To: Eric R. Zarnikow, Associate Administrator, Office of Capital Access
Grady Hedgespeth, Director, Office of Financial Assistance
From: Debra S. Ritt, Assistant Inspector General for Auditing
Subject: Report on the Assessment of the Community Express Pilot Loan Program Report No. 10-12
This report presents the results of our assessment of SBA's Community Express pilot loan program. The program was designed to increase SBA lending to predesignated geographic areas and to New Market borrowers (i.e., women, minorities, and veterans). SBA, aware that the program's historical credit performance had generally been worse than that experienced by other SBA loan programs, requested that the Office of Inspector General determine whether the program is properly structured to ensure success and to minimize the risk of fraud, waste, and abuse. We were also asked to consider the effectiveness of recent program changes.
To respond to the request, we determined whether: (1) the Community Express program has increased the number and dollar amount of loans provided to select New Market groups, while keeping credit risk at an acceptable level; (2) small businesses receiving Community Express loans benefited from lower interest rates and fees than that of other SBA loan programs, and if lenders adequately justified their fees; (3) technical assistance was provided by lenders to increase the potential success for Community Express borrowers; and (4) the most appropriate measures have been established by which to evaluate program success.
To assess program results, we compared the volume, dollar value, geographic distribution, and other characteristics of Community Express loans to that of other 7(a) loans of $250,000 or less that were approved between May 1, 1999 and June 30, 2009. We also interviewed six lenders who had either stopped or reduced their participation in the Community Express program, and discussed lender participation rates with SBA and the National Association of Government Guaranteed Lenders.
To evaluate the risk of the program, we compared the l2-month Community Express purchase and charge-off rates reported for October 2004 to June 2009 to that of the other 7(a) programs. We also compared the l2-month purchase and charge-off rates for the two most active Community Express lenders, who were responsible for 71 percent of the loan volume (and 33 percent of the dollars loaned), to the rates charged to SBA's 7(a) portfolio averages. We determined whether the rates were uniformly high for all Community Express lenders. We also reviewed lender loan files and interviewed lender personnel to determine if lender underwriting practices increased credit risk. Further, we obtained information on the overall subsidy rates and the net cash flow for the Community Express program and other 7(a) programs.
To evaluate interest rates, we compared the average interest rates that each of the 11 most active lenders charged for their Community Express loans to the rates they charged for their other 7(a) loans. We also compared lender information on loan packaging fees for the Community Express program to the fees they charged for other 7(a) and non-SBA loans, which was available for only 7 of the 11 lenders.
To determine whether borrowers received technical assistance, we reviewed a sample of 99 loans that were disbursed between February 2005 and December 2008, and a sample of 31 loans disbursed from October 2008 (when program changes went into effect) to June 2009. We reviewed the loans for evidence that the technical assistance requirements were met. To corroborate data in the loan files, we interviewed 63 of the 130 borrowers we were able to reach, and 59 of 70 technical assistance providers. Our audit scope and methodology is further detailed in Appendix I, and a listing of the sampled loans is provided in Appendix II.
Finally, to determine whether SBA established the most appropriate performance measures for the program, we interviewed Office of Capital Access (OCA) personnel, and reviewed the FY 2008 and 2009 Federal Government Performance Reports. We conducted the audit between January 2009 and March 2010 in accordance with Government Auditing Standards prescribed by the Comptroller General of the United States.
The Community Express pilot program is a delivery method within SBA's 7(a) loan program, and was established in May 1999 to address the difficulty that New Market small business owners had in accessing capital from traditional lending markets. While SBA has not issued Community Express regulations, the policies and procedures for this program are outlined in several procedural notices and an August 2000 Community Express Program Guide, which have been incorporated into Standard Operating Procedure (SOP) 50 10 5(A), Lender and Development Company Loan Programs. The program generally conforms to the policies and procedures of the SBA Express loan program.
As discussed in program guidance, a major objective of the Community Express program is to combine technical and financial assistance to maximize the success potential of New Market small businesses. Technical assistance, which is to be provided by private concerns and SBA's partners, includes an assessment of borrowers' business weaknesses, the identification of training or counseling to address the weaknesses, and follow-up by the lender to encourage borrowers to complete the training. To offset some of the cost of the technical assistance, SBA provided a guaranty on Community Express loans that is similar to 7(a) loans-up to 75 percent for loans over $150,000 and up to 85 percent for loans of $150,000 or less.(1)
SBA Procedural Notice 5000-1068, Extended and Enhanced Community Express Pilot Program, was subsequently incorporated into SOP 50 10 5(A) and made major revisions to the Community Express program, which became effective in October 2008, These revisions included targeting pre-designated geographical areas in lieu of New Market groups; opening eligibility to all applicants for loans of $25,000 or less; allowing lenders to use SBA's on-line training for technical assistance; requiring that technical assistance be completed prior to disbursing loans of $25,000 or less; and requiring that additional courses be completed for loans disbursed over $25,000.
The OIG issued two prior reports addressing the Community Express program, and a memorandum addressing [FOIA ex. 4] one of the most active Community Express lenders. Management Advisory Report 6-34, Policies and Procedures for the SEA Express and Community Express Loan Programs, issued in September 2006, disclosed that the Community Express Loan Program Guide had not been officially cleared or issued by SBA, and that guidance on credit analysis and scoring was vague. We recommended that SBA promulgate regulations to govern the program, establish criteria for when lenders would be permitted to use credit scoring, identify the range of credit scores considered acceptable by SBA, and revise its guide to identify acceptable credit analysis methods. As of March 31,2009, SBA had taken action on some, but not all, of the recommendations.
Audit Report 7-08, The SEA Express and Community Express Loan Purchase and Liquidation Process, issued in December 2006, disclosed that SBA erroneously paid additional guaranties on 43 loans that were ineligible because lenders provided no evidence that borrowers received the required technical assistance. We recommended that SBA obtain evidence that technical assistance was provided to borrowers or seek recovery for those portions of SBA guaranties in excess of 50 percent. Since the report was issued, SBA has taken action to implement the recommendation.
RESULTS IN BRIEF
While the Community Express program was established to increase loans to New Market groups, only 12 percent of the growth in loans to these groups between fiscal years (FY) 2000 and 2008 can be attributed to the program. Further, Community Express loan volumes have fallen short of annual limits established for 7(a) pilot programs as the program has not been able to attract a sufficient number of new lenders or retain existing ones. As of June 30, 2009, two lenders were responsible for 71 percent of the loans and 33 percent of the dollars disbursed under the program. The most active of the two lenders was placed in receivership and, on April 16, 2010, sold to another bank, raising questions about the future viability of the program. Lenders we interviewed who left the program were discouraged by the greater risk associated with Community Express loans and the program's technical assistance requirements.
Despite the lower loan volumes, SBA believes the Community Express program has helped reach borrowers that are not served by other 7(a) programs, particularly borrowers seeking smaller loans ($25,000 or less), start-up businesses, or borrowers located in geographical areas not served by SBA's other loan programs. However, the other 7(a) loan programs have historically made a significantly higher percentage of smaller loans to New Market groups and start-up businesses than the Community Express program and served the vast majority of the geographical areas serviced by the Community Express program. SBA also believes that Community Express loans are now in greater demand and will become the primary vehicle for reaching new market groups as loan volumes in the SBA Express program, which provided the bulk of loans to new markets, have been declining.
Although SBA anticipated that the Community Express program would assume a high level of credit risk, loan purchases and charge-offs have far exceeded that of other 7(a) loan programs, making it costlier than all other 7(a) programs. As of June 30, 2009, the program had a negative cumulative net cash flow of $102.4 million, and was expected to increase the subsidy rate for the 7(a) program. According to SBA's projections, the Community Express subsidy rates will cause
a .12 percent increase in the overall 7(a) subsidy rate in FY 2010 and a .19 percent increase in the 7(a) subsidy rate for FY 2011.
While many Community Express lenders have contributed to the program's high purchase and charge-off rates, the two most active lenders, [FOIA ex. 4, 8] employed credit scoring practices that may have contributed to loan defaults that have increased the cost of the program. Based on credit scores, the lenders reduced loan amounts by as much as 80 percent, without regard to how the reductions would affect the borrower's business plans, intended use of proceeds, and projected cash flows. One of the lenders used a business credit-scoring model for assessing borrower repayment ability even though it did not meet SBA requirements for using credit scoring. Consequently, we believe 44 percent of the loans may have defaulted due to the lenders' credit scoring practices.
Community Express loan recipients generally were charged higher interest rates by the more active lenders and lower interest rates by the less active lenders. We noted that the two most active Community Express lenders, who were responsible for 71 percent of the loans made under the program, focused on making loans of $25,000 or less. Making these smaller loans allowed them to charge interest rates that were at or near the limit permitted by SBA. Less active lenders tended to provide lower interest rates and to make larger loans. Additionally, loan packaging fees charged for Community Express loans were in line with those the lenders generally charged for other 7(a) loans, but lenders we contacted were unable to justify the basis for their fees.
Finally, while a higher percentage of businesses received technical assistance since SBA made it a requirement for loan disbursement in October 2008, the amount and type of technical assistance provided did not always match borrower needs. On average, borrowers with the least experience were assessed as needing less assistance than those borrowers with greater experience. Where technical assistance was provided on-line or by SBA partners, lenders continued to be compensated with a higher guaranty percentage even though there was no cost associated with the on-line training and services provided by SBA partners.
In January 2010, SBA extended the Community Express pilot through December 2010 without assessing the pilot's success or the cost/benefits of the program. In addition, although the Community Express program has been a pilot since 1999, SBA did not establish measurable goals and outcomes for evaluating the program until FY 2008 and has yet to assess the program's effectiveness. We believe that it is not reasonable to continue a program that is heavily reliant on one high-risk lender for most of the loan activity, and whose costs have had a significant impact on the 7(a) program subsidy rate. We understand that SBA is interested in maintaining the Community Express program as a vehicle for reaching new market groups because the SBA Express program, which has contributed the most to this market, has experienced significant declines in loan volumes. However, we believe that SBA would be better served by focusing its efforts on how to make the SBA Express program work instead of making the Community Express program permanent.
We recommended, therefore, that SBA not extend the Community Express program in its current form when it expires in December 2010 and re-evaluate the need for the program. If SBA retains the program, we recommended that it take steps to increase lender participation; reduce program risk; reconsider how program costs should be financed; clarify the appropriate uses of credit scoring; improve the type and quality of technical assistance provided to borrowers; and establish criteria for assessing borrower technical assistance needs and for measuring the program's success. We also recommended that SBA revise its program procedures to limit its guaranty to 50 percent on loans where technical assistance is provided by SBA partners or SBA's online training; and that it repair $18,960 in guaranties on 4 loans it purchased above the 50-percent guaranty level. Finally, we recommended that SBA annotate 30 loan files for a potential repair of $268,190 should the loans default.
Management concurred with two recommendations (recommendations 3 and 4), and agreed to act swiftly to address the remaining ten recommendations related to terminating or changing the program, which would require more careful consideration. The OIG considers the Agency's actions to be responsive to the report, and agrees that careful deliberation of the audit results is needed to ensure that the needs of the underserved markets are addressed in the most prudent and cost effective manner.
The Community Express Program Has Not Been as Effective as Other 7(a) Programs in Increasing Loans to New Market Groups
While intended to increase loans to New Market groups, between FY s 2000 and 2008, Community Express loans comprised only 12 percent of the growth in overall loan volume and 6 percent of loan dollars disbursed to New Market groups. In comparison, other 7(a) programs made 88 percent of the loan growth and 94 percent of the dollars disbursed to New Market groups, primarily the SBA Express program, as shown in Figures 1 and 2. Data supporting Figures 1 and 2 are provided in Appendix III.
Figure 1. Percentage of New Market Loans by 7(a) Program from FY 2000 to FY 2008
To view Figure 1, please see the attached document.
Figure 2. Percentage of Dollars Disbursed to New Market Groups by Loan Program from FY 2000 to FY 2008
To view Figure 2, please see the attached document.
As illustrated in Figures 3 and 4, the SBA Express program has been responsible for the majority of annual increases in loans to New Market groups, while the Community Express program has contributed the least of all of the 7(a) programs to the growth in New Market loans. While the Community Express program reached its highest New Market loan volume in FY 2007, this level was only 17 percent of the New Market loans approved that year. The SBA Express program and the Other 7(a) Delivery Methods collectively produced more New Market loans annually with the exception of FY 2008. Data Supporting Figures 3 and 4 are provided in Appendix III.
Figure 3. New Market Loan Volumes by SBA Program from FY 2000 to FY 2008
Figure 4. Loan Amounts Disbursed to New Market Groups from FY 2000 to FY 2008
To view Figures 3 and 4, please see the attached document.
Although the Community Express program has historically comprised a relatively small percentage of the loans disbursed to New Market groups, in FY 2008 the Community Express share of loans to New Market groups increased to 22 percent. This increase in market share; however, was not due to a higher volume of Community Express loans, but instead attributable to a decrease in the overall volume of SBA 7(a) loan resulting from the economic downturn. Economic conditions led to a decline in lending in FY 2008, most notably in the SBA Express program, which had generated the largest percentage of loans to New Market groups.
Further, with one exception, Community Express loan volumes never approached annual loan limits for 7(a) pilot programs. By statute, the percentage of loans that can be made under a 7(a) pilot program is limited to not more than 10 percent of the total number of 7 ( a) loans guaranteed in any fiscal year. As shown in Table I below, a comparison of program loan levels to that of the annual loan limits disclosed that the volume of loans during the first 7 years of the program fell significantly below maximum loan limits. However, in FY 2008, as a result of a significant reduction in loan production for all 7(a) loan programs, the annual loan limit was significantly reduced; closing the gap between the annual loan limit and Community Express loan volumes that year.
Table 1. Comparison of 7(a) and Community Express Loan Approvals From FY 2000 to FY 2009(2)
To view Table 1, please see the attached document.
Additionally, in FY 2009, SBA limited the number of Community Express loans that each lender could make monthly in order to comply with the statutory limit. This had the effect of curtailing the loan volumes of all lenders, and may have further discouraged new lenders from participating in the program.
The Community Express Program's Loan Volume Was Impacted by the Lack of Lender Participation and Competing Programs
The Community Express program has suffered from a lack of lender participation. Only 275, or 6 percent, of the more than 4,438 lenders authorized to make 7(a) loans participated in the Community Express program from its inception through June 30, 2009. Of the 275 participants, two lenders were responsible for more than 71 percent of the loans and 33 percent of the dollars disbursed as of June 30, 2009. Lender participation was higher in the SBA Express program, which offered lenders the ability to make larger loans and charge higher interest rates than the Community Express program, while still using an expedited lending process. As of June 30, 2009, 1,957, or 44 percent of all 7(a) lenders participated in the SBA Express program even though it offered a 50-percent guaranty as opposed to the 75-percent or 85-percent guaranty offered by the Community Express program. SBA Express lenders are also not required to provide technical assistance, which is an additional expense that Community Express lenders must bear if they do not use SBA partners or SBA on-line training.
To determine why lender participation had declined, we interviewed six lenders who stopped making Community Express loans, which revealed that:
Two lenders did not have sufficient borrower demand for Community Express loans;
Two were uncomfortable with the technical assistance follow-up requirement, especially the repercussions if the follow-up was not done or was improperly done;
One believed the process involved too much paperwork without assurance of loan approval; and
One lender preferred to make SBA Express loans because it believed they were less risky than Community Express loans.
Due to limited lender participation in the Community Express program, effective October 1, 2008, SBA made program changes to attract new lenders, including contacting many of the small volume lenders to discuss future plans for their use of the program. As of June 30, 2009, 24 new lenders joined the program, who collectively originated 47 loans. While this is a large increase in lenders for the Community Express program, it is a relatively small number compared to the 4,438 lenders participating in other 7(a) programs.
Further, the most active lender in the program was put into receivership by the Federal Deposit Insurance Corporation and, on April 16, 2010, sold to another bank. This occurrence will adversely impact the volume of loans made under the program as the new lender has stated it will not participate in the program due to the higher default risk and the greater servicing requirement. Unless SBA succeeds in growing lender participation, the viability of the program is doubtful. Because SBA has not been able to attract a significant number of other lenders, it may want to consider combining the Community Express program with other programs that service the same market groups.
The Community Express Program Provides Loans to Smaller Borrowers and in Geographical Areas Served by Other 7(a) Programs
SBA officials believe that while fewer New Market loans are made under the Community Express program than that of other 7(a) programs, the Community Express program has helped reach borrowers that are not served by other 7(a) loan programs. For example, SBA officials have reported that the program helps meet the needs of borrowers seeking smaller loans ($25,000 or less). As shown in Table 2, the vast majority of Community Express loans made between FY 2000 and FY 2008 were $25,000 or less. However, other 7(a) programs made a significantly greater percentage of New Market loans in this size category than the Community Express program.
Table 2. Size of New Market Loans Approved from October 2000 to June 2009
To view Table 2, please see attached document.
Further, SBA officials believe that one group served predominantly by the Community Express program is start-up small businesses. However while 42.5 percent of Community Express loans made since the start of the program in October 1999 to June 2009 went to start-up New Market businesses, these loans comprised only 5.2 percent of all New Market 7(a) loans made to start-up businesses during the same period. This would indicate that other 7(a) loan programs are predominantly meeting the needs of start-up New Market borrowers, and that the Community Express program is not significantly contributing to the financing of start-up business concerns.
We also examined the geographical areas served by the Community Express and other 7(a) loan programs between FY 2000 and June 2009. Our analysis showed that other 7(a) programs were largely serving the same areas that were serviced by the Community Express program. Specifically, the other 7(a) loan programs made loans to businesses in the same zip codes as that of Community Express program 93 percent of the time. Only 634 of the 8,666 zip code areas serviced by the Community Express program were not reached by other 7(a) programs.
Finally, SBA believes the Community Express program is particularly in demand now that SBA Express loan volumes have significantly declined since 2007. In 2007 SBA Express represented 66 percent of the total 7(a) loan approvals. This declined to about 43 percent under the Recovery Act due to a number of large lenders reducing or discontinuing their participation. Since the SBA Express program made the largest percentage of loans to New Market groups of all the 7 ( a) programs, the volume of loans to this market has significantly decreased in the past couple of years. Therefore, SBA believes that the Community Express program is needed more today than ever. However, we believe that since both the Community Express and SBA Express programs have lender participation issues, SBA would be better served by addressing the reasons for the decline in the SBA Express program, which is a permanent program and less risky than the Community Express program.
The High Cost of the Community Express Program Is Expected to Significantly Increase the 7(a) Subsidy Rate in FYs 2010 and 2011
When SBA initiated the Community Express pilot in 1999, it believed that the program would experience higher costs due to the smaller size of the loans and the target market. Although this was the expectation at the outset, we found the cost of the program to be significantly higher than SBA may have anticipated or would have considered cost effective. Historically, the Community Express program has had a high level of credit risk, as evidenced by the program's purchase rates. As shown in Figure 5, the Community Express 12-month purchase rates increased from about 3.4 percent in October 2004, to about 11 percent in June 2009. Additionally, the 12-month purchase rate(4) for the Community Express program significantly exceeded that of all other 7(a) loan programs.
Figure 5. 12-Month Purchase Rates by Loan Program from October 2004 to June 2009
To view Figure 5, please see attached document.
Further, as shown in Figure 6, the program's 12-month charge-off rates, which represent losses incurred by SBA after loan defaults and collateral liquidation, increased from about 1.77 percent in June 2004 to almost 8 percent in June 2009, and consistently exceeded the charge-off rates for all other loan programs during that period. As of June 30, 2009, SBA had charged off a total of $68.3 million in Community Express loans. Data supporting Figures 5 and 6 are provided in Appendix IV.
Figure 6. 12-Month Charge-Off Rates by Loan Program from October 2004 to June 2009
To view Figure 6. please see the attached document.
We compared the 12-month purchase rates of the two most active lenders with that of the other Community Express lenders, and found that the rates were uniformly high among lenders. For example, as of June 30,2009, the 12-month purchase rates for the two most active lenders were 7.5 percent and 7.7 percent, or about 2 percent to 2.2 percent above the 7(a) portfolio average. For all other Community Express lenders, the rate averaged 8.02 percent-or 2.5 percent above the 7(a) portfolio average. Further, as June 30, 2009, the 12-month charge-off rate for one of the lenders was 6.1 percent, or about 3.2 percent higher than the 7(a) portfolio average-while the other lender's rate was 1.7 percent, or below the portfolio average. For all other Community Express lenders, the 12-month charge off rate averaged 4.8 percent-or 1.9 percent above the 7(a) portfolio average.
As a result of the high purchase and charge-off rates, as of June 30, 2009, the Community Express program had a negative cumulative net cash flow of $102.4 million and a negative cash yield of 12.57 percent. The cumulative cash flow is determined by subtracting all program cash outlays, such as payments to honor guaranties for defaulted loans, from program revenue, such as lender fees and liquidation recoveries. The net cash yield is computed by dividing the cumulative net cash flow by the cumulative guaranteed SBA disbursed amount. No other 7(a) loan program had as high a negative net cash yield as the Community Express program. For example, as of June 30, 2009, the cumulative net cash yield for other programs ranged from a positive 1.6 percent for Rural Express loans to a negative 9.3 percent for regular 7(a) loans.
This negative cash flow will significantly impact the FY 2010 and 2011 subsidy rates for Community Express loans of $25,000 or less, which are projected to be 11.63 and 23.78, respectively. This is significant when compared to the expected FY 2010 and 2011 subsidy rates for similar-sized loans made by all other 7(a) programs, which are projected to be 3.80 and 5.23, respectively.
Moreover, the high subsidy rate of the Community Express program, which is largely comprised of smaller loans, is projected to significantly increase the 7(a) subsidy rates for FY 2010 and 2011. Generally, the 7(a) subsidy rate is heavily influenced by larger loans, which have historically performed better, keeping the overall subsidy rate(5) relatively low. However, as presented in Table 3, the Community Express program will cause a .12 percent increase in the overall 7(a) program subsidy rate in FY 2010, which equates to an increase of about $21.8 million that must be borne by 7(a) lenders and borrowers. A similar comparison for FY 2011 shows that the Community Express program will increase the overall 7(a) subsidy rate by .19 percent or by about $31.2 million. The subsidy rate increases represent funds that could be put to better use should SBA terminate the Community Express program.
Table 3. Community Express Program Impact on Subsidy Rates
To view Table 3, please see the attached document.
SBA will need to determine whether the higher costs associated with the program are reasonable given that other loan programs are largely serving the intended market. Additionally, SBA will need to determine whether the all 7(a) lenders should continue to absorb the higher costs associated with the Community Express program, or whether the higher costs should be financed through other means, such as increased fees or interest rates charged the borrowers.
Credit Scoring Practices of the Two Most Active Community Express Lenders Have Increased Program Risk
While many Community Express lenders have contributed to the program's high purchase and charge-off rates, the credit practices of the two most active Community Express lenders may have contributed to loan defaults that have increased the cost of the program. [FOIA ex. 4, 8]
In addition, one of the lenders was cited in a Cease and Desist Order from the Federal Deposit Insurance Corporation in Calendar Year 2007 as engaging in unsatisfactory lending and collection practices. This lender received two subsequent Cease and Desist Orders in May 2009 and March 2010, and in April 2010 was closed by its state regulator and sold by FDIC.
According to SOP 50 10 5, lenders participating in pilot loan programs are required to use appropriate, prudent, and generally accepted industry credit analysis processes, which may include a business credit scoring model as long as the lender is using the model for similarly-sized non-SBA guaranteed commercial loans. This credit analysis process is intended to help lenders determine whether there is a reasonable expectation that the borrower will repay the SBA loan in a timely manner.
Although credit scoring is allowed for assessing applicant repayment ability, the two most active Community Express lenders also used the credit scoring process to determine the size of loans they would approve for individual borrowers-an application that SBA may not have originally envisioned. In our first two samples of 99 loans, 76 were subjected to loan size determinations based on credit scoring. Of these, 54 were approved for lesser amounts than originally requested. The reductions ranged from 20 percent to 80 percent of the original loan requests. As shown in Table 4, 24 of the 54 loans, or 44 percent, had defaulted as of June 30, 2009. This default rate is relatively high when compared to the default rate for the remaining 22 loans that lenders approved for the original amounts requested. Only 2, or 9 percent, of the 22 loans that were not reduced defaulted.
Table 4. Loan Reductions Made by the Two Most Active Lenders that Resulted in Defaults
To view Table 4, please see the attached document.
By using credit scoring to determine loan size, lenders did not consider how the reduced loan amounts would affect the small business concerns' cash flow and intended use of loan proceeds needed to succeed. Borrowers requested loan amounts based on business plans, which included financial analysis showing how they would use the loan proceeds. However, in lieu of analyzing the business plans, lenders assigned credit scores to loans based primarily on factors relating to the business principal's personal creditworthiness. The higher the numeric score, the larger the loan amount the borrower could receive. We believe that reducing loan amounts without consideration of these factors increased credit risk because borrowers either had to find alternative sources of capital or modify their business plans, which, in some cases negatively impacted their chances for success. For example:
In September 2007, a borrower requested a $12,500 loan for rent payments and inventory purchases for a start-up specialty market, but was approved for only $5,000. The loan defaulted within 24 months of the disbursement date. According to the borrower, the reduced loan proceeds prevented the purchase of sufficient inventory and caused him to become delinquent on the rent.
In December 2007, a borrower requested a $35,000 loan for a start-up video production company to purchase equipment, materials, and for working capital. However, the borrower was approved for a $15,000 loan, which was transferred into liquidation status in June 2009. According to the borrower, the reduced loan prevented her from obtaining a technical consultant and developing the business website.
In April 2008, a borrower requested a $17,000 loan to buy equipment for a long-haul trucking firm. The borrower was approved for $10,000 and accepted the loan because he thought he could obtain additional capital from another source. The loan defaulted in September 2008. The borrower believed that the reduction prevented him from acquiring all the needed equipment.
SBA officials told us that reducing the loan size provided borrowers with opportunities to obtain capital that otherwise would not have been available at a reduced risk exposure level to SBA and the lenders. Therefore, if the loans defaulted, the associated losses would be nominal. While this may be true, SBA regulations require that in making loan decisions, consideration is given to the borrower's repayment ability and the business' potential for long-term success. Specifically, 13 CFR 120.150 requires that loans be of such sound value as to reasonably assure repayment. In determining repayment ability, lenders are to consider the borrower's ability to repay the loan with earnings from the business and the potential for the firm's longer-term success. However, because in many instances the loans were for specific business equipment that could not be purchased with the reduced loan amounts, the smaller loans reduced the likelihood that the businesses would succeed. Additionally, the loan defaults, regardless of size, can actually harm the borrower's credit rating and ability to acquire further capital.
Additionally, one of the lenders, who was a Small Business Lending Company (SBLC) and is currently one of the most active Community Express lenders, used a business credit-scoring model for assessing borrower repayment ability, even though it did not meet SBA's requirements for using credit scoring. SOP 50 10 5 allows lenders to use business credit scoring models only if they are using the models for similarly-sized non-SBA guaranteed commercial loans. Further, it explicitly prohibits lenders that do not use credit scoring for their similarly sized non-SBA loans from credit scoring SBA loans. SBA's procedure is intended to provide assurance that lenders are applying generally accepted industry credit analysis processes consistent with those used for commercial lending.
As an SBLC, the lender made only SBA-guaranteed loans; and therefore, did not have a "tried and true" credit scoring process that could be relied upon for scoring the SBA-backed loans. Further, as a non-depository financial institution, SBLCs are allowed to sell the guaranteed portion of their SBA loans on the secondary market to maintain liquidity. Therefore, lacking a credit scoring process for nonSBA loans, coupled with the lender's ability to sell the SBA loans on the secondary market, fostered an environment where the lender was able to make bad loans and then profit from their sale on the secondary market. Absent assurance that the lender's credit scoring process was prudent, the Agency should not have permitted the lender to rely on credit scoring to underwrite its SBA loans. Instead, SBA should have required the lender to perform a full cash flow and repayment analysis on all of the loans it approved.
The high number of loan defaults indicates that the two lenders did not use appropriate, prudent, and in the case of the one lender, generally acceptable credit analysis processes. This was further substantiated when in 2007, one of the lenders received a Cease and Desist Order from its Federal regulator for, among other things, engaging in unsatisfactory credit administration and underwriting. Had SBA established more stringent credit requirements for the Community Express pilot, it may have been able to better control program risk. For example, many of the credit factors used in the credit scoring methodology of the two most active lenders are unrelated to the business activity of borrowers, and therefore, are not measures of creditworthiness. These factors include items such as whether the principal owns his/her residence, length of time at location, and monthly gross sales, which are meaningless without the businesses' monthly expenses. Further, although SBA has taken enforcement action against one of the lenders, it has not taken action against the second lender, and both continue to maintain their delegated lending authority.
Interest Rates and Fees Charged for Community Express Loans Are Comparable to the Rates Charged for Other SBA Loans
Under the SBA Express program, lenders can charge interest rates between 4.5 and 6.5 percent over the prime rate or LIBOR,(6) depending on loan size. For all other SBA loan programs, including the Community Express program, the interest rates generally cannot exceed 2.75 percent over the prime rate or LIBOR, unless the loan amount is $25,000 or less. Loans of this size may carry an interest rate of up to 4.75 percent over the prime rate or LIBOR.
Our review of the loan portfolios of the 11 most active Community Express lenders disclosed that the interest rates charged by these lenders for loans of $25,000 or less were generally the same as the rates they charged for SBA Express and other 7(a) loans. Specifically, a comparison of the interest rates ofloans made by individual lenders on the same day showed no differences in the interest rates charged among the various programs. The rates were also within the allowable program limits. Our results seem to indicate that Community Express borrowers, whom SBA concluded face unique problems, challenges, and risk in starting and developing their businesses, are not receiving an additional interest rate benefit under the Community Express program.
Community Express loan recipients generally were charged higher interest rates by the more active lenders and lower interest rates by the less active lenders. We noted that the two most active Community Express lenders, who were responsible for 71 percent of the loans made under the program, focused on making loans of $25,000 or less. Making these smaller loans allowed them to charge interest rates that were at or near the limit permitted by SBA. Less active lenders tended to provide lower interest rates and to make larger loans. Additionally, loanpackaging fees charged for Community Express loans were in line with those the lenders generally charged for other 7(a) loans, but lenders we contacted were unable to justify the basis for their fees.
Loan packaging fees charged for Community Express loans also were in line with those generally charged for other 7(a) loans. For example, the packaging fees charged for the 130 sampled loans ranged from $250 to $956, which were similar to the fees charged for other 7(a) loans. However lenders were unable to justify the basis for their fees. We asked 7 of the 11 most active Community Express lenders for which packaging fee information was available to provide formal calculations supporting their fees. However, none of the seven were able to do so. They explained that their fees varied, depending on cost components, such as credit reports, administrative time used to review and approve the loan, processing of the application, and the complexity of the loan itself.
Improvements Are Needed to Ensure that Technical Assistance Requirements Are Adequately Assessed and that a Higher SBA Guaranty is Justified
Under the Community Express program, participating lenders are required to ensure that an approved provider clearly identifies the technical assistance needs of the borrower and develops and implements an appropriate plan. Further, the lender must follow up to strongly encourage the borrower's completion of each milestone in the technical assistance plan. Despite these requirements, we determined that prior to October 2008, lenders had not consistently ensured that borrowers completed the required technical assistance. In October 2008, when SBA made borrower completion of the technical assistance a requirement for loan disbursement, the percentage of lenders providing technical assistance significantly increased. However, we found that the number of training courses assessed for each borrower generally did not correspond to the borrower's level of experience or age of the businesses. Additionally, lenders received higher guaranty fees even though they did not incur additional costs for the technical assistance.
Technical Assistance Completion Improved After October 2008, Although Assessments Did Not Reflect Borrower Experience Levels
As shown in Table 5, a review of 99 loans disbursed prior to October 2008 disclosed that borrowers completed the required technical assistance or received follow-up for only 35, or 35 percent of the loans. After technical assistance became mandatory prior to loan disbursement, the percentage of borrowers completing technical assistance increased. Specifically, borrowers associated with 24, or 77 percent, of the 31 sampled loans disbursed after September 2008, completed technical assistance or experienced substantial follow-up by lenders.
Table 5. Lender Compliance with Technical Assistance Requirements
To view Table 5, please see the attached document.
While technical assistance completion rates increased due to the October 2008 program changes, the amount and type of technical assistance provided did not always reflect borrower needs. Procedural Notice 5000-1013, Technical Assistance Requirements under the Community Express Program, issued in February 2007, states that the amount and type of technical assistance required by a borrower depends on the maturity of the business and the principal's business experience. However, our analysis of 130 loans identified an inverse relationship between the age of the business or experience of the borrower, and the amount of technical assistance identified as being needed. As shown in Tables 6 and 7, the average number of courses assessed per borrower increased with the age of the business and the principal's years of experience. These figures seem to indicate that assessments made by the technical assistance providers did not adequately reflect the technical assistance needed by borrowers.
Table 6. Average Number of Courses Assessed Based on Borrower Experience
To view Table 6, please see the attached document.
Table 7. Average Number of Courses Assessed Based on Age of Business
To view Table 7, please see the attached document.
On average, borrowers with the least amount of experience were assessed as needing fewer courses. For example:
- A start-up firm specializing in substance abuse education whose principal had no business experience was assessed as not needing technical assistance.
- Nine start-up concerns whose principals each had less than I year of business experience were each assessed as needing only one training course.
- Seven borrowers whose principals had operated their businesses for at least 6 years and who each had at least 10 years of business experience were each assessed as needing from 5 to 8 courses.
Because the number of courses generally did not match the borrowers' experience or age of the businesses, SBA should consider whether lenders need better guidance for instructing providers how to assess borrower technical assistance requirements. We noted that SBA has not defined what technical assistance should consist of, and therefore, has no way to measure the quality of assistance provided or its effectiveness. This may have contributed to the variances in technical assistance that were noted by the audit.
According to SBA, lenders do not feel comfortable with the concept of providing technical assistance, as this is outside their normal lending function. With the exception of on-line training offered by one lender, all of the other lenders used private providers or SBA partner organizations to provide the technical assistance. Further, according to a senior SBA official, technical assistance is also outside the normal business activity of OCA that oversees the program. This may explain why SBA has not defined the components of technical assistance, or developed procedures to measure the quality and effectiveness of the assistance provided.
Lenders Received Higher Guaranties than Were Justified
A major objective of the Community Express program is to combine technical and financial assistance to maximize the success of small businesses. To achieve this objective, SBA requires that lenders offer training and counseling to borrowers. While borrowers are ultimately responsible for taking the required training, lenders are strongly encouraged to follow-up with borrowers to ensure they complete the technical assistance requirements. To offset some of the lender's expenses associated with the technical assistance requirements, SBA provides a guaranty of up to 75 percent for Community Express loans over $150,000 and up to 85 percent for loans of$150,000 or less.
Of the 130 loans reviewed, we identified 72 for which lenders received higher guaranties than were justified, either because technical assistance had not been provided or the lenders had not followed up appropriately with borrowers to encourage their completion of the technical assistance requirements. Of the 72 loans, 22 with 85-percent guaranties were purchased which resulted in $71,799 (7) in unsupported costs. In February 2007 SBA established procedures to reduce the guaranty purchase level to 50 percent on loans not meeting the technical assistance requirements. However, it did not follow its procedures when purchasing 4 of the 22 loans that were approved after February 2007. Because the guaranty percentage is directly related to the technical assistance requirement, SBA should recover $18,960, which constitutes unsupported costs for the portion of the loan guaranties in excess of 50 percent.
Of the remaining 50 loans that had not defaulted, 5 are paid in full, 15 were approved prior to May 2007, leaving 30 that will need to be annotated for a possible repair of $268, 190 in the event of default. The $268,190 represents the amount in excess of the 50-percent guaranties.(8) A listing of the purchased and current loans and the reasons for the guaranty repairs or loan annotations are provided in Appendix V.
Finally, in addition to the 34 loans that require a repair or file annotation, we identified 11 that were approved after October 2008, for which the technical assistance was provided by an SBA partner or through SBA's on-line training. Although current program guidance allows lenders to use other sources for technical assistance and be reimbursed the higher guaranty, we believe that SBA should revise program procedures to limit the guaranty to 50 percent whenever technical assistance is provided on-line or by SBA partners.
SBA Established Appropriate Performance Measures But Did Not Evaluate Program Success
Although the Community Express program has been a pilot since May 1999, SBA did not establish measures to evaluate the performance of the program until FY 2008. In September 2008, SBA issued Procedural Notice 5000-1068, Extended and Enhanced Community Pilot Program, which established factors that SBA planned to use to evaluate the program's overall effectiveness and its implications for the Agency's portfolio and subsidy rate. These factors, which we determined were appropriate measures, included the:
Number of loans processed;
Number and percentage of loans processed where the applicant's principal office is located in a distressed community;
Demographics compared to other 7(a) programs;
Number of loans in rural areas;
Number of loans to start-up businesses:
Number of lenders participating and distribution of loans among lenders;
Delinquency, purchase, and loss rates compared to other SBA loan products;
The Office of Credit Risk Management's portfolio analysis and projections based on data in the Loan Management and Accounting System; and
Portfolio management credit scores compared to other 7(a) programs.
Despite establishing performance measures in 2008, SBA has not performed an evaluation of the program. During the audit OCA management stated that a detailed program evaluation would be premature because program changes implemented in October 2008 have not been in place long enough to see significant trends. Management further stated that OCA plans to conduct an evaluation in 12 to 18 months after the implementation of program changes. We made multiple requests for evidence that SBA has evaluated the success of the Community Express program, but no response was received as of the report issue date.
We recommend that the Associate Administrator for the OCA:
Not extend the program, in its current form, beyond the current pilot deadline of December 31,2010, which would result in $53 million in 7(a) subsidy costs that could be put to better use.
Evaluate the need for and viability of the Community Express loan program given that the objectives for New Market lending are being met by other 7(a) loans programs and that one lender will be making most of the loans under the program going forward.
Repair $18,960 in guaranties on the 4 loans purchased above the 50-percent guaranty level for which technical assistance was not completed.
Annotate the loan files for the 30 current loans where technical assistance was not provided for a possible repair of $268,190 should the loans default.
If the program is retained, we recommend that the Associate Administrator:
Develop a plan for increasing lender participation that considers lender feedback on how best to incentivize lenders to participate in the program, which includes the removal of lender loan limits that were imposed in FY 2009. Because SBA has not been able to attract a significant number of other bank lenders, it may want to consider combining the Community Express program with other programs that service the same market groups.
Develop guidance clarifying the appropriate uses of credit scoring and prohibiting the use of credit scores to establish loan size.
Take steps to reduce loan defaults that are increasing the cost of the program, including implementing stricter credit standards aimed at improving the quality of loans made by the most active lender.
If program costs cannot be reduced, determine whether they should continue to be financed through the subsidy rate, which is projected to increase by $31.2 million in FY 2011, or passed onto the borrower through higher fees.
Revise program procedures to limit the guaranty to 50-percent on loans for which technical assistance is provided by SBA partners or SBA's online training.
Provide criteria to lenders for assessing technical assistance needs of borrowers.
Establish annual goals for measuring the success of the Community Express program and measure program accomplishments against these goals.
Establish a process for periodically evaluating the cost/benefit of the program.
AGENCY COMMENTS AND OFFICE OF INSPECTOR GENERAL RESPONSE
On June 11, 2010, we provided OCA and the Office of Financial Assistance (OFA) with the draft report for comment. On August 16, 2010, OCA and OF A submitted a formal response, which is contained in its entirety in Appendix VI. Management commented that the OIG conducted a thorough review and that the report will provide the Agency with a valuable framework for evaluating the Community Express program. Management also concurred with recommendations 3 and 4, and agreed to act swiftly to address the remaining ten recommendations related to terminating or changing the program. To provide careful consideration of these recommendations, SBA advised the OIG that its senior management team has already commenced extensive analysis of the report and program data, and is undertaking a comprehensive assessment of the needs of borrowers in underserved communities and the lenders who serve those communities. They expect to have a well thought out plan for the Community Express program by October 31, 2010.
The OIG considers the Agency's actions to be responsive to the report, and agrees that careful deliberation of the audit results is needed to ensure that SBA acts prudently and cost effectively in addressing the needs of underserved markets for access to capital.
Please provide your management decision for each recommendation on the attached SBA forms 1824, Recommendation Action Sheet, within 30 days from the date of this report. Your decision should identify the specific action( s) taken or planned for each recommendation and the target date(s) for completion.
We appreciate the courtesies and cooperation of the Offices of Capital Access, Financial Assistance, and Credit Risk Management. If you have any questions concerning this report, please call me at (202) 205-[FOIA ex. 2].
APPENDIX I. AUDIT SCOPE AND METHODOLOGY
The objectives for the audit were to determine whether: (1) the Community Express program has resulted in increases in the number and dollar amount of loans provided to select New Market groups; while keeping credit risk as an acceptable level; (2) small businesses receiving Community Express loans benefit in terms of interest rates and fees as compared to other SBA loan programs, and if lenders can adequately justify their fees; (3) technical assistance (TA) has been provided by lenders to increase the potential for success for Community Express borrowers; and (4) SBA has established the most appropriate measures by which to evaluate program success.
To assess program results, we compared the volume, dollar value, and geographic distribution of loans of $250,000 or less to New Market groups that were made by the Community Express program to those made by other 7(a) loan programs between May 1, 1999 and June 30,2009. We also compared the number ofloans made to start-up businesses and the size of loans made under the Community Express program to those made under other 7(a) loan programs during the same period. Our comparative analyses were based on information in SBA's Loan Accounting System. We also interviewed five lenders who had stopped participating in the Community Express program to determine their reasons for leaving the program, and discussed lender participation rates with SBA and the National Association of Government Guaranteed Lenders.
To evaluate the risk of the program, we compared the 12-month Community Express purchase and charge-off rates that were reported for October 2004 to June 2009 in SBA's Loan Accounting System to that of the other 7(a) programs. We also compared the 12-month purchase and charge-off rates of the two most active Community Express lenders who were responsible for 71 percent of the loan volume to the rates charged to SBA's 7(a) portfolio averages. To determine whether the purchase and charge-off rates were uniformly high among Community Express lenders, we compared the rates for the two lenders with that ofthe other Community Express lenders. We also reviewed SBA's risk ratings for the two lenders and lender loan files, and interviewed lender personnel to determine if lender underwriting practices increased credit risk. We evaluated SBA policies and procedures for sufficient internal controls to protect against unnecessary losses and fraud. To determine whether loans defaulted because borrowers received smaller loans than requested, we interviewed selected borrowers.
Further, we obtained information from the Chief Financial Officer on the overall subsidy rates for loans of $25,000 or less and the net cash flow for the Community Express program and other 7(a) programs to determine how the cost of the Community Express program would impact the overall subsidy rate for the 7(a) program in FY 2010 and 2011.
To evaluate interest rates, we compared the average interest rates that each of the 11 most active lenders charged for their Community Express loans to the rates they charged for their other 7(a) loans. We computed average interest rates based on information reported in SBA's Loan Accounting System. We also compared lender information on loan packaging fees for the Community Express program to the fees they charged for other 7(a) and non-SBA loans, which was available for only 7 of the 11 lenders. The loan packaging fees for Community Express loans was obtained from SBA loan files, and the fees charged for other 7(a) loans was obtained through interviews with personnel at the 11 Community Express lenders.
To determine whether borrowers received technical assistance, we reviewed a sample of 99 loans that were disbursed between February 2005 and December 2008, and a sample of 31 loans disbursed from October 2008 (when program changes went into effect) to June 2009. We reviewed the loans for evidence that the borrower needs were assessed, action plans were developed to address the needs, and the borrowers completed either the technical assistance or the lenders followed up with the borrowers. To corroborate data in the loan files, we interviewed 63 of the 130 borrowers we were able to reach, and 59 of 70 technical assistance providers. We made on-site visits to one borrower and one TA provider. In addition, we judgmentally sampled 10 purchased Community Express loans to evaluate SBA's purchase review procedures. We requested documentation from the applicable loan servicing center supporting the decision to honor or not honor the loan guaranty.
Finally, to determine whether SBA established the most appropriate performance measures for the program, we interviewed Office of Capital Access (OCA) personnel, and reviewed the FY 2008 and 2009 Federal Government Performance Reports. In addition, we reviewed selected monthly Portfolio Analysis Committee reports to identify any analysis made by management regarding the Community Express program. We conducted the audit between January 2009 and March 2010 in accordance with Government Auditing Standards prescribed by the Comptroller General of the United States.
APPENDIX II. LIST OF LOANS REVIEWED
To view Appendix II, please see attached document.
APPENDIX III. NEW MARKET LOANS BY LOAN PROGRAM
To view Appendix III, please see the attached document.
APPENDIX IV. 12-MONTH PURCHASE RATES AND CHARGE OFF RATES BY 7(A) PROGRAM
To view Appendix IV, please see the attached document.
APPENDIX V. LOANS REQUIRING REPAIR OR ANNOTATION
To view Appendix V, please see the attached document.
Appendix VI. Management Comments
To view Appendix VI, please see the attached document.
3 Data is for the first 9 months of FY 2009.
4 A purchase occurs when SBA has honored the guaranty on a defaulted loan. The 12-month purchase rate is computed by dividing total gross purchases during the past 12-months by the total value of the outstanding loan portfolio, plus total gross purchases during the past l2-months. The 12-month charge-off rate is computed by dividing total gross dollars charged-off during the past 12 months by the total gross outstanding plus gross dollars charged-off during the past 12 months.
5 Federal loan programs, such as the 7(a) loan program, have either a positive subsidy rate (i.e., incurs cost to the Government) or a negative subsidy rate (i.e., makes a profit) before administrative costs are considered.