U.S. Student Loan ABS

U.S. Student Loan ABS Dominion Bond Rating Service THE CORE CURRICULUM MARCH 2005 FREDERIC P. VIGNERON MICHAEL NELSON

INTRODUCTION Student loan securitizations fall into two broad categories: U.S. Government-Backed Student Loan Securitizations; and Private (or Alternative ) Student Loan Securitizations. Some student loan securitizations include both U.S. government-backed and alternative loan portfolios, either commingled or segregated into separate classes. The demand for student loan monies, especially in the form of non-government backed alternative loans, is growing and is expected to continue to expand in the foreseeable future. The primary engine for this growth is the widening gap between the all-in cost of post-secondary education and the resources of students and their families. Simply put, the cost of both college and graduate school education in the U.S. has risen, and continues to rise, disproportionately to: (1) periodic increases in U.S. federal government maximum education grant and loan levels, (2) the overall cost of living index, and (3) median household incomes in current U.S. dollars. This economic pressure, in combination with the ever-increasing overall U.S. domestic (as well as international) demand for higher education, has produced an unprecedented need for alternative funding sources. A comparison of student loan funding and securitization statistics indicates the vast majority of alternative student loans are ultimately financed by the capital markets through the use of securitization vehicles. Accordingly, the ever-increasing volume of student loan asset-backed securities (SLABS) has intensified institutional investor preferences for objective measures of their creditworthiness, such as a Dominion Bond Rating Service ( DBRS ) rating. The following table summarizes the volume of student loancollateralized security issuance from 2001 to 2004. This data does not include municipal-related student loan issuance. For 2004, market sources generally agree that total student loan-related issuance, including municipal-related issues, was approximately 60 billion. Total U.S. SLABS Issuance Year to Date Dec. 31, 2004 Dec. 31, 2003 Dec. 31, 2002 Dec. 31, 2001 (US millions) % Change % Change % Change Data Sallie Mae FFELP Stafford/PLUS loans 10,002.0 73.3% 5,772.0-47.7% 11,033.0 71.3% 6,441.0 SLM Consolidation loans - -100.0% 4,256.0 115.4% 1,976.0 - - SLM Private credit student loans 2,535.0-27.6% 3,503.0 407.7% 690.0 - - SLM Total securitization sales 12,537.0-7.3% 13,531.0-1.2% 13,699.0 112.7% 6,441.0 SLM Asset-backed commercial paper 4,186.0 - - - - - - SLM On-balance sheet consolidation loan VIEs 17,124.0 3.2% 16,592.0 - - - - SLM Total securitization financings 21,310.0 28.4% 16,592.0 - - - - SLM Total Sallie Mae Loan Securitizations 33,847.0 12.4% 30,123.0 119.9% 13,699.0 112.7% 6,441.0 SLM Other Student Loan Securitizations 16,175.7 11.2% 14,540.6 19.5% 12,172.9 147.8% 4,911.4 Calc. Total Student Loan Securitizations 50,022.7 12.0% 44,663.6 72.6% 25,871.9 134.1% 11,050.7 ABA Sources: Asset-Backed Alert (ABA); Sallie Mae (SLM Corporation). This report has been drafted to provide guidance: (1) to originators, lenders, or facilitators as they prepare new securitizations, and (2) to institutional investors as they evaluate their portfolio allocations. Information comes from sources believed to be reliable, but we cannot guarantee that it, or opinions in this Study, are complete or accurate. This Study is not to be construed as an offering of any securities, and it may not be reproduced without our consent.

U.S. Student Loan ABS: The Core Curriculum - Page 2 GOVERNMENT VERSUS ALTERNATIVE STUDENT LOAN MARKETS (1) U.S. Government Loan Programs: Program Compliance-Driven Servicing The U.S. government offers two basic student loan programs to students and/or their parents: (a) direct, and (b) guaranteed loans. The Federal Direct Student Loan Programs (FDSLP) are exclusively funded by the U.S. government. Consequently, the capital markets do not play a role in these programs (except indirectly through the U.S. Treasury market). Federally guaranteed loans, however, are almost always funded by the non-government capital markets. The origination and packaging of these loans is, basically, a Higher Education Act of 1965 (HEA)-authorized multi-step process. (1) Academic institution financial aid offices refer borrowers to a choice of banks and other eligible lending institutions including: (a) pension funds; (b) insurance companies; (c) one of the state or private not-for-profit state agencies; and (d) subject to certain limitations, the academic institutions themselves. (2) The loans are approved and disbursed (i.e. originated ) by the lender of choice in accordance with government guarantee eligibility rules, as monitored by the HEAauthorized, state-sponsored, not-for-profit guarantee agencies (which act as the primary, DOE-reinsured guarantors for the loans). (3) Secondary market institutions purchase and package for sale to the capital markets eligible loans from the great majority of originators that chose not to keep the loans on their books. These institutions include: - First and foremost, Sallie Mae, the former government-sponsored enterprise (GSE) and dominant student loan market player; - The growing segment of other non-government financial institutions with student loan securitization expertise; and - The individual state-designated private, not-for-profit agencies through either: (a) tax-exempt, or (b) taxable financings (if their tax-exempt statutory issuance limits have been reached). Secondary market institutions also often act as primary market originators by making loans to students directly and then packaging their own portfolios for securitization. Under the current federal rules, these loans are ultimately guaranteed for usually at least 98% (if not often 100%) of principal and interest through the federal government s reinsurance of state-sponsored guarantors. Hence, obligor credit risk has, historically, not been a primary consideration for this form of collateral. All of these programs, however, are very closely monitored by the Department of Education (DOE) with respect to servicer compliance. Non-compliance can jeopardize the government s guaranty, thereby potentially compromising the ultimate creditworthiness of securitizations backed by student loan pools serviced by these servicers. The primary risk to the ultimate recoverability of invested funds lies with the servicer s ability to satisfy precisely the meticulous administrative requirements set forth by the DOE. Collateral-specific initial and ongoing servicer due diligence is, therefore, indispensable to the proper underwriting of federally guaranteed loan securitizations. As a result, the federal government conducts its own reviews of the major servicers. DBRS believes that understanding the government review methodologies is an appropriate component of the due diligence process. Note: As of the publication date of this study, proposed changes to the HEA (such as the reduction of loan guaranty percentages) would require a reassessment of both student loan pool credit risk and their prepayment assumptions. Federal Programs Federal Pell Grant Federal Supplemental Educational Opportunity Grants Federal Work-Study Federal Perkins Loans Grants and Work-Study x x x Direct Loans x Third-Party Funded/ Government Guaranteed Loans Federal Stafford Loans (subsidized) x x Stafford loans (unsubsidized) x x Parent loans for undergraduate students (plus) x x Consolidation Loans x x Securitized Loans

U.S. Student Loan ABS: The Core Curriculum - Page 3 U.S. Government Guaranteed Student Loan Program Flow Chart Servicer P&I Payments Remittances Servicing Fee Special Allowance/ Interest Subsidy Payments Guaranty Claim Reinsurance Payment U.S. Department of Education Reinsurance Claim Student Borrower Financial Aid & Loan Applications Loan Disbursement Lender 98% or 100% Guaranty Payment Guaranty Fee Guarantor School Financial Aid Office School-Certified Loan Application Guaranty Application Guaranty Issuance (2) Alternative Student Loans: Collateral Performance- Driven Servicing Alternative student loans do not benefit from ultimate government guarantees. Consequently, the critical underwriting criteria for alternative student loans are driven by the individual obligors creditworthiness. In this respect, SLABS are comparable to other forms of consumer creditbacked securitizations. Alternative student loan securitization obligor credit risk can be either partially or wholly mitigated at: (a) The individual loan level through guarantees issued by private guarantors such as The Education Resources Institute (TERI); (b) The security level by one of the monoline financial guarantee insurers; or (c) The structural level through the use of a classic multiclass, senior/subordinated cash flow waterfall. Many alternative student loan securitizations incorporate a combination of these credit enhancement techniques. The challenge for originators of student loans centers on the relative lack of credit history for most student borrowers. While still essentially indispensable to the underwriting process, even FICO scores are of limited value for this segment of the borrowing population. FICO scores are more meaningful, however, for graduate students (who, being older, usually have several years of credit history) and for parent co-signed loans (which reflect the creditworthiness of a parent). As with government programs, the servicing process is also a critical component of the private student loan industry. For alternative loan originators/lenders, however, the definition of successful servicing is performance-related and less a function of satisfying a rigid set of government rules. Success in this segment is defined by the ability of the servicer to collect on the loan itself rather than simply relying on the guarantees (even for the privately guaranteed transactions). In short, the servicing paradigm for alternative loans is performance driven rather than compliance driven. (3) Government-Subsidized Versus Market-Priced Loan Interest Rates From the perspective of student borrowers, DOE-sponsored loans are less expensive than options available from the alternative lenders. Loans with rates below unsecured consumer debt market rates are available to students through the U.S. government direct loan programs because they are directly funded by the federal government itself, the most competitive of borrowers. For U.S. governmentguaranteed, but non-government-funded loans, however, the government entices the private sector to participate through a Special Allowance Payment (SAP) interest subsidy program paid directly to the lenders. The government also pays lenders additional interest subsidies for those loans that do not accrue interest when the borrowers are in school. Conversely, alternative loan borrowers do not benefit from these subsidies and are obligated to pay a coupon more accurately reflecting their true credit risk as well as their lenders funding costs.

U.S. Student Loan ABS: The Core Curriculum - Page 4 REQUIRED INFORMATION While there are substantial differences between the government and private student loan markets, most of the core analytical characteristics for both collateral segments often overlap or are essentially identical. For practical purposes, therefore, DBRS has created a single list of required information. Required Student Loan Pool Data Pool Statistics (1) Outstanding principal balance (2) Total accrued interest (3) Total outstanding principal and accrued interest (4) Number of borrowers (5) Number of co-borrowers (6) Number of loans (7) Annual interest rate weighted average (8) Annual interest rate in repayment weighted average (9) Original term to maturity (months) weighted average (10) Remaining term to maturity (months) weighted average (11) Delinquency status loan pool (12) FICO, DTI: all loans weighted average (13) FICO, DTI: co-signed loans weighted average (14) FICO, DTI: non-co-signed loans weighted average (15) Guaranty/insurance program claim rates Loan & Obligor Level Statistics (16) Principal balance per loan (max./min./mean/median) (17) Principal balance per borrower (max./min./mean/median) (18) Annual interest rate (max./min./mean/median) (19) Annual interest rate in repayment (max./min./mean/median) (20) Original term to maturity (months) (max./min./mean/median) (21) Remaining term to maturity (months) (max./min./mean/median) (22) FICO, DTI: All loans (max./min./mean/median) (23) FICO, DTI: co-signed loans (max./min./mean/median) (24) FICO, DTI: non-co-signed loans (max./min./mean/median) (25) Obligor school/degree type (26) Loan status (in-school/repayment/deference/forbearance/claim, etc.) (27) Loan geographic concentrations state (28) Loan geographic concentrations zip code (29) Dependency status (independent; indep. w/ & w/out dependents) (30) Obligor age (max./min./mean/median) (31) Obligor default history (32) Student grade point average (max./min./mean/median) (33) Student age (max./min./mean/median) Pool Academic Institution Statistics (34) Academic institution concentration (by loan balance & obligors) (35) Academic institution type (36) Cohort rate weighted average (37) Cohort rate (max./min./mean/median) Pool Other Data (38) Loans by guarantor distribution (39) Loans by DOE SAP interest rate index (40) Loans by servicer distribution Available? (Y/N) Applicable? (Y/N)

U.S. Student Loan ABS: The Core Curriculum - Page 5 ANALYTICAL METHODOLOGY (1) Portfolio Valuation DBRS assesses the portfolio quality based on applicant-provided pool and loan-/obligor-level data as to both credit risk and estimated overall payment speed within the context of both the proposed transaction s cash flow structure and, on a comparative basis, with previously issued student loan securitizations. Particularly in the fragmented private student loan market, access to historical performance, including claims to guarantors, becomes a crucial element in the analysis. While cohort default history is incorporated into the guarantee formula for U.S. government loans, such information is also vital in determining enhancement levels for private loan programs. If private lenders incorporate proprietary credit scores in determining underwriting standards, DBRS will review such data to determine if the scores correlate with loan performance. Portfolio analysis will also include determining how factors such as borrower status (parent or child), field of study, academic institution, geographic concentration, and loan term affect portfolio performance. (2) Cash Flow Modeling DBRS will analyze the proposed transaction and request ratings-driven cash flow stress scenarios, applying assumptions consistent with the student loan pool characteristics. For example, DBRS will provide prepay and loss stresses by rating category based on the particular transaction. Key analytical considerations will include the following: (a) Loan default volume, severity, and timing, (b) Applicable guaranty/insurance levels, claim rates, and payment delays, (c) DOE Special Allowance Payment disbursement delay, (d) Loan pool prepayment and extension risks, (e) Any form of pool dilution, (f) Actual (total) losses, (g) Formula and conditions for servicer advances related to the replacement or purchase of loans not serviced according to contracted procedures, (h) Interest rate and basis risks, and (i) Applicable reinvestment rates. Such analysis will take into consideration timing of the student loan life cycle, such as indicated below. Additionally, DBRS will perform both static pool and vintage analysis on relevant student loan portfolios. DBRS will also reconcile the securitization structure s sources and uses of funds. GOVERNMENT GUARANTEED STUDENT LOAN LIFE CYCLE Phase 1 Phase 2 Phase 3 Origination Interim Repayment Scheduled Accrual & Repayment Periods +/- Adjustments = Total Interim periods applicable to Stafford loans only; not to PLUS or Consolidation. Deferment + 6-12 months; max. three years Forbearance + 3-6 months; max. three years Apply School Grace Repayment Maturity Variable + 6-9 months + 10-30 years Default = x years Subject to: ½ time enrolment Begins at: Graduation; Withdrawal; or < ½ time enrolment - Months to maturity from claim payment date (+ 9-12 months min./max. guarantor response) Discharge - Months to maturity from discharge date

U.S. Student Loan ABS: The Core Curriculum - Page 6 (3) Structural and Legal Evaluation DBRS will examine all applicable legal opinions to assess the strength of the proposed transaction s bankruptcy-remote structure. As student loan-backed securities are originated/issued by one of three types of entities (state agencies, not-for-profit organizations, and for-profit companies), the appropriate sections of the U.S. Bankruptcy Code should be applied in assessing the potential for special-purpose entity (SPE) consolidation with respect to the obligations of the student borrower, the loan seller, and the SLABS SPE itself. In providing an exception to an individual debtor s bankruptcy discharge for any debt, Section 523(a)(8) of the Code addresses student loans specifically: for an educational benefit overpayment or loan made, insured or guaranteed by a governmental unit, or made under any program funded in whole or in part by a governmental unit or non-profit institution, or for an obligation to repay funds received as an educational benefit, scholarship or stipend, unless excepting such debt from discharge under this paragraph will impose an undue hardship on the debtor and the debtor s dependents. 11 U.S.C. 523(a)(8). The following table summarizes the key issues applicable to the transaction participants for all three categories of originator-issuers. Type Transaction Participant Bankruptcy Type Issuer Student/Parent Borrower Loan Seller SLABS SPE State Agencies Not-for-profit Organizations ( N-F-P ) For-profit Companies Chapter(s): 13 Chapter(s): 9 Chapter(s): N/A Cannot be discharged under Subject to verification of State agencies do not issue the Code except by showing seller classification as SLABS through SPVs. State substantial hardship. governmental unit, agency student loan bonds Application of Chapter 13 bankruptcy risk for state are collateralized direct may reduce other debts, agencies effectively issues of the state agency. however. Poor lender considered diminimis. bookkeeping may compromise loan enforceability. Chapter(s): 13 Chapter(s): N/A Chapter(s): N/A Cannot be discharged under May file for voluntary N-F-Ps do not issue SLABS the Code except by showing bankruptcy, but unlikely through SPVs. N-F-P student substantial hardship. given the inability of loan bonds are collateralized Application of Chapter 13 creditors to force the seller direct issues of the N-F-P. may reduce other debts, into involuntary bankruptcy however. Poor lender if the seller is properly bookkeeping may compromise classified as an N-F-P within loan enforceability. its domicile state. Counsel s opinion that the seller meets the N-F-P exception to code is advisable. Chapter(s): 13 Chapter(s): 7 or 11 Chapter(s): 7 or 11 Could be discharged only if the loan process is not connected in any way to a notfor-profit organization. Otherwise, cannot be discharged under the Code except by showing substantial hardship. Application of Chapter 13 may reduce other debts, however. Poor originator/servicer bookkeeping may compromise loan enforceability. Standard review of selleroriginator counsel s true sale and non-consolidation opinions applicable to all ABS assets. Standard review of issuer counsel s bankruptcy-remote opinion applicable to all ABS assets.

U.S. Student Loan ABS: The Core Curriculum - Page 7 DUE DILIGENCE DBRS will schedule meetings with the appropriate officers of the proposed transaction s originator/facilitator and its master servicer. For government-guaranteed loans, the servicer s ability to be in compliance with the DOE guaranty regulations is stressed. For alternative loans, collection procedures are key. The financial strength of the servicer is also taken into consideration, as is that of any selected alternate servicer. CONCLUSION The analysis of student loan asset-backed securities requires a special combination of industry knowledge and product-focused securitization expertise. While the credit risk of student loan pools resembles other forms of non-collateralized consumer debt, this asset class is distinctive with respect to: (1) the relatively limited credit history of most borrowers, (2) the loans irregular stop and go payment terms, and (3) the industry s highly specialized operational requirements. SLABS cash flow structures reflect these characteristics. Consequently, DBRS has designed analytical criteria that reflect the unique nature of this asset class and its securitization structures. As the student loan securitization market evolves, DBRS will continue to reassess its methodology in a manner that is both analytically comprehensive for investors and operationally transparent for issuers. ANALYST CONTACT INFORMATION Frederic P. Vigneron Ph: 212 806 3254 fvigneron@dbrs.com Michael Nelson Ph: 212 806 3251 mnelson@dbrs.com