. United States General Accounting Office Briefing Report to Congressional &A0 Requesters September 1990 GUAIZANTEED STUDENT LOANS. Profits of Secondary Market Lenders Vary Widely $ GAO/HRD-!IO-13OBR 8 GAO United States General Accounting Office Washington, D.C. 20548 Human Resources Division B-235076 September 28, 1990 The Honorable Edward M. Kennedy Chairman, Committee on Labor and Human Resources United States Senate The Honorable Augustus F. Hawkins Chairman, Committee on Education and Labor House of Representatives The federal government subsidizes higher education loans to students. In fiscal year 1988, federal interest subsidies for Stafford student loans (formerly Guaranteed Student loans) were about $22 billion. Lenders, such as banks, savings and loan associations, and credit unions, make below-market rate loans (generally 8 percent) to students and bill the federal government for the interest subsidies. In 1986, when the Congress reduced the federal subsidy rate by 0.25 percent for most new loans, lenders warned that resulting profit reduc- tions would make the guaranteed loans unattractive investments. To provide a better basis for determining the appropriate subsidy rate on student loans, you requested us to determine . the lenders’ rates of return or profitability on Stafford loans in their portfolios, . the reasons for varying levels of profitability among institutions that hold such loans, and . the effect of the 1986 subsidy reductions on these lenders’ profitability. As agreed with your offices, our report focuses on the activities of lenders that purchase Stafford loans in the “secondary market.” These lenders purchase the loans from originating lenders (those that made the loans), thereby providing them money to make new loans. Originating lenders’ portfolios may contain many kinds of loans-such as home mortgages, auto loans, and credit card receivables. In contrast, many lenders in the secondary market either deal almost exclusively with stu- dent loans or separately account for their student loan activities. While secondary market lenders may not be representative of originating lenders, they are more likely to maintain the financial data we needed to determine the profitability of their student loan business. Page 1 GAO/I-lRD9O-13oBR Profitability of Guaranteed Student Loans 8235076 We judgmentally selected 10 institutions that are major loan holders in the three main kinds of secondary markets for student loans: l Commercial banks-Chase Manhattan Bank and Wachovia Bank and Trust Company. l The federally chartered secondary market-the Student Loan Mar- keting Association, known as Sallie Mae. l State-level agencies or institutions- the California, Colorado, Indiana, Nebraska, Pennsylvania, and Virginia agencies, and’& New England Education Loan Marketing Corporation (Nellie Mae), which serves four New England states. The banks and Sallie Mae are for-profit institutions; the state institu- tions are not.’ Information that lenders reported to the Department of Education indicates that these 10 institutions held (1) about 34 percent of all Stafford loans outstanding at the end of fiscal year 1988 and (2) made about 71 percent of all secondary market purchases of Stafford loans during the year. The Congress has changed the level of interest subsidies paid to lenders several times since the inception of guaranteed student loan programs in 1965. Effective October 1, 1980, the subsidy for lenders using financing for which interest is taxable was set at the difference between the interest rate paid by students-generally 8 percent-and a rate 3.5 per- cent above the yield on 91-day Treasury bills. Subsidy levels for Staf- ford loans financed from tax-exempt sources on or after that date were set at one-half of the subsidy for taxable financed loans, provided total interest paid to lenders was at least 9.5 percent. In 1986, the Gramm-Rudman-Hollings budget sequester temporarily reduced the subsidy rate factor for new loans made between March 1 and September 30, 1986, from 3.5 to 3.1 percent. The reduction applied to the first four quarterly subsidy payments for each loan. Subse- quently, the Higher Education Amendments of 1986 set the subsidy rate factor at 3.25 percent for new loans made after November 15, 1986, with funds obtained from taxable sources. Subsidies for loans purchased with tax-exempt funds were not affected by either of the 1986 revisions. 1Although some of the institutions we reviewed are nonprofit entities and do not earn “profits” as such, we use the term “profitability” of student loans as the difference between income earned on the loan portfolios and the costs associated with financing and servicing the loans, the costs of operating the agency, and applicable taxes. Page 2 GAO/HRD!Kk130BR Profitability of Guaranteed Student Loans B235076 We analyzed records obtained from the Department and the 10 institu- tions for the four fiscal years 1985-88, and interviewed Department and lending institution representatives and other knowledgeable parties. We conducted our work between January 1988 and February 1990 in accor- dance with generally accepted government auditing standards. A more detailed description of our methodology is in appendix I. Annual after-tax rates of return varied considerably during fiscal years Results of Our 1985-88 among and within the institutions we reviewed. Sallie Mae, the Analysis two commercial banks, and the Indiana agency were profitable during each of the 4 years. Five other secondary market lenders experienced losses in at least one year during the period.” The Pennsylvania agency had losses in all four fiscal years, while the other four lenders had annual rates of return ranging from a profit of 1.24 percent of their out- standing Stafford loan portfolios to a loss of 3.31 percent. (See pp. 21- 25.) In 1988, secondary market lenders’ net rates of return varied within a range of 4.26 percentage points of outstanding loans. Profit variations were due primarily to differences in the lenders’ financing, servicing, operating, and other costs, which varied within a range of 3.86 per- centage points. In contrast, gross revenues as a percentage of out- standing loans varied by only 1.35 percentage points. The 1986 subsidy reductions have had little effect on lenders’ revenues to date. The variations in profitability among lenders indicate that revenue and cost information does not provide a sufficient basis for determining appropriate subsidy levels. In fact, profitability by itself is not the only determinant of lender participation. The loan portfolios of all but 1 of the 10 institutions increased over the 4-year period, including the hold- ings of 4 agencies that were unprofitable in at least one of the years. These not-for-profit agencies were established for such purposes as serving as a secondary market for all lenders in their service areas by purchasing all loans offered without regard to risk or potential profitability. 2Colorado provided cost information, but did not provide other information needed to compute profitability. Page 3 GAO/HRD9O-13OBR Profitability of Guaranteed Student Loams B-235076 Cost Variations The institutions’ financing costs, principally interest, accounted for Significantly Affected about 76 percent of total costs in 1988. Costs as a percentage of out- standing loans varied within 1.34 percentage points. The factors Profitability affecting their financing costs included the timing, maturity, and mix of taxable and tax-exempt financing, and the mix of fixed and variable rate financing. Costs unrelated to financing-servicing, operating, and other costs- varied by 3.71 percentage points in fiscal year 1988. These costs were lowest for Sallie Mae (1.42 percent) and highest for the California agency (5.13 percent). The higher costs at several institutions were due in part to unique events or circumstances. For example, California’s 1988 costs included a provision for future losses on delinquent loans of 3.34 percent-the agency may incur significant losses if the Department of Education or the state guaranty agency3 determines that certain of its delinquent loans were not properly serviced and refuses to pay default claims. The Colorado agency’s 3.58-percent rate was caused in part by expenses related to its transition from in-house servicing of loans to a contract arrangement. Interest Subsidy The 1986 reductions had little, if any, effect on the institutions’ profit- ability, primarily because they applied to only a small portion of their Variations Had Little 1988 outstanding loans. The Gramm-Rudman-Hollings budget sequester Effect on Profitability reduction in the subsidy rate was temporary and applied only to new loans made between March 1 and September 30, 1986. The reduction to 3.25 percent required by the Higher Education Amendments of 1986 applies only to loans made after November 15, 1986, and financed with taxable funds. On average, these changes applied to about 18 percent of the Stafford loans held by the 10 institutions at the end of fiscal year 1988. We estimate that the maximum reduction in overall profitability for any institution was 0.1 percent of outstanding loans in 1988. The reductions had no effect on the Colorado and Pennsylvania agencies, which relied entirely on tax-exempt financing during the year. However, for some institutions in some years, the reductions could be significant. For example, the reduction for one agency was 0.1 percent of outstanding loans compared to its rate of return for that year of 0.29 percent. 3Borrowers’ interest and loan principal payments are guaranteed by guaranty agencies, which are in turn insured by the Department. Page 4 GAO/HRD90130BR Profitability of Guaran teed Student Loans F&235076 The 1986 subsidy reduction of 0.25 percent can be expected to reduce revenues more in the future as (1) loans subject to the lower subsidy rate make up more of the taxable funded portions of portfolios and (2) state limits on the use of tax-exempt debt cause state agencies to rely more on taxable borrowing. However, the effect of the subsidy reduc- tion on the institutions’ profitability will likely continue to be minor compared with the effect of variations in financing, servicing, and other costs. (See p. 28.) Loans Financed With Tax- Agencies that use tax-exempt funds to purchase Stafford loans at times Exempt Funds Can Earn earn higher interest revenues than do lenders using taxable funds to finance their loan portfolios because: More RevenueThan Other . Loans l The 1986 reduction of 0.25 percent in the subsidy rate factor did not apply to student loans made or purchased with tax-exempt funds, which continue to receive subsidies at the pre-1986 level. At the end of fiscal year 1988, such loans accounted for about 55 percent of the Stafford loan portfolios of all seven state secondary market institutions studied. l The Higher Education Act provides loans purchased with tax-exempt funds a minimum rate of return of 9.5 percent. In periods of relatively low interest rates, lenders receive higher rates of interest on these loans than on loans made or purchased with taxable funds that are not pro- tected by an interest rate floor. For example, during fiscal year 1986 the gross return on tax-exempt financed loans remained at the floor of 9.5 percent, while the return on taxable financed loans to first-time bor- rowers ranged from 8.75 to 10.88 percent. (See p. 26.) Eliminating the 9.5-percent revenue floor and reducing the subsidy rate factor on tax-exempt financed loans to 3.25 percent would be consistent with the treatment of loans financed with taxable funds and would reduce federal interest subsidies. However, such actions would reduce revenues of state-level agencies, which included the least profitable institutions in our study. The Department of Education and 9 of the 10 secondary market lenders Agency Comments we reviewed provided written comments on a draft of this report. The Department and several of the institutions provided technical com- ments, which we incorporated where appropriate. Several lenders also noted that the institutions vary widely in their operations and profit- ability, and some advised us that their costs have increased since the Page 5 GAO/HRMO-13OBR Profitability of Guaranteed Student Loans Et-235076 completion of our study. Our evaluation of their comments begins on page 36. Their comments appear in appendixes III through XII. We are sending copies of this briefing report to the Department of Edu- cation, other congressional committees, and other interested parties. Should you wish to discuss its contents, please call me on (202) 275- 1793. Major contributors to this report are listed in appendix XIII. Franklin Frazier Director, Education and Employment Issues Page 6 GAO/HBD-B@13OBR lbfltability of Guaranteed Student Loana . Page 7 GAO/HRDWMOBB Profitability of GuaranteesI Student bona Contents Letter 1 Profitability of 12 Stafford Student Objectives Scope and Methodology 12 13 Loans Held by The 10 Institutions’ Loan Holdings Have Increased 20 Secondary Markets Profitability Varied Widely Among Secondary Market 21 Lenders Varied Widely Revenues Were Similar 25 Costs Varied 29 Conclusions 35 Agency Comments and Our Evaluation 36 . Appendix I 38 Methodology Data Collection Data Validation 39 39 Data Limitations 40 Appendix II 43 Data Supporting Figures Appendix III 50 Comments From the Department of Education Appendix IV Comments From Sallie Mae Appendix V Comments From Chase Manhattan Page 8 GAO/HRDSl%13OBR Profitability of Guaranteed Student Loans Content.3 Appendix VI Comments From Wachovia Appendix VII .58 Comments From the California Agency Appendix VIII 59 Comments From the . Indiana Agency Appendix IX 60 Comments From the Nebraska Agency Appendix X Comments From Nellie Mae Appendix XI Comments From the Colorado Agency Appendix XII Comments From the Pennsylvania Agency Appendix XIII Major Contributors to This Briefing Report Page 9 GAO/HRD9@13OBR Profitability of Guaranteed Student Loam Contents Tables Table 1: Stafford Loan Holdings Generally Increased 21 Despite Unprofitable Operations (Fiscal Years 1986- 88) Table II. 1: Ten Institutions’ Loan Holdings Doubled 43 (Fiscal Years 1985-88) (Data for Fig. 6) Table 11.2:Nine Institutions’ Profitability Varied Widely 43 (Fiscal Year 1988) (Data for Fig. 7) Table 11.3:For-Profit Institutions Were Consistently 44 Profitable (Fiscal Years 1985-88) (Data for Fig. 8) Table 11.4:Not-for-Profit Agencies’ Returns Varied (1988) 44 (Data for Fig. 9) Table 11.5:State Governmental Agencies’ Returns Varied, 45 but Each Has Had Losses (Fiscal Years 1985-88) (Data for Fig. 10) Table 11.6:Institutions’ 1988 Gross Revenues Were Similar 45 (Data for Fig. 11) Table 11.7:9.5-Percent Interest Revenue Floor Increased 46 Returns on Tax-Exempt Financed Loans (Data for Fig. 12) Table 11.8:Loans Subject to Reduced Subsidies Are 47 Increasing (Fiscal Years 1987-88) (Data for Fig. 13) Table 11.9:Costs Varied Among 10 Lenders (Fiscal Year 47 1988) (Data for Fig. 14) Table II. 10: Tax-Exempt and Taxable Borrowing Costs 48 Were Similar (Fiscal Years 1985-88) (Data for Fig. 15) Table II. 11: Proportion of Loans Purchased With Tax- 48 Exempt Funds Has Declined (Fiscal Years 1985-88) (Data for Fig. 16) Table II. 12: Servicing and Operating Costs Varied (Fiscal 48 Year 1988) (Data for Fig. 17) Table II. 13: Taxes Reduced Federal and Commercial 49 Lenders’ Profitability (Fiscal Year 1988) (Data for Fig. 18) Figures Figure 1: Objectives 13 Figure 2: Methodology 14 Figure 3: Reviewed 10 Secondary Markets 15 Figure 4: State Agencies Reviewed 17 Figure 5: How Profitability Is Calculated 18 Page 10 GAO/HRlMO-13OBR Profltabiity of Guaranteed Student L.oans Contents Figure 6: Ten Institutions’ Loan Holdings Doubled (Fiscal 20 Years 1985-88) Figure 7: Nine Institutions’ Profitability Varied Widely in 22 1988 Figure 8: For-Profit Institutions Were Consistently 23 Profitable (Fiscal Years 1985-88) Figure 9: Not-For-Profit Agencies’ Returns Varied (Fiscal 24 Years 1985-88) Figure 10: State Governmental Agencies’ Returns Varied, 25 but Each Has Had Losses (Fiscal Years 1985-88) Figure 11: Institutions’ 1988 Gross Revenues Were Similar 26 Figure 12: 9.5-Percent Interest Revenue Floor Increased 27 Returns on Tax-Exempt Financed Loans (Fiscal Years 1985-88) Figure 13: Loans Subject to Reduced Subsidies Increased 29 (Fiscal Years 1987-88) Figure 14: Costs Varied Among 10 Lenders (1988) 30 Figure 15: Tax-Exempt and Taxable Financing Costs Were 32 Similar (Fiscal Years 1985-88) Figure 16: Proportion of Loans Financed With Tax- 33 Exempt Funds Has Declined (Fiscal Years 1985-88) Figure 17: Servicing, Operating, and Other Costs Varied in 34 Fiscal Year 1988 Figure 18: Taxes Reduced Federal and Commercial 35 Lenders’ Profitability in Fiscal Year 1988 Page 11 GAO/HRD9@13OBR Proiitability of Guaranteed Student Loans Profitability of Stafford Student Loans Held by SecondaryMarkets Varied Widely The costs of federal interest subsidies for guaranteed student loans rose Objectives from $1.3 billion in fiscal year 1980 to $3.3 billion in fiscal year 1985. In 1986, the Congress reduced the interest subsidy rate by 0.25 percent for most new loans. At that time, some lenders indicated that the reduction would make student loans unattractive. To assess the profits lenders were making on these loans and to provide a basis for assessing the ade- quacy of federal interest subsidies, the House Committee on Education and Labor and the Senate Committee on Labor and Human Resources asked us to determine (1) the profitability of Stafford loan portfolios held by major secondary market institutions receiving federal interest subsidies on these loans, (2) the reasons for variations in the profit- ability of these portfolios, and (3) the effect of the 1986 reduction in the interest subsidy rate on their profitability. (See fig. 1.) After discussions in early 1988 with the committees, we focused our efforts on the profitability of Stafford loan portfolios of those lenders that make up the secondary market for student loans, that is, financial institutions that purchase Stafford loans from banks, savings and loan associations, credit unions, and other financial institutions that make loans to students. In contrast to the originating lenders, whose portfolios may contain many kinds of loans-such as home mortgages, auto loans, and credit card receivables-many lenders in the secondary market either deal almost exclusively with student loans or separately account for their student loan activities. Page 12 GAO/HRD-SO-130BR Profitabihy of Guaranteed Student Loans Profitability of Stafford Student Loans Held by Secondary Markets Varied Widely Fiaure 1 GAO Objectives l Determine profitability of student loans . l Examine variations l Determine effect on profits of 1986 interest subsidy reduction At the beginning of our review, we held a conference with secondary Scopeand market officials and others knowledgeable in student loan finance as we Methodology developed our review methodology. We also contracted with an expert on government-sponsored enterprises to identify and describe the legal and institutional factors that affect the three major kinds of institutions that make up the secondary market for Stafford loans-commercial banks, state agencies, and the federally chartered Student Loan Mar- keting Association (Sallie Mae)-and their reasons for participating in the secondary market for student loans. (See fig. 2.) Page 13 GAO/HRD90-130BR Profitability of Guaranteed Student Loans Profitability of Stafford Student Loana Held by Secondary Markets Varied Widely Figure 2 GAS Methodology l Held conference with major loan purchasers to discuss review approach . l Used consultant to identify and assess factors affecting competitiveness among purchasers l Analyzed financial activities of 10 major purchasers We focused on the activities of 10 judgmentally selected Stafford loan secondary market institutions that were major loan holders during fiscal years 1985-88. At the end of fiscal year 1988, secondary market institu- tions held about two-thirds of all Stafford loans. The 10 we analyzed held about $13.5 billion, or one-third of all Stafford loans, and made about 71 percent of all reported secondary market purchases during fiscal year 1988. (See fig. 3.) Page 14 GAO/HRD-9lX13OBR Pmfitmbtkity of Guuanteed Student Loans Profitability of Stafford Student Loans Held by Secondary Markets Varied Widely Figure 3 GACI Reviewed IO Secondary Markets l Sallie Mae 0 Commercial banks . *Chase Manhattan l Wachovia l State agencies lDesignated not-for-profit (CA, IN, NE, Nellie Mae) @Government agencies (CO, PA, VA) They represent the three major kinds of secondary market institutions: . The federally chartered Student Loan Marketing Association (Sallie Mae) is a stockholder-owned, for-profit corporation, established by the Congress as a national secondary market for federally guaranteed stu- dent loans. With a portfolio of about $9.4 billion of Stafford loans at the end of fiscal year 1988, Sallie Mae is by far the largest holder of these loans, l Commercial banks are stockholder-owned, for-profit lending institu- tions. We selected two banks that, in addition to purchasing Stafford Page 16 GAO/HRD90-13OBB Profitability of Guaran teed Student Loana Profitability of Stafford Student Loam Held by Secondary Markets Varied Widely loans, were also major originators of such loans. Chase Manhattan Bank’ and Wachovia Bank and Trust Company together held $1.1 billion of Stafford loans at the end of fiscal year 1988. l State agencies are either governmental or not-for-profit agencies. They purchase student loans from private lenders, often for resident-bor- rowers of the states in which they were established. A principal feature that differentiates them from other secondary lenders is that they may use tax-exempt financing to purchase Stafford loans. We selected seven state agencies to provide a cross-section of the different types of state secondary markets. Three of these are private, not-for-profit agencies that serve single states (California, Indiana, and Nebraska); one is a pri- vate, not-for-profit agency that serves Massachusetts, New Hampshire, Connecticut, and Rhode Island (New England Education Loan Marketing Corporation, or Nellie Mae); and three are state governmental agencies that serve single states (Colorado, Pennsylvania, and Virginia). (See fig. 4.) To determine Stafford loan profitability for the 10 institutions in each of the four years during the 1985-88 period, we analyzed their Stafford loan costs and revenues expressed as a percentage of the average bal- ance of their outstanding Stafford loan portfolio for each year.2 Although the seven state agencies do not generate “profits” as such, we use the terms “profit” and “loss” to refer to each of the 10 institutions’ net rates of return on Stafford loans (net income or loss as a percentage of the average balance of outstanding Stafford loans). 1We excluded from our analysis Stafford loans held by Chase Lincoln First Bank and Chase Man- hattan, St. Thomas. 2Colorado provided cost data, but did not provide other data needed to compute profitability. Page 16 GAO/HRD9@13OBR Profitability of Guaranteed Student Loans Profitability of Stafford Student Lonna Held by Secondary Markets Varied Widely Figure 4: State Agencies Reviewed W[ Not-for-profit Agencies GowmmsntalA~s m Page 17 GAO/HRMU%13OBR Profitability of Guaranteed Student Loans Profitability of Stafford Student Loans Held by Secondary Markets Varied Widely Figure 5 GAO How Profitability Is Calculated Gross revenues Less costs: . @Costof funds *Servicing costs @Operating costs aTaxes, where applicable Equals profitability, or net rate of return Revenues to lenders, whether they make loans or purchase them in the secondary market, consist mostly of interest paid by students and interest subsidies paid by the Department of Education. Borrowers interest and loan principal payments are guaranteed by 1 of 59 state or nonprofit guaranty agencies, which are in turn insured by the Depart- ment. The federal interest subsidies include (1) students’ interest while they are in school and during grace and deferment periods after they leave and (2) an additional subsidy, referred to as a special allowance Page 18 GAO/HRD!M%13OBR Profitabiity of Gnaran teed Student Loans Profitability of Stafford Student Loans Held by Secondary Markets Varied Widely payment, throughout the life of the loan that is intended to give lenders a near-market interest rate.3 Secondary market lenders incur costs to borrow the funds to purchase and service loans and to pay operating and other expenses. l Costs of funds include lenders’ interest expenses and other costs of issuing debt, such as letters of credit, underwriting, and bond attorneys’ fees. . Servicing costs include the costs of billing, collecting, and accounting for loan payments; encouraging borrowers to make scheduled payments; and filing claims with the guaranty agency when students default. Some lenders service their own loans, while others contract for the servicing of all or a portion of their portfolios. l Operating and other costs include administrative costs and provisions for loan losses. To calculate profits for Sallie Mae and the two commercial banks, we also deducted taxes from revenues. To facilitate comparisons among agencies, we included in our analyses only revenues and costs directly associated with the Stafford loans held by each. We excluded, for example: 9 Arbitrage revenues that state agencies earned by issuing tax-exempt securities and temporarily investing portions of the proceeds in higher yielding investments until they purchase student loans. l Revenues that Wachovia and the Colorado, Indiana, and Pennsylvania agencies, or their affiliates, received for servicing loans held by other lenders. l Revenues that Sallie Mae earned from sales of letters of credit and loans it made to facilitate other lenders’ student loan programs. In addition, some of the 10 institutions failed to provide all of the data we requested. A complete description of our methodology, including data limitations, is in appendix I. 3The interest rate students pay has been 8 percent on loans to first-time borrowers since 1983. For students borrowing Stafford loans for the first time after June 1988, interest will increase to 10 percent after the fourth year of repayment. The special allowance is paid quarterly and, for taxable faced loans, is the difference between the borrower’s interest rate and the average bond equivalent rate on 9lday T~asury bills plus 3.26 percent. Page 19 GAO/HRD9@199BR Profitability of Guaran teed Student Loans Profitability of StatTord Student Loans Held by Secondary Markets Varied Widely The 10 institutions we reviewed were among the 40 largest holders of The 10 Institutions’ guaranteed student loans at the end of fiscal year 1988.4 As shown in Loan Holdings Have figure 6, their outstanding Stafford loans rose from $7.4 billion at the Increased end of fiscal year 1985 (22 percent of all outstanding loans) to $13.5 billion at the end of fiscal year 1988 (34 percent). Sallie Mae was by far the largest holder. Its $5.1 billion student loan portfolio at the end of fiscal year 1985 increased to about $9.4 billion at the end of fiscal year 1988. Four other institutions reviewed were among the 10 largest holders of guaranteed student loans at the end of fiscal year 1988- Nellie Mae, Chase Manhattan, and the Nebraska and California agencies. Figure 6: Ten Institutions’ Loan Holdings Doubled (Fiscal Years 1985-88) 14 Dofm in BHlions 13 4 3 2 1 0 \ W-B 1985 1988 1987 1988 End of F-1 Year State Agencies cbfnmerdal BenILs Sallie Mae Note: Excludes many of the Pennsylvania agency’s Stafford loans (for example, $340 million for 1988) that-while federally insured-were ineligible for federal interest subsidies pnnclpally because the bor- rowers’ incomes exceeded federal maximums. 4At that time, about $40 billion of the $45 billion of outstanding guaranteed student loam were Staf- ford loans. Page 20 GAO/ERB!I@12OBR Profitability of Guaran teed Student Loans Profitability of Stafford Student Loana Held by Secondary Markets Varied Widely As discussed in more detail below, net rates of return varied widely among nine institutions during the 4-year period. However, profitability was apparently not the only factor influencing lenders’ continued partic- ipation in the program. As shown in table 1, of the five agencies that reported losses in at least one year during the period, four increased their loan portfolios substantially while one reduced its portfolio slightly. Each of the lenders that had losses are not-for-profit or state agencies generally established for purposes other than making profits. Some of the reasons for which these lenders were created include (1) to serve all lenders and borrowers in their service areas regardless of the costs and risks of certain kinds of loans and (2) to purchase loans that lenders have difficulty selling to for-profit secondary market institutions. Table 1: Stafford Loan Holdings . Generally Increased Despite Loan holdings Unprofitable Operations (Fiscal Years Years change 1986-88) Institution unprofitablea (percent) Sallie Mae None 82 Chase Manhattan None 11 Wachovia None 135 California 1988b 196 Indiana None 48 Nebraska 1985 230 Nellie Mae 1985-86 258 Pennsvlvania 198588 408 Virainia 1987-88 -8 Talifornia and Chase Manhattan data are for calendar years bNo data on profitability for fiscal years 1985 and 1966. The variations in profitability among the institutions were more often a Profitability Varied result of differences in costs than in revenues. In 1988,s for example, net Widely Among rates of return varied within a 4.26~percentage-point range, from a Secondary Market 3.31-percent loss to a 0.95~percent profit. (See fig. 7.) However, gross revenues as a percentage of outstanding loans varied by only 1.35 per- Lenders centage points (9.03 to 10.38 percent). Costs as a percentage of out- standing loans varied over a broader, 3.34~percentage-point range (9.00 to 12.34 percent). ‘The California agency and Chase Manhattan data are for calendar year 1988. All other data are for the fiscal year ending September 30,1988. Page 21 GAO/‘HFKL9&13OBR Profitability of Guaran teed Student Loans Profitability of Stafford Student Loans Held by Secondary Markets Varied Widely Figure 7: Nine Institutions’ Profitability Varied Widely in 1988 Nef Rate of Rafurn (Percent) -4 Note. Data not avallable for the Colorado agency Among profitable lenders in 1988, Sallie Mae had the highest profit (about 0.95 percent after taxes) and Chase Manhattan had the lowest (0.18 percent). During that year, three of the nine agencies (California, Pennsylvania, and Virginia) had negative net rates of return (losses of 3.31, 0.40, and 0.18 percent, respectively). For-Profit Secondary Sallie Mae, Chase Manhattan, and Wachovia were consistently profitable over the 4-year period, with Sallie Mae’s rates of return being the Market Lenders Were highest and Chase Manhattan’s the lowest. (See fig. 8.) According to Consistently Profitable officials at Chase, profits on the bank’s Stafford loan portfolio were lower because of additional investments made in equipment and staff in anticipation of substantial increases in the size of its student loan opera- tion. Stafford loans made up relatively small portions of the two com- mercial banks’ assets-about 0.88 percent at Chase Manhattan and about 1.48 percent at Wachovia as of the end of fiscal year 1988. In contrast, student loans were a major portion of total assets for Sallie Mae and most of the state agencies. Page 22 GAO/HRD-90-13OBR Profitability of Guaranteed Student Loans Profitability of Stafford Student Loans Held by Secondary Markets Varied Widely Figure 8: For-Profit Institutions Were Consistently Profitable (Fiscal Years 1985 Net Rate of Return (Percent) 88) 2.0 -1 .o -1.5 -2.0 1985 1966 1987 Fiscal Year - Sallie Mae -1-1 Chase Manhattan m Wachovia State Not-for-Profit While Sallie Mae and the banks consistently earned a profit during the 4-year period, the four not-for-profit agencies’ net rates of return varied Institutions’ Net Returns TT--2 - -1 v ar1eu considerably. Although all four had positive returns in 1987, two had losses in earlier years (Nebraska in 1985 and Nellie Mae in 1985 and 1986), and California had a large loss in 1988. (See fig. 9.) According to agency officials: l The Nebraska agency’s 1.71-percent loss in fiscal year 1985 resulted tn part from high interest costs for long-term fixed interest rate securities that the agency had issued in prior years when interest rates were higher. Nebraska lowered its cost of funds considerably, from 10.63 per- cent in fiscal year 1985 to 7.73 percent in fiscal year 1986, by issuing lower yield securities to replace earlier higher yield securities, thereby improving its net rate of return in subsequent years. l Nellie Mae’s losses in fiscal years 1985 and 1986 of 0.43 and 0.57 per- cent, respectively, resulted in part because it did not receive special allowance subsidy payments for some of its loans during these years. Page 23 GAO/IiRD9&130BR Profltabllity of Guaranteed Student Loana Profitability of Stafford Student Loana Held by Secondary Markets Varied Widely These loans were purchased with funds Nellie Mae raised by issuing tax- exempt securities before it obtained approval of its plan for using such funds to finance Stafford loans. The loans were therefore ineligible for special allowance payments until Nellie Mae received state approval. . The California agency incurred a 3.31-percent loss in calendar year 1988 largely because it included in its costs a provision for future losses on delinquent loans of 3.34 percent. The agency may incur significant one- time losses if the Department of Education or the guaranty agency determines that certain delinquent loans were not appropriately ser- viced and therefore refuses to pay default claims. Figure 9: Not-for-Profit Agencies’ Returns Varied (Fiscal Years 198588) Net Rate of Return (Percent) 2 -- 7 mmmmmmZ~OmOO*O* *a--- \\ \ -2 -3 -4 1995 1966 1997 1999 Fiscal Year - California -1-1 Indiana m Nebraska nnnn Nellie Mae Note: California agency data not avallable for fiscal years 1985 and 1986. Page 24 GAO/HBDf@13OBR ProfItability of Guaranteed Student Loans Profitability of Stafford Student Loam Held by !?uxondary Markets Varied Widely State Governmental The Virginia agency earned a profit in fiscal years 1985 and 1986, but incurred losses in 1987 and 1988. The Pennsylvania agency had losses in Agencies Have Had Losses all four years. (See fig. 10.) The Colorado agency did not provide suffi- cient revenue data to determine its profitability during the 4-year period. Virginia’s losses were attributed to its high cost of funds-the highest reported of the 10 institutions in 1986, 1987, and 1988. A Virginia agency official explained that the agency had issued fixed rate tax- exempt bonds at a time when interest rates were higher. The Penn- sylvania agency, as was the case with Nellie Mae, lost potential revenue because it was initially not eligible to receive special allowance pay- ments. It began to receive the subsidy payments in January 1987, after its plan for the use of tax-exempt financing was approved. Figure 10: State Governmental Agencies’ Returns Varied, but Each Has Had 2 Net Rate of Return (Percent) Losses (Fiscal Years 1985-88) 1985 1986 Fiscal Year - Pennsylvania -9-9 Virginia RevenuesWere Similar nues as a percentage of outstanding loans varied in 1988 within a Page 26 GAO/ERD9@13OBR Profitability of Guaranteed Student Loans Profitability of Stafford Student Loans Held by Secondary Markets Varied Widely 1.35-percentage-point range-from 9.03 to 10.38 percent of their Staf- ford loan portfolios6 Figure 11: Institutions’ 1988 Gross Revenues Were Similar 11 Gross Revenue Rate (Percent) 10 9 6 7 6 5 4 3 2 1 0 5 Note: Colorado agency data not available. Although the special allowance payment for tax-exempt financed loans is generally one-half of that for taxable-financed loans, loans financed with tax-exempt funds are guaranteed a gross interest revenue rate of at least 9.5 percent.’ When the Treasury bill rates to which subsidies are tied are relatively low- as was the case in recent years-the revenue rates on tax-exempt financed loans can approach, or even exceed, those on taxable financed loans. Figure 12 illustrates the effect of the 9.5- percent floor on lenders’ gross interest revenues for their tax-exempt funded loans during fiscal years 1985-88. The floor raised agencies’ ‘Some of the variations in revenues resulted from the use of different reporting periods. Two agen- cies used the year ended December 31, and the others used September 30, 1988. Average Treasury bill rates in the quarter ending December 31, 1988, rose above earlier levels, thereby increasing annual interest subsidy revenues for the two agencies. ‘Stafford loans made or purchased with tax-exempt funds before the beginning of fiscal year 1981 earn special allowance payments at the same rate as loans made or purchased with taxable funds. Page 26 GAO/HRNNl-13OBB Profitability of Guaranteed Student Loana Profitability of Stafford Student Loans Held by Secondary Marketa Varied Widely revenues on tax-exempt financed loans in periods of relatively low Trea- sury bill rates. Figure 12: g&percent Interest Revenue Floor increased Returns on Tax-Exempt Financed Loans (Fiscal Years 1985-88) f 11.0 Intenst Rata (Percent) 1999 1999 1967 1999 Fiscal Year B Interest rate paid (with 9.5-percent floor) -1-1 Interest rate had there been no floor Quarterly revenues of lenders who used tax-exempt financing were as much as 1.12 percent higher than they would have been without the interest rate floor in 13 of the 16 quarters during the 4-year period. For example, we estimate that in fiscal year 1988 the seven state agencies we reviewed received about $8 million more than they would have without a 9.5~percent interest revenue floor. For all agencies that use tax-exempt financing, we estimate that the provision increased revenues by about $19 million in that year. However, the institutions that bene- fited from the subsidy reduction exemptions included the least profit- able of the 10 we studied. Furthermore, in 7 of the 16 quarters, Treasury bill rates declined to the point that the 9.5~percent interest revenue floor provided the state agencies higher interest revenue for tax-exempt financed loans than for taxable financed loans, which have no minimum special allowance payment. Page 27 GAO/HRD9@12OBB Profltablllty of Guaranteed Student Loana Profitability of Stafford Student Loans Held by Secondary Markets Varied Widely Federal Cost Reduction Two congressional changes were enacted in 1986 to reduce federal interest subsidy costs that resulted in slightly lower revenues for most Initiatives Have Had Little lenders. The first was temporary; the second remains in effect: Effect on Lenders’ Revenuesto Date l The Gramm-Rudman-Hollings sequester reduced the interest subsidy rate factor used to calculate special allowance payments by 0.4 percent for loans made between March 1 and September 30,1986. This reduction remained in effect for four quarterly payments on each affected loan. l The Higher Education Amendments of 1986 reduced the interest subsidy rate factor for taxable financed loans made after November 15, 1986, by 0.25 percent, from 3.5 to 3.25 percent. This provision applies for the life of these loans. Although the Gramm-Rudman-Hollings reduction applied to all Stafford loans, in practice it did not affect tax-exempt financed loans. The 91-day Treasury bill rates were low enough that these loans earned the min- imum 9.5-percent return provided for by law. In contrast, institutions that held taxable financed loans experienced reductions in revenue due to the sequester. Because of the short duration of the cut (it applied only to loans made during a 7-month period), as of September 30, 1986, this temporary subsidy rate reduction affected no more than 5.1 percent of any of the 10 agencies’ portfolios. The Higher Education Amendments excluded tax-exempt financed loans from the 0.25-percent reduction in the subsidy rate factor. Because many of the loans held by five state agencies, and all of the loans held by two agencies, were made or purchased with tax-exempt rather than taxable financing, the rate reduction had little, if any, effect on their gross revenues. Moreover, because the reduction applies only to loans made after November 15,1986, many of the taxable financed loans in their portfolios were unaffected as of the end of fiscal year 1988. As a result, none of the 10 agencies’ revenues decreased by the full 0.25 per- cent as of September 30,1988. Revenue reductions due to the revised 3.25-percent subsidy rate factor ranged from zero for the Colorado and Pennsylvania agencies, which held only tax-exempt financed loans, to slightly more than 0.1 percent of the loan portfolio balance for Chase Manhattan Bank, which had almost half of its loans subject to the 3.25- percent rate factor. (See fig. 13.) Although Chase Manhattan’s O.l-per- cent revenue reduction is not as significant as the variations in its costs, it is significant when compared to its 1988 net rate of return before taxes of 0.29 percent. Page 28 GAO/HRINML12OBR Profitabiity of Guaranteed Student Loans Profitability of Stafford Student L.oam Held by Secondary Markets Varied Widely Figure 13: Loans Subject to Reduced Subsidies Increased (Fiscal Years 1987-88) 50 Percent of Outstanding Loans Subject to Reduced Subsidies Note: The Colorado and Pennsylvania agencies held only tax-exempt financed loans, which are not subject to the subsidy reduction. As the number of taxable financed loans subject to the reduced subsidy rate increases, the new rate will have a greater effect on lenders’ reve- nues. The rate cut, however, will continue to have no impact on lenders’ Stafford loan portfolios financed with tax-exempt funds. As shown in figure 14, the 10 institutions’ 1988 costs (cost of funds; Costs Varied servicing, operating, and other costs; and applicable taxes) as a per- centage of their outstanding loans varied from 8.48 percent for Sallie Mae to 12.34 percent for the California agency, a range of 3.86 per- centage points. While the cost of funds was the 10 lenders’ largest cost element, it varied less as a percentage of outstanding loans than ser- vicing costs or operating and other costs. Page 29 GAO,‘HRD90139BB Profitability of Guaran teed Student Loans Profitability of Stafford Student Loans Held by Secondary Markets Varied Widely The cost of funds varied from 7.06 percent of outstanding loans for Sallie Mae to 8.40 percent for the Virginia agency-a range of 1.34 per- centage points. In contrast, servicing costs varied by 2.13 percentage points, ranging from 0.80 percent for Sallie Mae to 2.93 percent for the Colorado agency. Operating and other costs varied by 3.89 percentage points, ranging from 0.15 percent for Wachovia to 4.04 percent for the California agency. Figure 14: Costs Varied Among 10 Lenders (1988) 13 Costs as Percent of Outstanding Loans . . . . ..- . . . .. ..-w [ Taxes Servicing, Operating, and Other Costs Costs of funds Funding Cost Variations Many factors influence the secondary market lenders’ cost of funds, including (1) the tax status of the securities issued, such as taxable or tax exempt; (2) the timing and terms of the debt issue-including the interest rate in effect at the time of issue, whether the rate is variable or fixed, and the length of the repayment period; and (3) the costs of Page 30 GAO/HRD9@13OBR Profitability of Guaranteed Student Loans Profitability of Stafford Student Loans Held by Secondary Markets Varied Widely issuing the debt and obtaining credit enhancements, such as letters of credit. The interplay among these factors and the volatility of interest rates make it difficult to analyze and isolate the reasons for funding cost differences. For example, while market interest rates on tax-exempt financing are generally lower than on taxable financing, a secondary market agency may incur higher average interest costs for its tax-exempt financed debt than for its taxable debt. This could occur as market interest rates declined, if it had issued long-maturity, fixed-rate, tax-exempt debt at high market interest rates, while its taxable debt was shorter-maturity and/or floating-rate. The Colorado and Pennsylvania agencies (which relied exclusively on tax-exempt financing during the period) and the Virginia agency (which used tax-exempt financing for almost two-thirds of its loan portfolio) were among those with the highest costs of funds in 1988. In addition, for four of the five state agencies that had both taxable and tax-exempt debt, the average cost of tax-exempt debt exceeded their average cost of taxable debt in at least 1 of the 4 years for which we collected data. One of the agencies had a higher average cost for its tax- exempt debt in all 3 years that it held Stafford loans financed with both taxable and tax-exempt debt. However, while the cost of tax-exempt debt exceeded that of taxable debt for some institutions in some years, overall the average cost of outstanding taxable debt exceeded the average cost of tax-exempt debt outstanding in all 4 years. (See fig. 15.) Page 31 GAO/HRD9@13OBR Profitability of Guaranteed Student L.oans Profitability of Stafford Student Lams Held by Secondary Markets Varied Widely Figure 15: lax-Exempt and Taxable Financing Costs Were Similar (Fiscal 10.0 Average Cosl of Funds (Percorn Interest Rate) Years 1985-88) 7.5 7.0 6.5 lM5 1966 1987 1666 Fiscal Year - Taxable debt - - -- Tax-exempt debt However, the 1986 Tax Reform Act (P.L. 99-514) reduced the interest rate advantage of new tax-exempt borrowing and the availability of tax- exempt funds. The act amended the Internal Revenue Code to (1) reduce personal and corporate income tax rates, thereby lessening the tax advantage of investments yielding untaxed interest income, and (2) sub- ject tax-exempt student loan bonds to an alternative minimum tax that requires certain investors to pay income tax on their interest, notwith- standing the tax-exempt status of the bonds. For new taxable and tax- exempt debt, these changes tend to narrow the difference between interest rates. The 1986 Tax Reform Act also reduced the availability of tax-exempt funds by restricting, in stages, the amount of tax-exempt debt a state could issue each year. For calendar year 1988, this volume cap limit was $50 per capita, or $150 million for each state-whichever was greater. For the six state agencies that provided data for all 4 years, the propor- tion of Stafford loan portfolios financed with tax-exempt borrowings declined from over 75 percent of outstanding loans at the end of fiscal year 1985 to less than 50 percent at the end of fiscal year 1988, although the dollar volume of tax-exempt loans rose over the period. Page 32 GAO/IIRD~13OBR Profitabi ty of Guaranteed Student Loans - Profltabllty of Stafford Student Loans Held by Secondary Markets Varied Widely (See fig. 16.) For example, an official at the Virginia agency told us it was unable to issue additional tax-exempt debt to purchase Stafford loans in 1987 and 1988 because the agency did not receive state approval for an allocation under the state’s volume cap for new tax- exempt bond issues. Figure 16: Proportion of Loans Financed With Tax-Exempt Funds Has Declined Outstanding Loans (Billions of Dollars) (Fiscal Years 1985438) 2.0 1985 1966 1997 Fiscal Year I 1 Financed with taxable debt 1 Financed with tax-exempt debt Servicing and Other Costs Nonfund costs, which include the cost of servicing and all other Stafford loan-related costs other than the cost of funds, varied somewhat among Varied the 10 institutions in fiscal year 1988 (see fig. 17). Sallie Mae, with the largest portfolio, had the lowest nonfund costs that year (1.42 percent of outstanding loans). However, there was no apparent connection between the size of the other institutions’ portfolios and their nonfund costs. Rather, differences in lenders’ servicing and operating costs reflect their operating policies and experiences. Among circumstances officials described to us to explain their nonfund costs were the following: Page 33 GAO/‘HRD9@13OBR Profitability of Gunran teed Student Loam Profitability of Stafford Student Loans Held by Secondary Markets Varied Widely l Though the California agency reported relatively low servicing and operating costs, its fiscal year 1988 total nonfund costs exceeded the other nine institutions’ costs, reflecting a 3.34-percent provision for losses on delinquent loans. According to an agency official, some defaulted loans may not be reimbursed by the guaranty agency or the Department of Education if either determines that they were not prop- erly serviced. l The Colorado agency’s high fiscal year 1988 nonfund costs (3.58 percent of outstanding loans), according to an agency official, reflected expenses related to its transition from in-house to contracted servicing. 9 The Pennsylvania agency reported high nonfund costs in all 4 years. The agency services loans for other lenders in addition to its own, and it used the proceeds from its loan-servicing operation to help subsidize loans to borrowers who do not qualify for federal subsidies under Staf- ford loans. l According to a bank official, Chase Manhattan’s relatively high nonfund costs (2.19 percent in fiscal year 1988) increased from previous years, in part due to additions to its staff and equipment in anticipation of expanding its student loan activities. Figure 17: Servicing, Operating, and Other Costs Varied-in Fiscal Year 1966 5.5 Servicing and Operating Costs (Percent) 5.0 4.5 4.0 3.5 3.0 2.5 I . Page 34 GAO/HRD90-130BR Profitability of Guaranteed Student Loans Profitability of Stafford Student Loans Held by Secondary Markets Varied Widely Taxes Substantially Unlike the seven state agencies, Sallie Mae and the two banks are sub- ject to income taxes. Sallie Mae pays federal corporate income taxes, but ReducedFor-Profit is exempt from state and local income taxes. The banks are subject to Lenders’ Returns both federal and state taxes. As shown in figure 18, the payment of taxes substantially reduced these three lenders’ net rates of return on Stafford student loans in fiscal year 1988. Figure 18: Taxes Reduced Federal and Commercial Lenders’ Profitability in Return as a Percentage of Outstanding Loans Fiscal Year 1988 1.6 . 0.6 l-J Applicable taxes Net rate of return (after taxes) IConclusions 1985-88 among the secondary market institutions that we reviewed. The variations resulted more often from variations in costs than from varia- tions in revenue. Page 36 GAO/HRD9@13OBR Profitability of Guaranteed Student Loam Prof¶tability of Stafford Student Loans Held by Secondary Markets Varied Widely In addition, the 1986 subsidy reductions had little or no effect on lenders’ revenues. For some lenders in some years, however, the reduc- tions could be significant when compared to profits because profit mar- gins were relatively narrow. Four of these institutions consistently earned a profit on their Stafford loans, including the two commercial lenders and Sallie Mae, all of which are for-profit entities. The other lenders incurred losses in 1 or more years. These lenders were not-for-profit or state agencies that entered the secondary market for reasons other than making a profit. The variations in profit levels, and the many reasons for them, indicate that profitability measures do not, in themselves, provide a sound basis for determining the appropriate special allowance factor. . The Department of Education and 9 of the 10 lending institutions we Agency Comments and reviewed commented on a draft of this report. The Department had only Our Evaluation a technical comment that we addressed in appendix II. Our evaluation of the comments received from the institutions are summarized below. 1. Several institutions suggested we more clearly emphasize that the 10 participants in the study may have used different assumptions or methods to allocate costs, and that 2 of the participants provided data on a calendar year rather than a fiscal year basis. While we requested comparable data from all institutions and identified possible inconsistent assumptions or allocations of costs, we recognize that differences among the lenders exist. We discuss data limitations in appendix I. 2. Several lenders stated that their costs have increased since the com- pletion of our review. According to these lenders, increases included higher letter-of-credit costs and higher administrative costs attributed to stricter enforcement of due diligence requirements. Lenders also stated that their revenues had been adversely affected by (1) Treasury Depart- ment regulations that reduced the benefits of using tax-exempt financing and (2) lower special allowance payments, which are having a greater impact on revenues each year. We recognize that profit levels of some institutions may have changed since our review. We state in the report that the impact of the reduction in the special allowance rate should be greater for some agencies in Page 36 GAO/EIRB9@130BR Profitability of Guaranteed Student Loana Profitability of Stafford Student Loans Held by Secondary Markets Varied Widely future years. Our analysis was limited to the 1985-88 period, and we did not attempt to forecast any future changes in lenders’ operations. How- ever, where appropriate, we have incorporated the lenders’ concerns in the report. 3. Our original draft of this report contained a consultant’s paper that discussed the legal and institutional factors affecting the secondary market in guaranteed student loans. In their comments on the report, some institutions suggested that the information from the paper was valuable, while others disagreed with some of the information the paper contained. While we believe the paper provided a useful description of the charac- teristics of the secondary market for student loans, we have deleted it from our report because of the controversy it generated among the insti- tutions and our concern that it would divert attention away from the major focus of the report. 4. The two commercial banks were concerned about public disclosure of the information they provided. We discussed the issue with officials of the two banks and agreed to (1) treat the detailed information that they provided, and which was not included in our draft report, as proprietary, and (2) identify in our report the institutions’ revenue, costs, and profitability analyses which were included in our draft report. 5. Several lenders suggested revisions and technical changes to increase the accuracy or clarity of the report. We made changes where appropriate. Page 37 GAO/HRD-!40-130BR Profitability of Guaranteed Student Loans Appendix I Methodology Early in our review, we held a conference with participants and other knowledgeable parties in the student loan community, such as repre- sentatives from the Department of Education, the Congressional Budget Office, and secondary markets, to help us develop our study approach. We also contracted with an expert on government-sponsored enterprises to identify and describe the legal and institutional benefits, limitations, and other factors that influence the efficiency, competitiveness, and profitability of the three major kinds of secondary market institutions. As agreed in discussions with congressional staff, we focused our efforts on a group of major secondary markets, that is, financial institutions that purchase Stafford loans from originating lenders, such as banks, savings and loan associations, and credit unions. Because many sec- ondary markets deal primarily in student loans, we expected that they would be more likely than originating lenders to maintain financial data that could be used to determine the profitability of their Stafford loan portfolios. We focused our analysis on the student loan holdings of 10 major sec- ondary markets during fiscal years 1985-88. These 10 accounted for about one-third of all Stafford loan holdings at the end of fiscal year 1988 and nearly three-fourths of all Stafford loan purchases lenders reported to the Department of Education for fiscal year 1988. The 10 institutions were judgmentally selected to represent the three basic kinds of entities: commercial banks, state agencies, and an institu- tion chartered by the federal government to provide a secondary market for student loans. As a basis for our sample selection, we used Depart- ment of Education data on dollar volume of Stafford loan holdings and purchases by secondary market institutions. Of the institutions selected-other than Sallie Mae, the dominant secondary market entity-two were commercial banks. To provide a cross-section of the different kinds of state agencies, we selected four not-for-profit corpora- tions and three state governmental agencies. Six of the lo-the two banks, the three state governmental agencies, and one of the state not- for-profit institutions- originate as well as purchase loans. All 10 were among the top 40 holders of guaranteed student loans in fiscal year 1988. We sent questionnaires to each of the 10 institutions requesting data for fiscal years 1985-88 regarding special allowance payments, revenues, and cost of funds and servicing, operating, and other costs not related to financing. Page 38 GAO/HRD9@13OBR Profitability of Guaranteed Student Loans Appendix I Methodology We mailed each of the 10 institutions three questionnaires: Data Collection . Special allowance payment questionnaire-requested, by year, a break- down of loan portfolio by the SAPfactor (3.5 percent, 3.25 percent, or other) used to calculate special allowance payments. . Cost of funds questionnaire-requested distribution of fiscal year-end loan balances by source of funding (taxable, tax-exempt, or other) and the cost of funds for and rate of return on student loans. Additional items on this questionnaire included letters of credit and their cost purchase price of portfolios (whether at par or at a premium or dis- count), and whether new borrowings were at fixed or floating interest rates. l Servicing and operating cost questionnaire-requested loan-servicing costs, operating costs, and other costs not related to the cost of funds; proportions of portfolio serviced by the institution or contracted out; and comments, including a description of efforts to constrain these costs. We requested cost and revenue data as a percentage of portfolio rather than in terms of dollar volume. In those cases where we determined from talking to responsible officials that they had based their cost and/ or revenue percentages on some other measure of portfolio, we asked them to recalculate using average daily loan balance. We tabulated data received in response to these questionnaires and used the data to calculate rates of profitability and to assess the relative importance of various factors to explain variations in profitability. Though we use the terms “profitability,” “profit,” and “loss” in dis- cussing net returns of all these institutions, we recognize that state agen- cies’ activities do not generate profits as such. To calculate net rates of return, or “profits,” we aggregated the cost of funds, servicing, operating, and other costs (all as percentages of loan balances) and then deducted the sum of these costs from interest rev- enue (made up of borrowers’ interest plus federal special allowance pay- ments). Where applicable-that is, for the two banks and Sallie Mae- we deducted taxes to obtain their net rate of return after taxes. We checked data validity principally by examining the internal consis- Data Validation tency of data provided; the consistency of those data within the context Page 39 GAO/HRIHK%13OBR Profitability of Guaranteed Student Loans Appendix I Methodology of relevant laws and regulations; and, to a limited extent, the consis- tency of questionnaire data with data reported to the Department of Education, such as institutions’ annual reports and financial statements, For example, we checked the volume of an institution’s loans subject to the reduced special allowance payment against outstanding loans funded with taxable loans. Because the reduction did not apply to loans from tax-exempt funds, any excess of 3.25-percent special allowance payment loans over taxable funded loans suggested an error in one of the totals. We also calculated a range of possible rates of return (interest revenue) based on formulas specified by law and compared these ranges with rates of return the institutions reported. We interviewed Department of Education officials and financial officials at the secondary market institutions to confirm our interpretations of the regulations. We reviewed reports by the Department, the Congres- sional Budget Office, and the Congressional Research Service, as well as other literature relating to student loan finance. When we had obtained corrected data or explanations of apparent inconsistencies, we sent review copies of our compiled and derivative data to financial or executive officers at each of the 10 institutions, requesting that they make any needed changes. Nine of the 10 institutions sent confirmation of the data. Some of these included additional revisions. California sent us additional financial data on which to base the requested data but asked us to perform the calculations. To do so, we allocated operating costs and the cost of funds between taxable and tax-exempt funds in the same proportion that the agency allocated outstanding debt. We conducted our work between January 1988 and February 1990 in accordance with generally accepted government auditing standards. Data Limitations certain data limitations remain. Data Validity Except where our data analysis revealed inconsistencies, we did not attempt to verify or validate the data institutions provided us. Page 40 GAO/HRIh9@13OBR Profitability of Guaran teed Student Loana Appendix I Methodology Imprecision Due to Use of Some data represent estimates rather than exact values. For example, Virginia’s agency cautioned that some of its data are estimates and that Estimates because of the use of average balances, its data should not be construed as exact. As another example, the California data are estimates based on that agency’s guaranteed student loan portfolio; the agency does not maintain separate cost data on its Stafford loan holdings. Fiscal Year Variations We requested data for fiscal years ending September 30. However, only 2 of the 10 entities operate on the federal fiscal year. Of the eight that operate on other fiscal years, all but two provided cost and revenue esti- mates based on the federal fiscal year. Of the two entities that did not provide data based on the federal fiscal year, one pointed out that because of year-end adjustments, conversion to a September 30 fiscal year would result in distorted data. As noted on affected figures, those two institutions’ data are by calendar year. They are therefore not directly comparable to the other institutions’ data, particularly when interest rates for the calendar year differ substan- tially from rates for the fiscal year. Moreover, we do not know how the institutions that converted their data for us handled year-end adjustments in completing our question- naires. One of the six that converted their data commented that the con- version probably entailed some sacrifice of precision. Trend Data To present a cross-section of the agencies represented, summary data and charts representing trends in cost of taxable and tax-exempt funds over time were developed using simple averages of the agencies’ costs. Because they are not weighted by loan volume, they do not reflect the aggregate costs of the 10 institutions’ portfolios financed with taxable funds as compared with those financed with tax-exempt funds. Further, because we included institutions’ data as available, averages do not represent the same number of institutions in each year. One agency was unable to separate guaranteed Stafford loan costs from costs of other student loan programs and was unable to provide cost of funds data for 2 of the 4 years. Another was unable to separate out guaran- teed Stafford loan revenue for any of the years. Page 41 GAO/HRD-W13OBR Profitability of Gum-an teed Student Loans Appendix I Methodology Differences in Operations Because of variations in the 10 institutions’ operations, costs do not always reflect the same cost elements. In figure 14, for example, ser- vicing costs may reflect in-house servicing, contracted servicing, or a mix of the two. Differences in Accounting In addition to differing in their operations, lenders differed in their methods of accounting for costs. For example, we asked institutions to Practices include in their cost of funds all costs incident to obtaining funds. Debt issuance costs institutions told us they had included in the cost of funds varied somewhat, and we did not attempt to eliminate those variations. Nor did we attempt to adjust institutions’ cost of funds for variations in their accounting practices with respect to some cost elements-pre- miums paid on loan purchases, for example. We recognize that, since the completion of our review, the financial con- dition of the institutions could have changed. For example, since 1988 the institutions’ borrowing costs have likely increased. Also, costs may have increased due to stricter loan servicing requirements imposed by the Department. Page 42 GAO/HRDfWl2OBR Profitability of Guaranteed Student Loans Appendix II Data Supporting Figures Table 11.1:Ten Institutions’ Loan Holdings Doubled (Fiscal Years 198588) (Data for Dollars in bullions Fig. 6) Outstanding amount of Stafford loans Seven state Fiscal year Sallie Mae Two banks institutions Total 1985 $5.144 $0.871 $1.423 $7.438 1986 6.271 0.854 1.753 8.878 1987 7.419 1.095 2.449 10.963 1988 9.357 1.068 3.054 13.479 Note: Thus table does not rnclude a major portion of Pennsylvanra’s guaranteed student loans. We excluded loans to students not eltgible for federal interest subsidles (about $8.2 mrllron, $22.1 mrllton, $71 5 mrllron. and $340.2 million at the end of fiscal years 198588, respectively). Accordtng to an agency offrcral, these loans were made to students who were ineligrble for federal interest subsrdtes because, for example, their incomes exceeded federal limrts. Nonetheless, according to this official, thetr loans are guaranteed lust as other Stafford loans by federally supported guaranty agencies. We also excluded from our analysis about $7.1 million of loans eligible for federal interest subsidies that Pennsylvania purchased wrth taxable funds the last day of fiscal year 1988 because the agency did not provide data . for these loans. Table 11.2:Nine Institutions’ Profitability Varied Widely (Fiscal Year 1988) (Data for Net rate of return in 1988 Fig. 7) Lender@ as a percent of poftfoliob Sallie Mae 0.95 ChaseC 0.18 Wachovta 0.34 CaliforniaC -3 31 Indiana 0.92 Nebraska 0 26 Nellie Mae 0 34 Pennsvlvania -0.40 Virainia -0.18 ?nsufficient data were available to calculate Colorado’s rate of return bNet rates of return were calculated after taxes, if applicable. CCalrfornra and Chase data are for calendar year 1968; other data are for the fiscal year endrng Sep- tember 30. 1988. Page 43 GAO/HRD-W13OBR Profitability of Guaranteed Student Loans Appendix ll Data Supporting Figures Table 11.3:For-Profit Institutions Were Consistently Profitable (Fiscal Years 1985 Net rate of return as a percent of portfolio 88)(Data for Fig. 8) After taxes Before taxes Yeap Sallie Maeb WachoviaC Chased Sallie Mae Wachovia Chase 1985 0 96 064 0.29 1.78 126 0.57 1986 083 0.46 0.29 1.53 0.90 0 57 1987 0.88 0.42 0.23 1.50 0.77 0.43 1988 0.95 0.34 0.18 1.47 0 56 n 79 ‘Chase data are for calendar years; other data are for fiscal years ending September 30 bUnlrke the other rnstrtutrons reviewed. Sallre Mae included In Its figures adjustments for expected increases In servicing costs as loans mature. These adjustments were 0 22. 0 18, 0 11, and 0 14 percent In fiscal years 1965-88, respectrvely These adlustments were treated as deferred Income rn Sallre Mae’s financral reports and as addrtrons to costs In the figures Sallre Mae provided to GAO. The figures pro- vrded by Sallie Mae indicate that taxes as a percentage of net rncome were about 46, 46, 41, and 35 percent In fiscal years 198586, respectrvely. Due to such items as tax-exempt rncome and tax benefits . In lease transactrons, Sallie Mae’s effective tax rates (taxes as a percentage of net Income from all sources) for all operatrons were 36.9, 350, 31 0, and 27.4 percent in calendar years 198568, respec- trvely. Unlrke the banks, Sallte Mae IS exempt from state and local taxes. CWachovra’s student loan operations were subject to state as well as federal Income tax Wachovra reported that its taxes as a percentage of net Income were about 49, 49, 45, and 40 percent tn fiscal years 1985-68, respectrvely Due to Income from tax-exempt secuntres, Investment tax credrts, etc Wachovra’s effective tax rates for all operations were lower-for example, about 22 percent in calendar year 1986. dChase was subject to federal and stnte corporate income tax The data Chase provrded rndrcate that taxes as a percentage of net income were about 49, 49, 47, and 38 percent in fiscal years 1965-88, respectrvely Due to losses from other operatrons (Income from tax-exempt Investments, etc ), Chase’s effective tax rate (total provision for taxes as a percentage of net income before taxes) was lower-for example about 20 percent in calendar year 1988. Table 11.4:Not-for-Profit Agencies’ Returns Varied (1988)(Data for Fig.9) Rate of return as a percent of portfolio Yeap Californiab Indiana Nebraska Nellie Mae 1985 c 0.46 -1.71 -043 1986 c 0.68 0.28 -057 1987 1.24 0.65 0.22 0.31 1988 -3.31 0.92 0.26 0.34 %alrfornra data are for calendar years 1967 and 1988 The other agencies provided data for fiscal years ending September 30 bCalrfornra’s agency did not provide sufficient data to calculate net rates of return in fiscal years 1985 and 1986 ‘Not available Page 44 GAO/HRD90-130BR Profitability of Guaranteed Student Loans Appendix II Data Supporting Figures Table 11.5:State Governmental Agencies’ Returns Varied, but Each Has Had Rate of return as a percent of portfolio Losses (Fiscal Years 1985-88) Fiscal year Pennsylvania Virginia (Data for Fig. 10) 1985 -2 04 0 70 1986 -371 0 11 1987 -1 71 -0 24 1988 -0.40 -0 18 Note Colorado’s agency provided insufficient data to calculate profrts. Table 11.6:Institutions’ 1966 Gross Revenues Were Similar (Data for Fig. 11) Lender Revenue as a Dercent of oottfolioa Sallie Mae 9.95 Chaseb 1038 Wachowa 9 66 Californiab 9 03 . Indiana 9 60 Nebraska 9 83 Nellie Mae 9 ss Pennsylvania 9 47 Virgmia 9 79 %hase and Cakfornia data are for calendar year 1988; other data are for the fiscal year endlng Sep- tember 30, 1988 bChase’s revenue was highest, at least In part, according to a bank official, because its data were for calendar year, not fiscal year, 1988 and Interest rates were higher in the fourth quarter of calendar year 1988 (the quarter following the end of fiscal year 1988). California also reported revenue for the calendar year, but its revenue was nevertheless the lowest, at least In part, according to the agency’s treasurer, because it did not receive interest subsidies for many of its loans as a result of servrcrng problems Except for Chase, Sallie Mae had the highest revenue (9.95 percent) and California had the lowest (9 03) Thus, revenue varied within a 0 92.percentage-point range Page 45 GAO/HRDSO-13OBR Profitability of Guaranteed Student Loans Appendix II Data Supporting Figures Table 11.7:OS-Percent Interest Revenue Floor Increased Returns on Tax-Exempt Interest rate Financed Loans* (Data for Fig. 1.2) Interest calculated Interest paid with without g&percent Fiscal year/quarter 9Spercent floor floor 1985 1 10.36 1036 2 9 98 9.98 3 9.64 964 4 9.50 942 - 1988 1 9.50 944 2 9.50 9 21 3 9.50 8.69 4 9- 50 -_ 838 1987 1 9 50 8.51 2 9.50 8.60 3 9.50 871 4 9.50 8.89 1988 1 9.50 8.84 2 9.50 872 3 9.50 0.97 A 9 !=a 937 aTotal interest lenders received from borrowers and the Department of Education on loans to first-time borrowers (all &percent loans) financed with tax-exempt funds. Page 46 GAO/HRD9&13oBR Profitability of Guaranteed Student Loam Appendix II Dab SUPPOW Fieures Table 11.8:Loans Subject to Reduced Subsidies Are Increasing (Ftscal Years Percent of portfolio subject to 3.25-percent special allowance promion 1987-88) (Data for Fig. 13) End of fiscal years Lender 1987 -______ 1988 Sallie Mae 3 -.____~ 20 Chase 33 46 Wachovla 5 28 Californiaa 25 -____ 25 Indiana” 3 6 Nebraskaa 1 39 Nellie Maea 0 11 Coloradoa,b 0 0 Pennsylvanlaa,c 0 ___- 0 VirginIaa 3 9 %tate agency. bColorado had only tax-exempt financed loans, which were not subject to the subsidy reductton ‘Except for taxable financed loans purchased on the last day of fiscal year 1988 that were not Included In any of Pennsylvanra’s data, all of the agency’s loans were financed from tax-exempt sources and thus were not subject to the reduction. Table 11.9:Costs Varied Among 10 Lenders (Fiscal Year 1988) (Data for Fig. 14) Costs as a percent of portfolio in 1988 Operating cost of Servicing and other Lender Lender funds costs costs taxes* Total Sallie Mae 7.06 0.80 0.62 0.52 9.00 Chaseb 7.96 1.29 0.90 0.11 10.20 Wachovia 7.30 1.65 0.15 0.22 9.32 Californiab 7.21 1.09 4.04 0.00 12.34 Indiana 7.16 1.03 0.49 0.00 8.88 Nebraska 7.51 1.50 0.56 0.00 9.57 Nellie Mae 7.31 1.12 0.92 0.00 9.35 Colorado 7.93 2.93 0.65 0 00 11.51 Pennsylvania 7.52 1.68 0.67 0.00 9.87 Viraima 8.40 1.05 0.52 0.00 9.97 aApplicable only to Sallie Mae and the two banks. bChase and California data are for calendar year 1988: other data are for the fiscal year ending Sep- tember 30, 1988. Page 47 GAO/IiRDMI-13OBR Profitability of Guaranteed Student Loans Appendix II Data Supporting Figures Table 11.10:Tax-Exempt and Taxable Borrowing Costs Were Similar (Fiscal Average borrowrng costs (figures are in percent) Years 1985-88) (Data for Fig. 15) Year Taxable Tax exempt 1985 9 22 9 14 1986 __~.- 7.77 7 62 1987 7 29 7 08 1988 7.86 7 09 Note Data shown are unwerghted averages for state agencies that reported the cost of both taxable and tax-exempt debt at some trme dunng the fiscal year 198588 period. The averages represent drf- ferent numbers of agencies In drfferent fiscal years. three In fiscal year 1985, four in fiscal year 1986: five In 1987 and 1988. Virglnia was not included In the fiscal year 1985 averages because it had no taxable financed loans In that year Californra was not included In the fiscal year 1985 and 1986 averages because It drd not provrde data on the costs of its tax-exempt and taxable debt In those years. Colorado and Pennsylvania were not Included in any of the averages because they reported no taxable financed debt during the fiscal year 198588 period. Chase and California data are for calendar years, other data are for the fiscal year endrng September 30, 1988 . Table II.1 1: Proportion of Loans Purchased With Tax-Exempt Funds Has Dollars in millions Declined (Fiscal Years 1985-88) (Data for Loan holdings Fig. 16) Loan holdings from taxable from tax- Percent debt and other Percent Fiscal vear exemot debt of total sources of total Total loans* 1985 $843 80 $216 20 $1,059 1986 736 60 482 40 1,217b 1987 815 53 710 47 1 ,525b 1988 967 49 1,008 51 1.975 Note: Data shown are totals for SIX state agencres that provided data for all 4 years %um of the columns does not equal the total due to rounding. bData for the end of fiscal years 1986 and 1987 Include about $4 mrllron of loans In Indiana’s portfolio financed from nerther tax-exempt nor taxable debt. Table 11.12:Servicing and Operating Costs Varied (Fiscal Year 1988) (Data for Lender Cost as a percent of portfolio* Fig. 17) 1 42 Sallie Mae Chase 2 19 Wachovia 1 80 California 5.13 Indiana 1 52 Nebraska 2.06 Nellie Mae 2.04 Colorado 3.58 Pennsylvania 2 35 Virninia 1 57 %hase and Calrfornia data are for calendar year 1988; other data are for the fiscal year ending Sep- tember 30, 1988. Page 48 GAO/I3RDSO-13OBB Profitability of Guaranteed Student Loans Appendix II Data Supporting Figmes Table 11.13:Taxes Reduced Federal and Commercial Lenders’ Profitability (Fiscal Profits as a percent of portfolios Year 1988) (Data for Fig. 18) Applicable Net rate of return after Lender Rate of return before taxes taxes taxes Sallie Mae 1 47 0.52 0.95 Chase 0.29 0.1 1 0.18 Wachovla 0.56 0.22 0.34 Thase data are for calendar year 1988; other data are for the fiscal year ending September 30, 1988. Page 49 GAO/HRD9@19OBB Profitability of Guaran teed Student Loans Amendix III CommentsFrom the Department of Education UNITEDSTATESDEPARTMENTOFEDUCATION OFFICEOF THE ASSKTANT SECRETARY FOR POSTSECONDARY EDUCA rlON Mr. Franklin Frazier Director, Education and Employment Issues United States General Accounting Office Human Resources Division Washington, DC 20540 Dear Mr. Frazier: Thank you for the opportunity to review GAO draft report, "Guaranteed Student Loans: Secondary Market Lenders' Profits Vary Widely" GAO/HRD 90-130, dated July 19, 1990. The Department offers the following technical comments to be taken into consideration when preparing the final report. Now on pp. 47 and 27. 12. u Calculations for fiscal year 1987, quarters 1, 2, and 3 for interest calculated without 9.5 percent floor in effect are correct if you consider each quarter alone. However, it does not represent the true effect depicted in the report. The illustration has totally ignored the fact that most new loans made during the sequester were made during the last 3 months (July, August, and September). If you have any questions, please contact Valerie Hurry of the Division of Quality Assurance on 708-9453. Sincerely, Leonard L. Haynes 111( Page50 GAO/HRIMO-13OBRProf'itabilityofGuaranteedStudentLoans Appendix IV CommentsFrom Sallie Mae STUDENT LOAN MARKETING ASSOClATlON 1050 Thomas Jefferson Street NW Washington 0 c 20007 202-298-2600 September 4, 1990 HAND DELIVERED Mr. Franklin Frazier Director, Education and Employment Issues United States General Accounting Office 441 G Street, N.W., Room 6739 Washington, DC 20548 Dear Mr. Frazier: Thank you for the opportunity to comment on the General Accounting Office's draft report "Guaranteed Student Loans: Secondary Market Lenders' Profits Vary Widely". We endorse the Report's conclusion that the variation in profitability among the secondary markets examined is largely attributable to their respective abilities to efficiently manage the cost of servicing student loans and to effectively contain their general operating costs. In our view, the report reaches an important conclusion in its findings that the least significant variation in costs among the secondary markets studied is their cost of funds, which as is pointed out in the Report is by far the largest cost element for all secondary markets. According to the Now on p. 29. Report (Page 33), the variation as to costs of funds is "less as a percent of outstanding loans than all other cost elements (servicing, operating and other non-fund costs) combined". Material deleted, see p. 37 Page 51 GAO/HRD-90-13OBR Profitability of Guaranteed Student Loans Appendix N Comments From Sallie Mae Mr. Franklin Frazier Page Two September 4, 1990 While the Report attempts to assess the impact on lender profitability of Congressionally mandated reductions in the special allowance formula, it does not go far enough in addressing the long term impact of such reductions. Specifically, we think GAO should have attempted to isolate the effect of the 1986 Gramm-Rudman- Hollings sequestration order (which reduced the special allowance . rate to T-Bill plus 3.0) on the loans actually affected by that order. This could have been accomplished by: 1) isolating that segment of each lender's loan holdings that was affected by the sequester order: 2) applying the proportional costs associated with each loan holder's portfolio to that segment: and 3) by reducing each loan holder's income on that segment of loans by the relative amount of the decrease in special allowance payments attributable to the Gramm-Rudman-Hollings sequester order. Similarly, GAO could have examined the future effects on profitability of those loans originated since November 16, 1986, the date the special allowance rate was reduced by 25 basis points under the Higher Education Amendments of 1986. Since loans subject to the reduction will eventually dominate a lender's holdings, such an analysis would have helped the Congress to better understand the full effect of long-term reductions in special allowances. Even now, because over $40 billion of GSLP loans have been made since the enactment of the 1986 Amendments, the effect on lenders' portfolios and their overall profitability is much more dramatic than would be true from an analysis which failed to look beyond fiscal 1988. This short sighted approach has seriously reduced the value of the Report. As Sallie Mae has previously stated in correspondence and conversation with your office, we believe that GAO must acknowledge that secondary markets are not a wholly suitable proxy for the universe of guaranteed student loan lenders. While the number of secondary markets serving the GSL program have remained relatively constant over the past several years, there has been a steady decline in the number of lenders originating loans under the GSLP. This indicates that the overall health of the program and the effect of program revisions, such as reductions in the level of special allowance payments received by program participants, may not be adequately evaluated by analyses that concentrate exclusively on the secondary markets. We strongly urge that some Page 52 GAO/HRD9@13OBR Profitability of Guaranteed Student Loans Appendix N Comments From Sallie Mae Mr. Franklin Frazier Page Three September 4. 1990 mention of the limitations associated with the study's concentration on secondary markets be added to the Report introduction. Lastly, we think the Report does not give adequate weight to the significant uncertainties regarding integrity of the data being reported. We suggest that the items listed in Appendix II under the heading "Data Limitations" be summarized and brought forward as part of the introduction to the Report and the accompanying summary. These limitations which include concerns regarding data validity, the use of inconsistent fiscal years, limited trend data and variations in the accounting practices of those organizations providing data to GAO, are significant enough that they should be brought directly to the reader's attention. l Material deleted, see p. 37 Sincerely, u Lawrence A. Hough President and Chief Executive Officer Enclosure cc: Mr. Jay Eglin Page 53 GAO/HRWlO-130BR Profitability of Guaranteed Student Loans Appendix V - comments From ChaseManhattan QlcHAsE = August 20, 1990 Mr. Franklin Frazier Director, Education and Employment Issues United States General Accounting Office 441 G. Street Northwest Washington, D.C. 20540 . Lear hr. Frazier: Re: Proposed Report to Congress on Student Loan Profitability This letter is in response to your July 19, 1990 letter to Charles Christiana. While Chase Education Finance supports the aforementioned study, we note that the limited number of commercial bank participants weakens any conclusions that may be drawn from the study. In Chase Education Finance's data collection package, the data were predicated on certain assumptions and estimates. Since there were no standard assumptions utilized by the participants it is probable that the variances cited with respect to gross income yields, funding and operating costs may, in part, be due to different assumptions and allocation methodologies within each institution. As a result, comparability of the results might be questionable. Chase Education Finance has an even more basic concern. We oppose the dissemination of the study in its present form and object to any release of our confidential or vroorietarv information. Chase Education Finance's intention in completing the data collection forms was to provide data for consolidation with other institutions. Moreover, identification of Chase is not, in our opinion, a critical element of the study and therefore anonymity should be afforded to us. In addition, since this study may be subject to release under the Freedom of Information Act, Chase Education Finance respectfully requests that all commercial and financial data of Chase and its participation in the study not be disclosed. It is our understanding that information which contains proprietary and confidential data is not subject to public disclosure under the Freedom of Information Act or under the GAO's regulatory policies. It is our opinion that the data provided is proprietary and confidential and should not be made available publicly. Page 54 GAO/HRl%SQ130BB Profitability of Guaranteed Student Loans Appendix V Comments Prom Chase Manhattan Mr. Franklin Frazier Page two August 17, 1990 Chase Education Finance appreciates the opportunity to review the draft and communicate its position. In view of Chase's objections, I assume you will consolidate the financial data and not release Chase's data separately. If that is not the case, I am available to discuss these issues with you in greater detail. Please feel free to call me at (813) 881-8080. Sincerely 1 i - ', . \ Stephen T. Iovino . President STI/sah Page 65 GAO/HRD43@13OBR Profitabiity of Guaran teed Student Loans Appendix VI - CommentsFrom Wachovia Wachowa Wachcwia Bank & TrustCompany.N.A. PO.Box xl!39 WmstanSalem. NC271503099 August L5, 1990 Mr. Franklin Frazier Director, Education and Employment Issues United States General Accounting Office 441 G Street, N. W. Washington, DC 20548 Re: Proposed Report to Congress on Student Loan Profitability Dear Mr. Frazier: This letter is in response to your July 19, 1990 letter to Kay Triplett. We appreciate the opportunity to comment on the draft report. We support the study of lender profitability by the General Accounting Office. Although your survey was limited as far as cormnercial banks are concerned, your conclusions show the declining profitability of student loan assets to an after-tax margin which is not very attractive. We cooperated in completing the "data collection instruments" citing the fact that our data input was based on certain estimates and assumptions. Some of the volatility in costs cited in your study is likely due to variances among respondents in estimating yields, funding costs and other cost allocations. We question any conclusion one might draw regarding absolute profit levels with such a small sample size. Assuming respondents used the same assumptions and estimates for each year's data, one could draw some conclusion regarding trend absent a conclusion about absolute levels. We assume that your report may be subject to release under the Freedom of Information Act. We respectfully request that Wachovia data and Wachovia's cooperation in your study be granted anonymity. We understand that commercial and financial information which contains privileged and confidential information is subject to exception from release under the Freedom of Information Act. We deem estimated yields, internal estimates of funding costs, and servicing and other cost estimates to be valuable proprietary information. This information is not available from other sources. Product profitability estimates may be a valuable resource internally, but should not be available for external publication. Page 8 of the draft report illustrates a plan of wide distribution absent any request under the Freedom of Information Act. Page 56 GAO/HRD-SO-13OBR Profitability of Guaranteed Student Loans Appendix VI Comments From Wachovia Mr. Franklin Frazier Washington, DC 2 August 15, 1990 we must protest the publication of the draft report in its current form and request that GAO carefully control copies of the draft report. Wachovia's intention in completing the data collection forms was to provide data which would be aggregated with other respondents. We do not believe that release of the identity of Wachovia Bank b Trust is necessary for the purposes of your study. Thank you for allowing us to state our position. I would be happy to discuss our position with you and can be reached at 919-770-4554. Sincerely, Richard B. Roberts Executive Vice President Page 67 GAO/HRB!3@13OBR Profitability of Guaranteed Student Loans Appendix VII CommentsFrom the CaliformiaAgency CALIFORNIA STLJDENT L~A~FINANCE CORPORATION August 17, 1990 Mr. Franklin Frazier Director, Education and Employment Issues United States General Accounting Office Human Resources Division Washington, D.C. 20548 . Dear Mr. Frazier: Thank you for including California Student Loan Finance Corporation (CSLFC) in the General Accounting Office's study regarding the profitability of student loans to secondary market lenders. We at CSLFC view our participation in the study as an opportunity to assist the General Accounting Office in educating Congress relative to the numerous influencing factors relating to our secondary market's profitability over the last several years. We have reviewed the draft report you sent to us in its entirety. Clearly, it is extremely thorough and very informative. However, we could not find where inherent risk in the guaranteed student loan program is explicitly discussed. An example of this risk is where legislated change to the program retroactively changed servicing requirements for loans which were originated or purchased in previous years. When this occurred, it created an immediate profitability risk, a risk which was not a reality nor perceived to exist when the affected loans were originated or subsequently purchased. This type of retroactive change significantly altered profitability levels for lenders and holders of student loan portfolios. Please let me know if you or your staff have any questions or comments. Again, thank you for allowing us to participate in this vitally important study. Sincerely!., .4 SK/th Page 58 GAO/HRJMO-18OBR Profitability of Guaranteed Student Loans Appendix VIII CommentsFrom the Indiana Agency Aeguct 24, 1990 Mr. Joaaph J. Eglln, Jr. 251 Nwr(h MMis streat Aaaistant Director Suita 1000 Human Resources Division Indianapolis. IN 46204 General Accounting Office 317-237-2000 Washington, D.C. 20548 Dear Mr. Eglin: Thank you for the draft report entitled “Guaranteed Student Loans Secondary Market Lenders Profits Vary Widely.” I have reviewed the report in detail and I believe it fairly rrpresents information provided to the General Accounting Office by our organization. I cannot coeaaent regarding the accuracy of the information in regard to other organizations. It is important to note, as a matter of update, that costs of operatiowa have increased substantially since those periods covered by the report and now represent 1.74X of outstanding student loan assets. This increase in cost of operations is largely attributable to the imposition of very prescriptive due diligence requirements of questionable value in the collection of loans. We continue to believe that greater efficiencies can be realized in costs of operations while enhancing collection SfKPbKN 1. CLINTON effectiveness if the level of regulatory direction is tied to PfWdWU delinquency and defsult rates. Through this approach, those organiaations who are ineffective in their collections would recciva increased regulatory oversight and those organizations which kavc proven themselves capable in collection of education loaad wwuld be permitted to retain that effectiveness unftc$ered by prescriptive due diligence requirements. We appreciate the opportunity to be involved in the study. ?leese call me if you have any questions about my comments. Indiana Secondary Market lot EdocahonLoans. Ino QAo~fB13BBit ProRtabiUty of Gaurmteed Student Loans Appendix IX CommentsFrom the NebraskaAgency !YEBHELP Nebraska llighcr Education Loan Program, lnc IiOil “0” Slreef PO Buu lll505 Lmcoln, SE 68501-2505 August 22, 1990 Mr. Franklin Frazier Director, Education and Employment Issues United State General Accounting Office Washington, DC 20548 Dear Mr. Frazier: . Thank yell for the opp,xtunity to comment on the draft of the GAO’s repori on t!:c profitability of guaranteed student loans held by secondary markets. As we understand them, the objectives of the report as directed by the House Committee on Education and Labor and the Senate Committee on Labor and Human Resources were to determine: l the profitability of student loans held by major secondary markets, l the reason for the variations in profitabihty, and . the effect of the 1986 reduction in the interest subsidy rate on profitability. It would be difficult for people not directly involved in the student loan industry to comprehend the difficult nature of this undertaking, and we applaud your efforts. Thr Nebraska Higher Education Loan Program, Inc. (NEBHELP) has several concerns about the report, however, which we will address in this letter. Our concerns include the scope of the report, major changes that have occurred since the period covered in the report that make the information in the report obsolete, and the impact of the Student Loan Marketing Association’s (Sallie Mae) inclusion in this report. Scope The scope of the report and the large number of variations in the agencies and d:lta studied preclude making any general conclusions related to the objectives of the report. Now on p. 36. To illustrate, in the first paragraph on page 41 of the conclusion you state. “the I’JSh subsidy reduction had little or no effect on lenders’ revenues.” The discussion on pasts Now on pp. 28-29. 30 - 32 and the data in Table III.8 in Appendix III suggest, however, that the subsidy reduction may not have efiected lenders’ revenues because secondary markets did not have significant loan volume in their portfolios subject to the reduced subsidies. A more accurate conclusion based on the information you provide would be. “The effect uf the 1986 subsidy reductions cannot be determined at this time since the subsidy reduction has yet to be passed from originating lenders to secondary markets.” We agree with your Now on p. 36. conclusion in the final paragraph of the conclusion on page 41: “The variations in profit levels. and the many reasons for them indicate that profitability measures do not. in themselves, provide a sound basis for determining the appropriate special allowance factor.” Dated Information The data used to generate the analysis and draw conclusions in this report was collected Page 60 GA0/HRD60430BR Profitability of Guaranteed Student Loans Appendix IX Comments From the Nebraska Agency Page 2, Franklin Frazier, August 22, 1990 from fiscal years 1985 through 1988. A number of significant events and changes have occurred in the student loan industry since 1988 and increased the costs associated with acquiring, owning, and servicing loans. These events and changes include the UES failure, changes in regulations, and most recently, the financial difficulty of the Higher Educational Assistance Foundation (HEAF). the nation’s largest student loan guarantor. UES failure The UES incident has created a dramatically different cost of funds structure. Due to both real and perceived risks, credit providers, particularly the Japanese banks, have made a wholesale exit from the student loan industry since 1988. As funds become less available, they become more costly. The fact that letter of credit fees have increased 30% - 40% since July, 1988 is proof of that statement. The resulting increased cost of obtaining credit facility has narrowed the aiready slim margins of many secondary markets and increased the need for maintaining the existing special allowance rate. Regulation changes Arbitrage regulations issued by the Treasury Department since fiscal year 1988 remove many of the benefits of utilizing tax exempt financing as vehicle for financing student loans. As discussed in the report, many state agencies and not-for-profit secondary markets have utilized tax exempt financing as the major source of financing student loan purchases. Typically, state agencies and not-for-profit secondary markets have accepted lower rates of return to fulfill the mandate of providing access and service to areas that for-profit lenders do not serve. The arbitrage earnings have allowed state agencies and not-for-profit secondary markets to subsidize otherwise unprofitable student loan operations and provide additional services and access to students. As the full extent of arbitrage restrictions is realized the possibility exists that not-for-profit and state agencies wiU have to curtail services to borrowers. Increased due diligence regulations implemented by the Department of Education in 1988 have increased the cost of servicing and operations and, directly influenced the secondary market profitability. In light of increased servicing and operation costs, it is inconceivable that further cuts can be made in the special allowance or any other facet of the program which reduces secondary market profitability. HEAF situation HEAF’s apparent collapse has created substantial doubt abcut the stability of the student loan industry. Statements by the Department of Education implying that the federal government’s guarantee applies only to the guarantee agency and not the lender has caused anxiety among originating lenders, secondary markets, and letter of credit providers. To date, several letter of credit providers have expressed strong concern regarding HEAF-guaranteed loans and others have requested that subsequent purchases not include HEAF paper. As the uncertainty persists, the possibility exists that student credit providers may cease any and aU involvement with student loan financing thus creating a serious access problem for students. These three areas of change have created an operating environment quite different from that of 1985 - 1988 when your study took place. While your report provides an excellent historical perspective on the profitability of secondary markets, it should not be Page 61 GAO/HRlM@13OBR Profitability of Guaran teed Student Loana Appendix IX Comments From the Nebraska Agency Page 3, Franklin Frazier, August 22, 1990 used to predict the future or set policies governing secondary markets. Sallie Mae The inclusion of Sallie Mae as just another secondary market skews the report and its conclusions. The federal agency status that Sallie Mae alone enjoys and the economies of scale created by their sizable portfolio and lending powers place Sallie Mae in a totally different competitive arena. Sallie Mae’s many advantages and few limitations make realistic comparisons to state or bank secondary markets impossible. The required parallels do not exist. As perceived today, the student loan industry presents greater risk than ever to credit providers. Increased risk means increased cost of funds. Since the federal government has, through arbitrage regulations, placed a cap on return to the secondary markets, special allowance provides a way to offset those increased costs. If secondary market income is cut by decreasing special allowance payments, secondary market liquidity drops, and access is reduced. If issued as drafted, your report has severe implications for the entire student credit industry and could result in restricted access to higher education. The conflict between the mandate of the student loan programs which is access, and the standards by which we, and other providers of those programs are increasingly judged (including profitability), is escalated by your report. Once again, I appreciate the opportunity to comment on this draft of your report and your attention to our concerns. If you have any questions, please contact me. Sincerely, Don R. BOW President Page 62 GAO/HRD9@12OBB prontability of Guaranteed Student Loans Appendix X CommentsFrom Nellie MAR / i x’ i,.. IX, / J/r A?- -. ,I’ i,” .. __--- --..-..- ~~ - --. _ -.. .- ’ The New England Education Loan Marketing Corptation :------- ----- ~ ----.-- l.,_ .. ” August 7, 1990 Mr. Franklin Frazier Director, Education and Employment ISSUM U.S. General Accounting Office Washington, DC 20548 Dear Mr. Frazier: Thank you for the opportunity to comment on the draft report of GAO regarding the profitability of guaranteed student loans to lenders and holders. I believe that the GAO staff has done an effective job of compiling and analyzing data provided by study participants who themselves are quite diverse in structure, financing and servicing characteristics, and portfolio composition. The draft study clearly demonstrates how political and economic factors effect program participants in different and often dramatic ways. It is important for Congress to know that these factors are delicately balanced and, when out of balance, result in program participants, including some of the largest in the nation, suffering diminished financial returns and even losses. It is telling that four of the five non-profit secondary markets realized lOSST?S in some years, snd that both commercial banks, while being Drofitable, achieved earnings well below those of other bank products. The only consistently profitable entity wss Sallie Mae. buttressed by the advantages of low cost "agency" borrowing and lower cost centralized servicing. It is also interesting to review how fast the statutory and regulatory environment (both Department of Education and Treasury) have changed. Simply over the period covered by the study we've seen: Reduction in SAP yield to T-Bill + 3.25% Gramm-Rudman-Hollings sequestration reduction to T-Bill + 3.1% Creation, expansion and reduction of SLS program Consolidation loan program Department of Education strict due diligence and cure regulations Private Activity Bond caps Change in Plan for Doing Business approvals 50 Bramrrcct-lilt Pnrk, Suite 300. Braintrce, Maswhuxrtr I)ZIX1-1761 617-H49-I32> WbEDU-LOAN Page 63 GAO/IilUMI%13OBlZ Profltabillty of Guaran teed Student Loans Appendix X Comments From Nellie Mae Mr. Franklin Frazier August 7. 1990 Page two Further, the Treasury regulations proposed in July 1989 and effective January, 1990 including SAP within the arbitrage calculation and limiting permissible operating expanses to 2%, effectively eliminates the use of tax-exempt financing for federal student loan programs. Material deleted, see p. 37 Two minor corrections: Now on p. 2. Page 3: the "New England Loan Marketing Association” should be “The New England Education Loan Marketing Corporation." Now on pp.23-24. Page 25: In 1985 and 1986 the US Department of Education was refusing to iSSIX approvals of many "plans for doing business" submitted by non-profit secondary markets. Such approval was necessary in order to receive special allovance payments, not interest benefits, when using tax-exempt funds. To continue our secondary market support of lenders, Nellie Mae did not receive SAP on loans funded with tax-exempt bonds until the Higher Education Act was amended to transfer responsibility for plan for doing business approval from the Secretary to the Governor of the State. Again. thank you for the opportunity to comment. I hope that you will take my comments here and those submitted earlier into consideration before releasing the final report. Very uly yours bW4AdT lrlc- p' Lawrence W. O'Toole President LWO/dms Attachment The New England Education Loan Marketing Corporation .._-__-~- -.~- -. I i Page 64 GAO/HRD9&13OBR Profitability of Guaran teed Student Loans Appendix XI CommentsFrom the ColoradoAgency August 17, 1990 Mr. Franklin Frazier Director Education and Employment Issues Human Resources Division United States General Accounting Office Washington, D.C. 20548 RE: Draft GAO Study Regarding Secondary Market Profitability Dear Mr. Frazier; Enclosed is our response to the draft of your organization's proposed report to Congress regarding the profitability of guaranteed student loans to secondary market lenders. During our review of the report draft, we did make several observations concerning the report’s findings and conclusions which we would now like to submit to your office for additional consideration before the final report is issued. Now on p. 3. Page 5 of the cover letter to Senator Kennedy states unequivocally that "The 1986 subsidy reductions had little, if any, effect on lenders' revenues .*I While this may be true for the period under review, we were not able to find a meaningful reference to what percentage of the study's portfolios was subject to this reduction. It would appear that, as the secondary markets continue to provide lender liquidity, and the loans within the portfolio continue to have declining balances through normal borrower repayment, the percentage of loans within the portfolio which is subject to lower subsidy will play an ever increasing part in the calculation of gross revenues as a percent of outstanding loans. Therefore, the statement quoted above should be modified to reflect its narrow application. The report's conclusion that Variations in profitability among (secondary markets) indicate that revenue and cost information does not provide a sufficient basis for determining appropriate subsidy levels" and that a number of the agencies you investigated showed Page 66 GAO/HRD9@13OBR Profitability of Guaranteed Student Loans Appendix= Comment.9 From the Colorado Agency Mr. Franklin Frazier August 17, 1990 Page Two losses from time-to-time suggests that there may have been insufficient scrutiny by Congress when reducing subsidy levels. These conclusions indicate that much more detailed research must be accomplished before subsidies are changed. Bond funded secondary markets earn income from student loan interest, special allowance payments, in-school interest if the loan is purchased prior to graduation, and investment income. As opposed to the free market, secondary markets cannot adjust interest rates to meet changing market conditions. They are confined to a legislatively-mandated rate structure: normal market oompotitive prioinq struotures do not exist in this industry. Hence, secondary markets are restricted in the earning potential on a student loan. Profitability of a secondary market hinges largely on costs. As Now on p. 3. pointed out on pages 4 L 5 of the draft, "profit variations were due primarily to differences in the landers' financing, servicing, operating, and other costs." Financing costs depend greatly on market conditions and timing of the issue. State secondary markets exist under a restrictive state volume cap, vhich can affect timing of a bond issue or portfolio purchase. If timing is off, financing costs can spiral or portfolios cannot be purchased. These restrictions do not apply to Sallie Mae or banks. We ask that the reference to financing costs being related to outstanding portfolio balance be corrected to reflect the Now on p. 29 relationship to outstanding DEBT (Page 33). The ability to be cost effective in issuing debt is hindered by state volume cap restrictions. Colorado would prefer to offer fewer, larger bond issues and access the financial markets with the obvious economies of scale, however, current volume caps on tax exempt issues make this impossible. Servicing costs have recently been escalating because of federal due diligence requirements. The study used data prior to the impact of the new due diligence regulations rendering the finding somewhat out of date already. The Office of Education has found technical violations of due diligence in almost every secondary market and servicer, the cost implications of which are unknown at this time. Also, those secondary markets using third-party servicing cannot directly control these servicing costs. Page 66 GAO/HBD9&13OBR Profitability of Gum-an teed Student Loans Appendix Xl Comment8 Prom the Colorado Agency Mr. Franklin Frazier August 17, 1990 Page Three This essentially leaves operating cost5 as the major control factor in "costs". For most organizations, this cost is a very small proportion of overall costs, much smaller than financing or servicing costs. Thus, state secondary markets are faced with a situation where cost control, to a large degree, is not directly under their influence. Now on p. 5. The report's heading statement on page 7 ("Loans financed with tax exempt funds can be more profitable than others".) is very misleading. The study defines profitability as gros5 revenues less Now on pp, 18 and 5 costs (page 17). What is being said on page 7 is that tax exempt financed loans may, UNDERCERTAIN NARKBT CONDITIONB, earn a higher special allowance (revenue) than loans financed by other means. Profitability includes costs; the report'5 statement doe5 not and is a major disservice to state secondary markets which use tax exempt financing. If this referenced statement is to remain part of the report, it should read as follows: tax -funds cerv rate more in rev-a I, If the term uprofitabilityn is used, then costs must be included. As Colorado is a non-profit, state secondary market, we find of particular interest the report's statement that for-profit Now on p. 22. secondary markets were consistently profitable (page 23), while also stating that those agencies which use tax exempt financing included some of the least profitable of the agencies studied. State secondary markets are under far more restrictions in terms of the markets they must serve. Enacting legislation requires we provide liquidity to all lenders for all eligible loan5 (guaranty still in effect, certain geographic requirements of either the borrower or the school etc.). The result is we frequently purchase and service the highest risk loans, without any off-setting compensation derived from increased subsidy (normal credit environments provide an increased rate of return for increased risk). Material deleted, see p. 37 Page 07 GA0/EJRD90430BR Profitability of Guaranteed Student Loana Appendix Xl Comments Prom the Colorado Agency Mr. Franklin Frazier August 17, 1990 Page Four We appreciate the opportunity to comment on the GAO report draft and sincerely trust our concerns will be seriously addressed prior to the final report being issued. Please feel free to contact me or my staff should you have any questions regarding our comments. Sincerely, William A. Stolf& President . Page 68 GAO/HBD90-13OBB Profitability of Guaranteed Student Loans Appendix XII CommentsFrom the Pennsylvania Agency PENNSYLVANIA HIGHER EDUCATION ASSISTANCE AGENCY 680 806 STREET HARRISBURG. PENNSYLVANIA 1,102-,346 August 31, 1990 Mr. Franklin Frazier Director, Education and Employment Issues U.S. General Accounting Office Washington, DC 20548 Dear Mr. Frazier: . This is in response to your letter of July 19, 1990 concerning PHEAA's comments on the draft report of the GAO regarding the profitability of guaranteed student loans to lenders and holders. A review of the draft report clearly indicates that GAO's staff has done a very good lob of compiling and analyzing data provided by the ten participants in the study which demonstrate quite a diverse approach to providing capital for secondary market purposes. Although PHEAA does hold approximately $40 million in Stafford loans purchased from various lenders', the statutory and public purpose is served by making loans for postsecondary education purposes to Pennsylvania residents at or below market rate levels to provide middle income families with a moderate cost source of credit to fund the costs of postsecondary education. To accomplish this goal, PHEAA must: a. Finance at tax-exempt rates. b. Subsidize the tax-exempt financings via an issuer contribution valued at five to ten percent of the face amount of the financing. C. Administer the direct loan program, including loan origination and servicing within the limitations of co?t recovery mechanisms controlled by the allowable spread inherent in tax-exempt financing. d. As Stafford loan eligibility continues to become less of a reality for the middle and upper income family, the need for PHEAA to meet this increasing demand for direct loans and the PBEAA "secondary market" activity is of the utmost importance and our program is not driven by concerns of profitability or competition. Because of the unique role of PHEAA, staff believes the Agency should be excluded from this secondary market report or placed in a separate category for the purposes of the report. Page 69 GAO/HRD99-13OBB Profitability of Guaran teed Student Loans -2- Also, it is important that the final report makes it clear to Congress that political and economic factors directiy affect the administration of each of the program participants and th,ese factors need to be considered before legislative changes are made. This is clearly evident when you look back at the numerous changes on both the statutory and regulatory level that have taken place which greatly impact on profitability of student loans to not only secondary markets but aLso direct lenders and guaranty agencies. Thank you for the opportunity to comment, and I will be looking forward to reviewing the final report. Sinqerely, . 'Thomas il. Fabian Executive Deputy Director TRF:mbm TF4.99900831/03 Page 70 GAO,‘iilWW19oBB ProfkabUity of Gumn teed Student Loans Qpendix XIII Major Contributors to This Briefing Report Joseph J. Eglin, Jr., Assistant Director, (202) 401-8623 Human Resources William A. Schmidt, Advisor Division, Washington, D.C. * Seatt1e Re@ona1 Office BeNjamin p. pfeiffer Evaluator Susie Anchell, Evaluator Keith C. Martensen, Evaluator . (lo4aaa) Page 71 GAO/HRD!W-13OBB Profitability of Guaranteed Student Loans
Guaranteed Student Loans: Profits of Secondary Market Lenders Vary Widely
Published by the Government Accountability Office on 1990-09-28.
Below is a raw (and likely hideous) rendition of the original report. (PDF)