No Cap: Congress Should Mandate Revolving Debt Off-Ramps, Not Impose Crude APR Rationing

Alexei Alexandrov consults various non-profits on housing, mental healthcare and homelessness, student lending, and works with various tech companies. Prior to that, Alexei worked at the Federal Housing Finance Agency, Amazon, Wayfair, and the Consumer Financial Protection Bureau, and taught at the University of Rochester.

Bipartisan proposals from President Trump, Senators Bernie Sanders, and Elizabeth Warren for a 10% credit card APR cap (price ceiling) address a real issue. The last few years have seen much higher credit card interest rates combined with hundreds of billions of dollars of revolving debt fueled by tens of millions of households barely making ends meet.

But an APR cap would likely hurt the households it intends to help. Instead, Congress should mandate a lower-APR installment off-ramp for borrowers who are revolving a high amount of not fully-amortizing debt for a long time. The specifics of “high amount” and “a long time” are flexible and up for debate. As a starting point for that policymaking, if a consumer is revolving more than $5,000 on a credit card account for over twelve months, the law should prohibit more borrowing on that account until the balance is paid off and convert the balance into a 36-month closed-end installment loan, with a maximum APR of 10%+Prime rate.

What Market Failure Are We Trying to Fix? 

Hint, it’s not lack of competition, over-abundant advertising, or junk fees. The credit card market is not perfectly competitive, with the three largest issuers at about 15-20% market share each, and the next three at about 8-10% each. But which markets are clearly more competitive, by a typical competition measure like HHI? (Restaurants, hotels, gas stations, and maybe apparel… but not airlines, car manufacturers, cellular service providers, grocery stores at the local level, or health insurers). Credit card issuers also appear to pass-through both their cost of funding (Fed’s interest rates) into interest charges and interchange fees into rewards almost 1-to-1, showing a sign of a reasonably healthy market. Compare and contrast this to, for example, checking and savings account deposit yields that are close to 0% regardless of whether banks are paid 0% for their reserves at the Fed or 5%. If we’re worried about competition, then singling out credit card issuers is a strange place to start.

Credit card issuers spend a lot on marketing, which some argue is a sign of monopoly power. Anecdotally, a healthy portion of this spend appears to attract transactors — card users that do not run balances - with rewards offering. Regardless, credit card issuers are not the only industry that spends big on marketing, and there are examples that arguably deserve much more policymaker attention — pharmaceuticals, alcohol, sugary foods, social networking sites, to list a few. Most bank branches at this point are arguably an advertising tool, and maintaining them is considerably more expensive than the credit card marketing spend. Again, if we’re worried that some advertisement is outright harmful and much of the rest is simply wasted money that could be channeled into lower prices instead (consumer packaged goods? cars?), then singling out credit card issuers is a strange place to start.

“Junk fees” are real. The CARD Act of 2009 fixed at least some of the issues. And over the last decade, late fees stayed at a roughly constant percent of balances. Moreover, the late and other fee revenue (not including annual fees, targeting mostly transactors going for perks) is roughly one-tenth of that of either interest charges or interchange fees from the merchants. So there is still work to be done here, but not through an APR cap that would incentivize issuers to charge more fees, and not as the top tier priority.

Short-Term Credit Products Trying to Serve Longer-Term Borrowing Needs are a Market Failure¹

Credit cards were never meant for long-term borrowing. They started as charge cards, to be paid off in full every month, with American Express in particular keeping the concept alive for decades. However, more recently, “about 12% of prime and 20% of subprime revolvers sustain balances continuously for 2 years or more.”²

Consumers revolving a low-dollar balance or revolving for a short time don’t pay that much interest. For example, the difference in interest costs of borrowing $1,000 for three months at a  17% APR (10%+Prime) vs. a  29% APR (deep subprime APR average in 2024Q4) is about $30.

The problem is consumers who end up borrowing for a long time and a large amount. The difference in interest cost between borrowing $5,000 for three years at a 29% APR vs a 17% APR is $1,800. And many consumers continue making just minimum payments for longer than three years, resulting in payday-like cases of more interest paid than the amount borrowed.

Credit cards were not designed as a product to borrow money for a long time, but many consumers end up using credit cards for this purpose. We have other products for that — personal loans (or collateralized options like home equity loans or lines of credit).

But personal loans are a much smaller market, with considerably less consumer awareness. And personal loans have three other weaknesses. First, personal loans are subject to more strict ability-to-repay requirements due to their closed-end nature and not having to pay simply minimum payments. Accordingly, some form of income verification is typically required, and a person who’s behind due to job loss or income irregularity might have problems qualifying. Second, it’s likely that many consumers who are getting into a sustained large credit card balance didn’t realize that they’ll be revolving this much for this long. Third, relatedly, credit card issuers have the benefit of having a sort of an insurance pool — the math works out, with borrowers not borrowing nearly as much as expected or paying off quickly averaging out borrowers borrowing more than expected and for longer than expected. Meanwhile, a borrower who realizes too late that they should have had a personal loan instead is already adversely selected — it’s clear to the issuers that they need more credit than it seemed and they might have a lower credit score by now. This is kind of like trying to get someone insured after it’s clear that they have a health concern.

Off-Ramp to Debt Repayment Fixes This

The fix is to build an off-ramp for consumers who end up revolving a lot and for a long time — for example, over $5,000 for more than one year. About 15% of accounts made only minimum payments on general purpose credit cards in 2024; and about 13% of accounts are in “persistent debt.” The $5,000 balance is at about the 90th percentile of credit card accounts. Conservatively, let’s assume that about half of the persistent debt accounts have more than $5,000 in balances — about five times more than credit card users in general. Then, about 7% of accounts would fall into the mandatory off-ramp.

The off-ramp could be structured exactly as a personal loan that the consumer, in retrospect, should’ve taken out instead, with a similar interest rate to what the consumer likely would have gotten at the time. In particular, a three-year, closed-end, same monthly payment amortizing loan, at a more reasonable APR of at most 10%+Prime (below average personal loan APRs).

Just like a personal loan, there are no top ups, at least through this credit card account – the consumer cannot take out more debt on this credit card until the loan is paid off. Effectively, the balance after one year converts into a personal loan, and there is no remaining credit line until the loan is paid off.

The off-ramp should be structured as a not-great outcome either for the borrower or for the credit card issuer, it’s not a prize for either. We also don’t want to penalize the credit card issuer too harshly, so that there are minimal effects on who gets qualified for credit cards in the first place. The 10%+Prime is not far from the interest that the consumer could have gotten on personal loans before getting too much revolving debt — so this off-ramp is worse for the issuers than what they get now, but not wildly unprofitable even for the 7% or so of borrowers who ever get there (likely fewer once credit card issuers start adjusting underwriting). The off-ramp would be likely to improve repayment for the few consumers in such a situation, improve predictability of repayment streams, and cut off further borrowing on this card. Warning consumers about the off-ramp would be needed, for example on their monthly statement, 6 months, 3 months, and 1 month before the off-ramp starts. These accounts might also get separately securitized, as securitizing closed-end credit had been historically easier than securitizing open-end credit, due to much more predictable payment flows.

Revolving forever while paying minimum payments is not exactly demonstrating ability to repay, it’s just demonstrating the ability to kick the can down the road. We must incentivize card issuers to take ability-to-repay more seriously. Is the consumer going to get into persistent debt? If yes, then give them a personal loan to begin with. If not, sure, give them a credit card, but in the 7% of cases when you’re wrong, do provide a reasonable off-ramp that looks like a personal loan that they should’ve gotten to begin with. Such an off-ramp is similar to servicing requirements mortgage lenders already have to comply with.

Such off-ramp monthly payments aren’t going to be considerably higher than minimum monthly payments. A $5,000 balance at 29% APR requires a minimum payment of about $175. A 10%+Prime APR cap three-year amortizing installment loan is about $180 per month.

The UK’s version of the CFPB has been worried about persistent debt issues for credit cards for about a decade, and implemented guidance addressing many of the points I discuss in this blog, apparently without unraveling the market. In the UK’s guidance, first warnings come 18 months into revolving debt, and real interventions kick in at 36 months, with no explicit dollar (pound) limit on the amount in persistent debt. At the 36-month point, credit card issuers “would need to help the customer by proposing ways of repaying more quickly over a reasonable period, usually between 3 and 4 years. For example, by transferring the balance on the credit card to a lower-interest personal loan,” and the credit card could be suspended especially for the borrowers who do not respond.

One Does Not Simply Impose a 10% APR Cap Without Expecting Less Credit for Subprime Borrowers

If enacted, a 10% APR cap would be a pyrrhic victory. Yes, current borrowers would see a temporary relief to the extent that they are already qualified for credit lines, but tens of millions of potential borrowers will be stranded, especially renters who do not have collateral. Lending to subprime borrowers, for an indefinite amount of time, with a hard-to-predict repayment schedule at 10% is not particularly attractive relative to, say, close-to-7% mortgage-backed securities that are insured by the Federal government from defaulting and are liquid.

The vast majority of superprime consumers are transactors, while the majority of subprime consumers are revolvers — making them easy to differentiate for issuers. About 14% of superprime (800+ score) credit card accounts carry a balance on non-promotional general purpose accounts. The corresponding number for subprime (under 660 score) is 59%.

With a 10% APR cap, superprime rewards cards will still be around, maybe with tighter qualification criteria and lower credit limits. It’s a minor inconvenience for the target audience of such cards, but would rule out subprime revolvers. And an occasional superprime revolver at 10% might not be hard to handle.

For subprime revolvers though, the door will be largely closed to revolve more than a minimal credit limit. The 10% interest rate cap is not enough for credit card issuers to make profit on this market segment. Naturally, this pushes borrowers into some alternatives that at least theoretically could work out well, like BNPL that allow for six or more weeks of interest-free revolving, but also into other alternatives that could be disastrous, like payday or title loans.

Politics and Implementation

This proposal could also serve as an off-ramp to any remaining Congresspeople who want to make it clear that they care about their constituents, but do not necessarily want to jump head first into the shallow pool of destroying subprime credit cards as we know them.

Credit card issuers should also embrace this proposal, and potentially even start implementing it voluntarily. It seems unlikely that the US will turn less populist soon. So the credit card issuers can either set a de facto market standard or wait and hope a bipartisan consensus to cap APRs and have a revitalized CFPB implement it never materializes. Implementing such off-ramps voluntarily immediately cuts off the worst examples of consumers getting caught in revolving debt, likely considerably lowering political heat, and leaving future proposals to marginal issues ($3,000, $5,000, or $10,000? 9 months, 12 months, or 18 months?). 

The largest issuers that also have substantial non credit card revenue — like Chase — might find it easier to step up, eventually forcing their other competitors to step up as well. They should read the room and lead the way.

Finally, versions of this proposal could apply to BNPL, payday, and title loans — all designed as short-term borrowing, but with many consumers ending up permanently revolving; however, implementation details deserve their own long blog. For BNPL, this could be implemented at a firm level (say, Affirm) — and as long as the borrower is paying on time, there is no issue whatsoever because they already have their pay-in-four (or similar) off-ramp that satisfies all the constraints above. But the off-ramp would be needed if the borrower starts falling behind on their payments, incurring fees or high interest charges. For higher-APR and fee products like payday and title loans, the off-ramp might need to come sooner — for example, the typical APR of 36% line could be the threshold for when the off-ramp kicks in sooner.

The opinions shared in this article are the author’s own and do not reflect the views of any organization they are affiliated with.

^[1]  As an aside, another large (if not larger) credit card market failure to fix is the high interchange fees, but I’ll save it for another OpenBanker piece.

^[2]  See, https://onlinelibrary.wiley.com/doi/abs/10.1111/joca.12360, see also https://files.consumerfinance.gov/f/documents/bcfp_data-point_credit-card-revolvers.pdf and https://onlinelibrary.wiley.com/doi/abs/10.1111/jmcb.13023.

Open Banker curates and shares policy perspectives in the evolving landscape of financial services for free.

If an idea matters, you’ll find it here. If you find an idea here, it matters. 

Interested in contributing to Open Banker? Send us an email at [email protected].