Sallie Mae Q3 2026 Earnings: Record $3.4B Originations and a $0.19 Loss

On October 22, 2026, Sallie Mae reported the largest origination quarter in company history — $3.4 billion of private student loans, up 15%, with graduate lending up 119% — and lost ($0.19) per share doing it. The arithmetic is pure student-lending seasonality amplified by policy change: $213.7 million of credit provisions and $220.7 million of peak-season expenses front-ran the interest income those new loans will earn for years, while loan-sale gains collapsed 97% to $3.7 million. Management then cut full-year EPS guidance by more than a third, to $1.95–$2.05. Our new SLM ledger models two audited annuals plus this quarter in plain-text double entry, so every line below reconciles to the filing.

The Headline Numbers​

Every figure below comes from the Form 10-Q and the earnings release (8-K Exhibit 99.1), which reprints the same statements:

Three things stand out. First, the revenue collapse is almost entirely the loan-sale line going quiet: non-interest income fell $136.6 million while net interest income fell only $33.7 million, so four-fifths of the top-line decline is gains that did not repeat, not a shrinking book. Second, the loss is front-loaded investment by the release's own accounting — CEO Jonathan Witter attributes the ($0.19) print "primarily due to provision and expenses related to our highest ever year-over-year originations growth," and the two lines together ($434.4 million) exceed total revenue ($375.5 million). Third, the share count keeps cushioning the per-share damage: diluted shares fell 12% year over year to 185.6 million on buybacks, so the ($0.19) would have been ($0.21) at last year's count. The quarter also swung hard sequentially, from +$59 million of net income in Q2 to ($32.3 million), as revenue fell 6% and provisions jumped 69% quarter over quarter.

Revenue Deep Dive​

Net interest income fell 9% to $339.3 million as both sides of the balance sheet shrank their contribution — interest income down 6%, interest expense down only 3%:

The mix shift inside interest income is the balance sheet rotating in real time. Cash balances fell by half from year-end (to $2.17 billion) while available-for-sale securities grew 62% (to $2.85 billion), so investment income more than doubled to $31.9 million while cash income halved to $25.4 million — parked cash becoming earning securities. Loan interest income fell 6% despite record originations because new disbursements earn for weeks in Q3 while the release's loan-sale program spent the last year shrinking the seasoned book: gains on sales printed $323.7 million in the first nine months of 2025 against $164.9 million this year. On the funding side, the release points to "a lower cost of funds, 4.21%, compared to 4.24% in the year-ago quarter, due in part to actions taken this year to extend the duration of our fixed-rate funding" — three basis points of relief, real but small against a $277 million quarterly funding bill, and deposits actually slipped 1.4% from year-end to $20.76 billion while short-term borrowings nearly doubled to $898.7 million.

Non-interest income is where the quarter breaks year over year, down 79%:

Only $3.7 million of loan-sale gains against $136.0 million a year ago is the single biggest delta in the entire income statement — bigger than the provision increase and the expense increase combined. The release's answer is timing plus pipeline: a new strategic partnership with "a leading global investment firm" closed during the quarter with an initial $143 million seasoned portfolio sale, and loans held for sale swelled 55% from year-end to $1.44 billion, which reads as inventory staged for future sales rather than demand destroyed. Other income growing 46% to $46.5 million is the quiet offset. The demand signal underneath is unambiguous in management's words: "the largest single quarter of originations in the Company's history," with "119% growth in Graduate Loan originations during the first academic year following PLUS reform," at an average approval FICO of 757 — up year over year — with cosigner rates "strong at 92%." Record volume at record credit quality is the bull case in one sentence; everything else this quarter is the cost of capturing it.

The Margin Story​

A lender's margin table is really a cost-of-growth table, and this quarter every ratio broke the wrong way:

Mechanically, two lines explain the whole trip from a 34.1% pre-tax margin to −15.7%: provisions took 24 points more of revenue and expenses took 26 points more, and revenue itself shrank by nearly a third, so both ratios were divided by a smaller number. Provisions deserve the closer look because they are two stories at once. The CECL story is benign — day-one provisions on $3.4 billion of originations, 15% more than last year, are the accounting entry for growth, and the allowance ratio on private education exposure actually improved to 5.89% from 5.93%. The credit story is less clean: net charge-offs rose 17% to $90.4 million, "primarily driven by misaligned third-party debt resolution practices affecting a small, high-ability-to-pay segment of borrowers progressing straight through delinquency to default," with management conceding "borrower adoption of these strategies has been slower than anticipated." Delinquencies ticked to 4.2% of loans in repayment from 4.0%, and forbearance rose to 2.3% from 2.0%. None of that is a crack — but it is why the full-year charge-off guide went up to $400–$420 million from $365–$385 million, and the release prices no further improvement into 2026.

The One Big Question: Is a ($0.19) Loss the Right Price for the PLUS Window?​

Federal PLUS reform just handed private lenders the biggest demand shock in a generation — graduate originations up 119% in the first academic year — and Sallie Mae chose to spend through it: provisions +19%, expenses +22%, loan-sale gains deferred, and a quarterly loss. The question is whether that trade was violently correct or quietly alarming, and the release gives both sides ammunition. The correct side: average FICO 757 and rising, 92% cosigned, delinquencies "within expectations" at 4.2%, and an allowance ratio that improved while the book grew — this is not subprime chasing. The alarming side is the guidance cut sitting two paragraphs below the record: full-year diluted EPS slashed to $1.95–$2.05 from $3.10–$3.20, a 37% cut at the midpoint, while the originations-growth target (12–14%) and the expense range ($750–$780 million) were both left untouched.

Read those two facts together and management is saying the volume is real, the spending is deliberate, and the earnings shortfall comes from credit and timing — higher charge-offs plus gains that slipped out of 2026. The nine-month math frames what is left: $1.68 of diluted EPS banked through September against a $1.95–$2.05 full year, so Q4 needs $0.27–$0.37 — a return to profit, but no heroics after a ($0.19) quarter. The balance sheet can fund the wait: deposits flat near $20.8 billion, the AFS book throwing off double the income, and $111 million of buyback capacity still authorized. What it cannot do is validate the trade — only 2027 net interest income, earned on loans disbursed this quarter, does that. A lender that originates $3.4 billion at FICO 757 has bought itself an earnings stream; this quarter's loss is the down payment, and the guidance cut is management admitting the financing costs more than it thought.

Tracking a $30B Lender in Plain Text​

Double-entry bookkeeping forces every dollar to reconcile, which is exactly what a bank balance sheet needs: twenty-nine billion of assets funded by deposits and borrowings down to the thousand. Read how we model every company, then look at the quarter as the ledger records it — income negative (credit), expenses positive (debit), summing to zero:

; FY2026Q3 Income Statement -- three months ended September 30, 2026
; Check: -375.467 + 213.675 + 220.702 + -26.649 + -32.261 = 0
 
2026-09-30 * "SLM Corporation" "FY2026Q3 Income Statement"
  Income:Revenue                               -375.467 MUSD  ; total revenue (NII 339.322: interest income 616.482 - interest expense 277.160; non-interest income 36.145: loan-sale gains 3.650 + securities (13.970) + other 46.465)
  Expenses:CostOfRevenue                         213.675 MUSD  ; provisions for credit losses (net charge-offs 90.386 excl. 19.1 clean-up-call buybacks; allowance ended at 1449.614)
  Expenses:OperatingExpenses                      220.702 MUSD  ; total non-interest expenses (compensation 98.443 + FDIC 5.455 + other operating 116.169 + acquired intangible amortization 0.635); +22% YoY on peak-season volume capture post-PLUS reform
  Expenses:IncomeTax                              -26.649 MUSD  ; income tax benefit on the pre-tax loss of 58.910
  Equity:Adjustments                               -32.261 MUSD  ; net loss offset, attributable to SLM Corporation (no NCI as filed; preferred dividends 3.643 are a distribution)

The balance-sheet number that tells the narrative is loans held for sale: $933.3 million at year-end 2025 to $1,442.1 million at September 30, up 55% in nine months while sale gains nearly vanished. That is $1.44 billion of staged inventory — either next quarter's gains or evidence the bid went away — and it is the line to watch when Q4 prints. The quieter tell is treasury stock at $4.26 billion, larger than all paid-in capital combined, which is why the ledger's equity account holds a debit balance: Sallie Mae has returned so much capital that cumulative buybacks exceed everything shareholders ever paid in.

The Multi-Year Arc​

Two audited years, the 2023 comparatives, and the nine-month trail — revenue, provisions, and the loan book that connects them:

The compounding story through 2025 was provisions discipline plus loan-sale gains: revenue grew 10% in two years while provisions fell 4%, and net income compounded 28% from $581.4 million to $744.8 million with the net margin widening five points to 37.5%. Note what did not compound — the loan book, dead flat near $20.5 billion for three years, because originations flowed out through sales almost as fast as they were made. That is what makes 2026 the regime change: nine-month provisions of $327.9 million have nearly matched all of FY2025 with the seasonally heaviest quarter still ahead, nine-month revenue is down 13% year over year while nine-month net income is down 35%, and the held-for-sale shelf is filling instead of clearing. If the PLUS window is structural, 2026 reads in hindsight as the year Sallie Mae stopped harvesting and started planting. If loan-sale demand stays soft into 2027, it reads as the year the gain-on-sale model broke at peak originations.

The Verdict: Bull vs. Bear​

Bull Case

• Record $3.4 billion originations at FICO 757 with 92% cosigned compound into net interest income for a decade; CECL day-one provisions make Q3 look worse than the economics by construction.
• Credit is holding where it counts: delinquencies 4.2% "within expectations," allowance ratio improved to 5.89%, and the charge-off increase is traced to identifiable third-party practices with remediation "gaining traction."
• Funding is stable and cheapening: cost of funds down to 4.21% on duration extension, deposits steady near $20.8 billion, and the rotated AFS book doubled investment income.
• Capital return continues through the loss: $132.2 million of buybacks in Q3 (5.1 million shares), $111 million of authorization left, and the $0.13 dividend held.
• The new strategic partnership plus $1.44 billion of held-for-sale inventory stages a loan-sale rebound that would refill exactly the line — gains — that caused this quarter's revenue hole.

Bear Case

• Management cut its own outlook 37% at the midpoint while keeping the volume and expense targets — it is telling you the shortfall is credit and timing, the two things it controls least.
• Loan-sale gains fell 97% to $3.7 million precisely when originations surged 15%; if the bid for seasoned paper has thinned, the originate-to-distribute model faces its worst quarter of pipeline with no exit.
• Net charge-offs up 17% with forbearance rising to 2.3% and borrower adoption of remediation "slower than anticipated" reads as early-cycle deterioration, not noise — and the charge-off guide went up, not down.
• Expenses up 22% "to capture new volume and customers following PLUS reform" may be the new run rate, not a one-time peak: competing for PLUS-refugee borrowers every fall costs money every fall.
• Q4 must print $0.27–$0.37 to reach even the lowered guide after a ($0.19) Q3, with no buyback-fueled share-count help left at anything like the 12% year-over-year shrinkage just banked.

Our Take: Buy the window, watch the exits. A 119% surge in graduate lending at rising FICO is the kind of demand shock lenders dream about, and spending 22% more to capture it while provisions front-run income is textbook-correct behavior for a balance sheet this clean — but the 97% collapse in sale gains and the 37% guidance cut mean the market will not pay for 2027 NII until Q4 proves the held-for-sale shelf can clear. The Q4 print is the whole thesis: gains north of $100 million and the loss quarter becomes legend; another single-digit gain and the originate-to-distribute model needs a rewrite.