WASHINGTON, July 14, 2026, 08:14 (EDT)
About a million borrowers have exited the government’s SAVE student-loan program as it winds down. But early data shows another issue: around 20% of these borrowers picked repayment plans that will also close in 2028. Only about 13% of the SAVE program’s original 7.5 million borrowers have left so far, so another wave of forced plan changes could be ahead before the current one wraps.
Loan servicers sent out 90-day notices to roughly 250,000 borrowers in early July and are sending another batch this week. Borrowers who don’t act within their window will be put on one of two default repayment plans. “We wanted to make sure that we had the operational support with our servicers,” Education Under Secretary Nicholas Kent told the Wall Street Journal. The Wall Street Journal
The $1.7 trillion federal loan portfolio is seeing more pressure. As of March 31, about 9 million borrowers were in default on $220 billion, and 3.5 million borrowers with active loans were over 30 days late. For many, the wrong repayment plan or a missed notice can turn a policy tweak into real cash-flow trouble at home.
Early data on where borrowers went is surprising. About half picked Income-Based Repayment, or IBR. Thirty percent chose the new Repayment Assistance Plan, or RAP. The rest—20%—went into Pay As You Earn or Income-Contingent Repayment. That doesn’t match what Education Department charts suggest: RAP payments come in lower than IBR payments for single borrowers earning $30,000 to $70,000 and no dependents. The Education Department hasn’t shared movers’ actual incomes, but the pattern fits with some borrowers wanting IBR’s shorter forgiveness or sticking to what they know.
| Plan selected | Share of first exits | Implied borrowers from first nearly 1 million | Main trade-off |
|---|---|---|---|
| IBR | Roughly half | Close to 500,000 | Forgiveness offered after 20 or 25 years, based on when the loan originated |
| RAP | About 30% | About 300,000 | 30-year repayment; main subsidies need on-time payment |
| PAYE or ICR | Roughly 20% | Roughly 200,000 | Both plans phase out July 1, 2028 |
Numbers are rough estimates, taken from the reported destination shares.
The 20% share could be more important than how many moved over so far. If that trend held for all 7.5 million SAVE borrowers, around 1.5 million would join a plan only to have to exit by July 2028. That’s a sensitivity scenario, not a prediction—early movers may not look the same as those who end up switching closer to a servicer’s deadline.
RAP isn’t always cheaper. Adjusted gross income is income after some tax changes. The department’s numbers show RAP is $22 a month more than IBR for those making $20,001 to $30,000. It saves $41 for borrowers earning $50,001 to $60,000. But RAP costs $110 more in the $90,001-to-$100,000 bracket.
| Adjusted gross income | IBR monthly payment | RAP monthly payment | RAP minus IBR |
|---|---|---|---|
| $20,001–$30,000 | $20 | $42 | +$22 |
| $30,001–$40,000 | $103 | $88 | -$15 |
| $50,001–$60,000 | $270 | $229 | -$41 |
| $70,001–$80,000 | $437 | $438 | +$1 |
| $90,001–$100,000 | $603 | $713 | +$110 |
| $100,001–$110,000 | $687 | $875 | +$188 |
The chart is based on a single borrower, no dependents, and the 2024 federal poverty guideline. Numbers below zero in the last column mean RAP costs less.
RAP has a tough payment-timing risk. If a payment comes in even a day late, the borrower misses that month’s unpaid-interest waiver and the government principal match, which is a direct reduction on the loan. Only payments made on time count for Public Service Loan Forgiveness. “Being late with a payment, by even just one day, will cost you,” Mark Kantrowitz, a higher-ed analyst, told CNBC. Rich Williams, a former Education Department official, said both protections depend on paying on time. eCFR
About 300,000 first-wave RAP users could get up to $50 in principal match per month, putting the total around $15 million monthly or $180 million a year if everyone qualified for the full amount. That’s just 0.01% of the federal loan portfolio. This doesn’t include the separate interest waiver and isn’t a budget figure. The real principal match will be lower, since it hinges on how much each borrower pays and how much actually goes toward reducing their balance.
Servicers have the most direct exposure among listed firms. Nelnet Inc. NYSE:NNI was handling $525.7 billion across government, private and consumer loans for 15.5 million borrowers as of March 31. Its servicing arm brought in $127.8 million revenue and $15 million net income for Q1. More account switches can mean extra calls and processing, though Nelnet said the latest federal contract pays less per borrower than the old deal. The current migration is more of an operational test than a real earnings driver for now.
Most borrowers are still staying inside the federal system, making a private refinance surge less likely right now. The bigger mover is the new annual borrowing limit set at $20,500 for most grad students and $50,000 for professional students. SoFi Technologies Inc. NASDAQ:SOFI originated $2.6 billion in student loans in Q1, a jump of 119% from last year. SLM Corp. NASDAQ:SLM came in at about $2.9 billion, up 5%. Both these figures are from before the July 1 loan caps. Private lenders are looking to the next academic year to see if the new federal limits start sending more borrowers their way.
The first read could shift. Just 40% of borrowers in active repayment used automatic debit in June. The Education Department’s temporary one-point interest rate cut could bring in more people and keep missed payments down. Early trends might not stick, since first movers may not be typical, and court fights have already unsettled some borrowing and forgiveness rules. Private lenders could see more business, but the new borrowers might have bigger funding gaps, so growth could lose value if credit quality drops.
Markets may be missing the timing on the SAVE shutdown, seeing it as an instant refinancing trigger. The first big moment is actually a test in 2026 to 2028, when borrowers face new servicing and payment pressures and a second migration to different plans stacks up. Only later does the private credit piece come in, and even then, it’s dependent on whether new federal caps produce gaps borrowers can actually fill.
