SLM Corp
For the period ended Dec 31, 2023. Filed Feb 22, 2024.
View on SEC.govAbout this filing
A company's yearly report to the SEC, with audited financial statements. It is the fullest account a company gives of its business, its finances and the risks it sees.
FY2023 · Filed Feb 22, 2024
Extracted filing text, with the paragraph anchors used by the comparison.
See the whole filing, with its images and exhibitsContents
Critical Accounting Policies and Estimates
Management’s Discussion and Analysis of Financial Condition and Results of Operations addresses our consolidated financial statements, which have been prepared in accordance with GAAP. Notes to Consolidated Financial Statements, Note 2, “Significant Accounting Policies” in this Form 10-K includes a summary of the significant accounting policies and methods used in the preparation of our consolidated financial statements. The preparation of these financial statements requires management to make estimates and assumptions that affect the reported amounts of assets and liabilities and the reported amounts of income and expenses during the reporting periods. Actual results may differ from these estimates under varying assumptions or conditions. On a quarterly basis, management evaluates its estimates, particularly those that include the most difficult, subjective, or complex judgments and are often about matters that are inherently uncertain. The most significant judgments, estimates, and assumptions relate to the following critical accounting policies that are discussed in more detail below.
We maintain an allowance for credit losses for the lifetime expected credit losses on loans in our portfolios, as well as for future loan commitments, at the reporting date.
In determining the lifetime expected credit losses on our Private Education Loan portfolio loan segments, we use a discounted cash flow method. This method requires us to project future principal and interest cash flows on our loans in those portfolios.
To estimate the future expected cash flows, we use a vintage-based model that considers life of loan loss expectations, prepayments, defaults, recoveries, and any other adjustments deemed necessary, to determine the adequacy of the allowance at each balance sheet date. These cash flows are discounted at the loan’s effective interest rate to calculate the present value of those cash flows. Management adjusts the effective interest rate used to discount expected cash flows to incorporate expected prepayments. The difference between the present value of those cash flows and the amortized cost basis of the underlying loans is the allowance for credit losses. Entities that measure credit losses based on the present value of expected future cash flows are permitted to report the entire change in present value as credit loss expense, but may alternatively report the change in present value due to the passage of time as interest income. We have elected to report the entire change in present value as credit loss expense.
In determining the loss rates used for the vintage-based approach, we start with our historical loss rates, stratify the loans within each vintage, and then adjust the loss rates based upon economic factors forecasted over a reasonable and supportable forecast period. The reasonable and supportable forecast period is meant to represent the period in which we believe we can estimate the impact of forecasted economic factors in our expected losses. At the end of the reasonable and supportable forecast period, we immediately revert our forecasted economic factors to long-term historical loss conditions. We use a two-year reasonable and supportable forecast period, although this period is subject to change as our view evolves on our ability to reasonably forecast economic conditions to estimate future losses.
In estimating our current expected credit losses, we use a combination of expected economic scenarios coupled with our historical experience to derive a base case adjusted for any qualitative factors (as described below). We also develop an adverse and favorable economic scenario. At each reporting date, we determine the appropriate weighting of these alternate scenarios based upon the current economic conditions and our view of the risks of alternate outcomes. This weighting of expectations is used in calculating our current expected credit losses recorded each period.
In estimating recoveries, we use both estimates of what we would receive from the sale of defaulted loans as well as historical borrower payment behavior to estimate the timing and amount of future recoveries on charged-off loans.
We use historical experience and economic forecasts to estimate future prepayment speeds. At the end of the two-year reasonable and supportable forecast for prepayments, we immediately revert to our historical long-term prepayment rates.
In addition to the above modeling approach, we also take certain other qualitative factors into consideration when calculating the allowance for credit losses, which could result in management overlays (increases or decreases to the allowance for credit losses). These qualitative factors include, but are not limited to, changes in lending policies and procedures, including changes in underwriting standards, changes in servicing policies and collection administration practices, state law changes that could impact servicing and collection practices, charge-offs, recoveries not already included in the analysis, the effect of other external factors such as legal and regulatory requirements on the level of estimated current expected credit losses, the performance of the model over time versus actual losses, and any other operational or regulatory changes that could affect our estimate of future losses.
82 SLM CORPORATION—2023 Form 10-K
The evaluation of the allowance for credit losses is inherently subjective, as it requires material estimates that may be susceptible to significant changes. If actual future performance in delinquency, charge-offs, and recoveries is significantly different than estimated, or management assumptions or practices were to change, this could materially affect the estimate of the allowance for credit losses, the timing of when losses are recognized, and the related provision for credit losses on our consolidated statements of income.
When calculating our allowance for credit losses and liability for unfunded commitments, we incorporate several inputs that are subject to change period to period. These include, but are not limited to, CECL model inputs and any overlays deemed necessary by management. The most impactful CECL model inputs include:
Economic forecasts;
Weighting of economic forecasts;
Prepayment speeds; and
Recovery rates.
Management overlays can encompass a broad array of factors not captured by model inputs, including but not limited to, changes in servicing policies, collection administration practices, state law changes that could impact servicing and collection practices, and observed differences between forecasted and actual results. In the fourth quarter of 2022, we further increased our allowance for credit losses to reflect higher expected future periodic defaults in both the near term (reasonable and supportable period) and long term. This change reflected our estimate that the elevated default rates experienced in the latter half of 2022 would continue into 2023 and then decline over time. This estimate of future losses, like other aspects of our estimate of current expected credit losses, is susceptible to significant changes.
Of the model inputs outlined above, economic forecasts, weighting of economic forecasts, prepayments speeds, and recovery rates are subject to estimation uncertainty, and changes in these inputs could have a material impact to our allowance for credit losses and the related provision for credit losses.
In the fourth quarter of 2022, we changed our loss model to include forecasts of college graduate unemployment, retail sales, and median family income in determining the adequacy of the allowance for credit losses. Prior to this change, we included forecasts of college graduate unemployment and the Consumer Price Index in our loss forecasting models. We obtain forecasts for these inputs from Moody’s Analytics. Moody’s Analytics provides a range of forecasts for each of these inputs with various likelihoods of occurrence. We determine which forecasts we will include in our estimation of allowance for credit losses and the associated weightings for each of these inputs. At December 31, 2021, December 31, 2022, and December 31, 2023, we used the Base (50th percentile likelihood of occurring)/S1 (stronger near-term growth scenario with 10 percent likelihood of occurring)/S3 (downside scenario with 10 percent likelihood of occurring) scenarios and weighted them 40 percent, 30 percent, and 30 percent, respectively. Management reviews both the scenarios and their respective weightings each quarter in determining the allowance for credit losses.
In 2023, we experienced slower prepayment rates due to the rising interest rate environment. Historically, when rates rise loan prepayments and consolidation activity by third parties decline, and when rates decline loan prepayments and consolidation activity increase. During 2023, our estimates of future prepayment speeds reflect the current interest rate environment and future expectations of increased prepayment speeds in line with market expectations of a decline in interest rates based on the scenarios produced by Moody's Analytics described above. Slower prepayment speeds increase the allowance for credit losses because the loss rates applied in the future periods are applied to higher loan balances. We experienced higher prepayments during the COVID-19 pandemic, when unemployment rates were elevated, than we would have expected based upon our experience during past financial crises. As a result, during 2021 we increased our estimate of prepayment speeds to reflect higher short-term and long-term prepayment experience, which had a beneficial impact on the allowance for credit losses at that time.
2023 Form 10-K — SLM CORPORATION 83
A 100-basis point increase or decrease in the following inputs to the CECL loss model is estimated to change the allowance as follows:
| Estimated Increase (Decrease) to the Allowance for Credit Losses(1) | ||
|---|---|---|
| (Dollars in millions) | +100 Basis Points | -100 Basis Points |
| College graduate unemployment rate(2) | $62,135 | $(85,010) |
| Prepayment speeds(2) | (16,453) | 16,528 |
| Recovery rates(3) | (17,402) | 17,402 |
(1) Based on our Private Education Loan Portfolio at December 31, 2023.
(2)The estimated impacts of changes were calculated for the two-year reasonable and supportable period, but were not calculated for the remaining periods since long-term assumptions used to calculate the allowance for the remaining periods will not change.
(3)The estimated change in the recovery rate is based on long-term assumptions.
A 100-basis point increase or decrease in the retail sales or median family income does not result in material changes to our allowance for credit losses. Increases in the weighting of economic forecasts resulting in greater weight given to more severe economic forecasts would result in an increase in the allowance for credit losses.
The estimated impacts of changes in the above table were calculated for the two-year reasonable and supportable period, but were not calculated for the remaining periods since long-term assumptions used to calculate the allowance for the remaining periods are based on longer term averages and only change when we determine there is a fundamental change that will affect long-term rates.
Below we describe in further detail our policies and procedures for the allowance for credit losses as they relate to our Private Education Loan and FFELP Loan portfolios. During the third quarter of 2022, we reclassified our Credit Card loan portfolio to loans held-for-sale and subsequently sold the Credit Card portfolio to a third party in May 2023.
Allowance for Private Education Loan Losses
In addition to the key assumptions/estimates described above, some estimates are unique to our Private Education Loan portfolio. Estimates are made on our Private Education Loans regarding when each borrower will separate from school. The cash flow timing of when a borrower will begin making full principal and interest payments is dependent upon when the student either graduates or leaves school. These dates can change based upon many factors. We receive information regarding projected graduation dates from a third-party clearinghouse. The separation from school date is updated quarterly based on updated information received from the clearinghouse.
Additionally, when we have a contractual obligation to fund a loan or a portion of a loan at a later date, we make an estimate regarding the percentage of this obligation that will be funded. This estimate is based on historical experience. For unfunded commitments, we recognize the related life of loan allowance as a liability. Once the loan is funded, that liability transfers to the allowance for Private Education Loan losses.
Key Credit Quality Indicators - Private Education Loans
We determine the collectability of our Private Education Loan portfolio by evaluating certain risk characteristics. We consider credit score at original approval and periodically refreshed/updated credit scores through the loan’s term, existence of a cosigner, loan status, and loan seasoning as the key credit quality indicators because they have the most significant effect on the determination of the adequacy of our allowance for credit losses. Credit scores are an indicator of the creditworthiness of borrowers, and the higher the credit scores the more likely it is the borrowers will be able to make all of their contractual payments. Loan status affects the credit risk because a past due loan is more likely to result in a credit loss than a current loan. Additionally, loans in the deferred payment status have different credit risk profiles compared with those in current pay status. Loan seasoning affects credit risk because a loan with a history of making payments generally has a lower incidence of default than a loan with a history of making infrequent or no payments. The
84 SLM CORPORATION—2023 Form 10-K existence of a cosigner lowers the likelihood of default as well. We monitor and update these credit quality indicators in the analysis of the adequacy of our allowance for credit losses on a quarterly basis.
In the second quarter of 2023, we changed how we collect on defaulted loans. Previously, we used a mix of in-house collectors and sales to third parties. We will continue to sell a segment of defaulted loans immediately after charge-off but will no longer sell retained defaulted loans (that have been subject to internal collection attempts for six months) to third parties and instead will continue our collection efforts using in-house collectors and third-party collectors. This improved our estimate of recovery rates for the year ended December 31, 2023. When we estimate the timing and amount of future recoveries on charged-off loans, we no longer include expectations of future sales on retained defaulted loans. We continue to monitor how we collect on defaulted loans and may modify the approach from time to time based on performance, industry conventions, and/or regulatory feedback.
For December 31, 2022, we used both an estimate of recovery rates from in-house collections as well as expectations of future sales of defaulted loans to estimate the timing and amount of future recoveries on charged-off loans.
Private Education Loans generally do not require borrowers to begin principal and interest repayment until at least six months after the borrowers have graduated or otherwise separated from school. Consequently, the loss estimates for these loans are generally low while the borrower is in school and then increase upon the end of the grace period after separation from school. At December 31, 2023 and 2022, 25 percent and 24 percent, respectively, of the principal balance of the Private Education Loan portfolio was related to borrowers who were then in an in-school (fully deferred), grace, or other deferment status and not required to make payments.
Our collection policies for Private Education Loans allow for periods of nonpayment (forbearance) for certain borrowers requesting an extended grace period upon leaving school or experiencing temporary difficulty meeting payment obligations.
As part of concluding on the adequacy of the allowance for credit losses for Private Education Loans, we review key allowance and loan metrics. The most relevant of these metrics considered are the allowance coverage of net charge-offs ratio; the allowance as a percentage of ending total loans and accrued interest to be capitalized and of ending loans in repayment and accrued interest to be capitalized on loans in repayment; and delinquency and forbearance percentages.
We consider a Private Education Loan to be delinquent if the borrower has not made a required payment prior to the 31st day after such payment was contractually due.
Adoption of ASU No. 2022-02, “Troubled Debt Restructurings and Vintage Disclosures”
On March 31, 2022, the FASB issued ASU No. 2022-02, “Troubled Debt Restructurings and Vintage Disclosures” (“ASU No. 2022-02”), which eliminated the accounting guidance for troubled debt restructurings (“TDRs”) while enhancing disclosure requirements for certain loan refinancings and restructurings by creditors when a borrower is experiencing financial difficulty. The enhanced disclosures are required to be provided for modifications made starting in the period of adoption. Information about modifications in periods before adoption is not required to be provided.
ASU No. 2022-02 also requires that entities disclose current-period gross charge-offs by year of origination. For entities that have adopted the amendments in CECL, the amendment is effective for fiscal years beginning after December 15, 2022, including interim periods within those fiscal years.
Early adoption of the amendments in ASU No. 2022-02 was permitted if an entity has adopted CECL. The amendments should be applied prospectively. For the transition method related to the recognition and measurement of TDRs, an entity has the option to apply a modified retrospective transition method. We elected to early adopt all aspects of ASU No. 2022-02 prospectively for the period beginning January 1, 2022. The adoption was immaterial to our consolidated financial statements.
Troubled Debt Restructurings
For the year ended December 31, 2021, the allowance for our TDR portfolio was included in our overall allowance for Private Education Loans. In estimating the expected defaults for our Private Education Loans that were considered TDRs, we followed the same discounted cash flow process described above but used the historical loss rates related to past TDR loans. The appropriate gross loss rates were determined for each individual loan by evaluating loan maturity, risk characteristics, and macroeconomic conditions. Our TDR portfolio was comprised mostly of loans with interest rate reductions and loans with forbearance usage greater than three months, as further described below.
We adjust the terms of loans for certain borrowers when we believe such changes will help our customers manage their student loan obligations and achieve better student outcomes, and increase the collectability of the loans. These
2023 Form 10-K — SLM CORPORATION 85 changes generally take the form of a temporary forbearance of payments, a temporary or permanent interest rate reduction, a temporary or permanent interest rate reduction with a permanent extension of the loan term, and/or a short-term extended repayment alternative. Forbearance is granted prospectively for borrowers who are current in their payments and may be granted retroactively for certain delinquent borrowers.
Prior to January 1, 2022, we classified a loan as a TDR due to forbearance using a two-step process. The first step was to identify a loan that was in full principal and interest repayment status and received more than three months of forbearance in a 24-month period; however, during the first nine months after a loan had entered full principal and interest repayment status, we did not count up to the first six months of forbearance received during that period against the three-month policy limit. The second step was to evaluate the creditworthiness of the loan by examining its most recent refreshed FICO score. Loans that met the criteria in the first test and had a FICO score above a certain threshold (based on the most recent quarterly FICO score refresh) were not classified as TDRs. Loans that met the criteria in the first test and had a FICO score under the threshold (based on the most recent quarterly FICO score refresh) were classified as TDRs.
A loan also became a TDR when it was modified to reduce the interest rate on the loan (regardless of when such modification occurred and/or whether such interest rate reduction was temporary). Once a loan qualified for TDR status, it remained a TDR for allowance purposes for the remainder of its life. About half our loans that were considered TDRs involved a temporary forbearance of payments and did not change the contractual interest rate of the loan.
Off-Balance Sheet Exposure for Contractual Loan Commitments
When we approve a Private Education Loan at the beginning of an academic year, that approval may cover the borrowing for the entire academic year. As such, we do not always disburse the full amount of the loan at the time of such approval, but instead have a commitment to fund a portion of the loan at a later date (usually the start of the second semester or subsequent trimesters). We estimate expected credit losses over the contractual period in which we are exposed to credit risk via a contractual obligation to extend credit, unless that obligation is unconditionally cancellable by us. The discounted cash flow approach described above includes expected future contractual disbursements. The portion of the allowance for credit losses related to future disbursements is shown as a liability on the face of the balance sheet, and related provision for credit losses is reflected on the income statement.
Uncollectible Interest
The majority of the total accrued interest receivable on our Private Education Loan portfolio represents accrued interest on deferred loans where no payments are due while the borrower is in school and on fixed-pay loans where the borrower makes a $25 monthly payment that is smaller than the interest accrued on the loan in that month. The accrued interest on these loans will be capitalized and increase the unpaid principal balance of the loans when the borrower exits the grace period after separation from school. The discounted cash flow approach described above considers both the collectability of principal as well as this portion of accrued interest that is expected to capitalize to the balance of the loan. Therefore, the allowance for this portion of accrued interest balance is included in our allowance for credit losses. The discounted cash flow approach does not consider interest accrued on loans that are in a full principal and interest repayment status or in interest-only repayment status. We separately capture the amount of expected uncollectible interest associated with these loans using historical experience to estimate the uncollectible interest for the next four months at each period-end date. This amount is recorded as a reduction of interest income. Accrued interest receivable is separately disclosed on the face of the balance sheet.
Allowance for FFELP Loan Losses
FFELP Loans are insured as to their principal and accrued interest in the event of default, subject to a risk-sharing level based on the date of loan disbursement. These insurance obligations are supported by contractual rights against the United States. For loans disbursed on or after July 1, 2006, we receive 97 percent reimbursement on all qualifying claims. For loans disbursed after October 1, 1993, and before July 1, 2006, we receive 98 percent reimbursement on all qualifying claims. For loans disbursed prior to October 1, 1993, we receive 100 percent reimbursement. Because we bear a maximum of three percent loss exposure due to this federal guarantee, our allowance for credit losses for FFELP Loans and related periodic provision expense are relatively small.
We use the gross loss approach when estimating the allowance for credit losses for the unguaranteed portion of our FFELP Loans. We maintain an allowance for credit losses for our FFELP Loans at a level sufficient to cover lifetime expected credit losses. The allowance for FFELP Loan losses uses historical experience of customer default behavior. We apply the default rate projections, net of applicable risk sharing, to our FFELP Loans for the current period to perform our quantitative calculation. Once the quantitative calculation is performed, we review the adequacy of the allowance for credit losses and determine if qualitative adjustments need to be considered.
86 SLM CORPORATION—2023 Form 10-K
The text of this filing, as the company filed it with the SEC, without its images or exhibits. This is not investment advice.