2000 Not started 7

7 The task force has put this procedure on hold until guidance on the
balances approach is issued by OMB.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 45

32 Procedures Related to All Loan

Programs Procedures completed or in process that apply to all loan programs:

Additional staff were hired by RD to assist with credit reform
implementation.

RD is working to resolve its frequency and timing of reestimate issues with
OMB.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 46

33 Procedures Related to All Loan

Programs Procedures that remain to be started that apply to all loan
programs:

Written policies and procedures for calculating estimates of credit program
costs need to be developed and implemented, including a formal supervisory
review process and a process to ensure that key cash flow assumptions are
reviewed and approved by budget, accounting, and program office staff.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 47

34 Procedures Related to All Loan

Programs Procedures need to be developed and

implemented to automate the credit subsidy estimation process to reduce
workloads and the chance of data entry errors.

Automating this process could include centralizing input sections in the
cash flow

models organized by program instead of by cohort 8 and

linking data files to cash flow model input sections to reduce the need for
manual data entry. 8 A cohort includes those direct loans or loan guarantees
of a program for which a subsidy appropriation is provided for a given
fiscal year even if disbursements occur in subsequent years.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 48

35 Procedures Related to All Loan

Programs The task force needs to conduct “lessons learned”

sessions with the Department of Housing and Urban Development staff to
understand how they implemented the balances approach and automated their
estimation processes.

The task force has put this procedure on hold until guidance on the balances
approach is issued by OMB.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 49

36 Implementation Plan Issues

RD has experienced a large number of delays in meeting scheduled due dates
and several due dates are not established. For example, of the 20 key
procedures remaining to be completed,

6 procedures' due dates were already missed, 4 procedures lack due dates,
and 10 procedures are due in the future.

Further, of the 5 recently completed procedures: 2 procedures were completed
after the due date

and 3 procedures lacked due dates.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 50

37 Oversight Issues

While there were individual efforts by Steering Committee members to
eliminate barriers impeding implementation efforts, the Steering Committee
as a whole

does not meet on a regular basis, was not made aware of missed or changed

scheduled due dates, and was not actively involved in overseeing progress

toward implementing the plan and resolving credit reform weaknesses.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 51

38 Resource Issues

Since April 1999, 2 staff were reassigned internally to work part time on
RD's credit reform implementation efforts; in May 2000, 2 additional staff
were hired to work full- time in this area.

The lack of sustained resource commitment to credit reform has impeded
implementation efforts.

Due to the complexity of the interest calculations for the single family
housing program and the limited resources currently available to the task
force, timely development of this cash flow model is unlikely. Further, no
funds were identified to hire contractors for this work.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 52

39 Conclusions

The majority of the RD's key procedures in the detailed implementation plan
are not completed, and several scheduled completion dates were missed or not
established.

The credit subsidy estimation process could be improved in a more timely
manner if more people were assigned in this area on a full- time basis.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 53

40 Conclusions

The task force's ability to resolve credit reform issues will be limited
without contractor support to develop the housing direct loan program cash
flow model.

Due to the magnitude of the RD's loan programs, an unqualified audit opinion
on USDA's consolidated financial statements will not be possible without the
successful implementation of credit reform for these programs.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 54

41 Recommendations

In order to improve RD's estimates of credit program costs in a timely
manner, we recommend that the Secretary of the Department of Agriculture or
his designee take the following actions:

Commit the necessary staff and/ or funding resources to promptly implement
the procedures included in the detailed implementation plan.

Establish scheduled due dates for all the procedures included in the
detailed implementation plan.

Enclosure GAO/ AIMD- 00- 286R RD's Credit Reform Implementation Page 55

42 Recommendations

Establish a process whereby the Executive Steering Committee routinely
monitors the progress of credit reform implementation efforts. With respect
to the detailed implementation plan, establish a process for the Executive
Steering Committee to

be notified when scheduled due dates will not be met so that corrective
actions can be taken and

approve changes to scheduled due dates.

(913905)
*** End of document. ***