By Brian Shearer | December 4, 2025
I have two girls and we’re in the middle of our K-Pop Demon Hunters phase, so apologies, I had no choice on the title. And for those of you who haven’t had to listen to the soundtrack 1,000 times, you can check out this video.
Now, with that caveat aside: Two weeks ago the American Bankers’ Association published a purported take-down piece attacking a report I authored in September showing that the bipartisan proposal to cap credit card rates at 10% would save billions of dollars without much downside. There are sleights of hand in this piece that are pretty typical of the banking industry lobbyists’ analysis on rate caps, which they claim are always a bad idea. It followed the four basic tactics in the corporate lobby playbook:
- Claim business is too complicated for policymakers to understand, with the goal of making them nervous that they might accidentally break something;
- Make it sound like policy proposals supported by long (in this case millenia old) historical precedent are actually radical departures from the norm;
- Cite to unrelated studies or data to make sweeping conclusions (and what do you know, it happens to support their clients’ cash cows); and
- Use generic Econ 101 principles when they simply don’t apply
Instead of just letting this go as more lobbyist noise, I decided to go through it to debunk the banking lobby’s points paragraph-by-paragraph. If we’re really going to do affordability policy in the next decade, we’re going to need to build up antibodies to the standard lobbyist playbook.
In other words, this is a take-down of the take down. Here we go…
The first paragraph starts off with a strawman:
A recent paper from Vanderbilt Policy Accelerator for Political Economy and Regulation prepared by former CFPB official Brian Shearer argues that capping credit card interest rates at 18%, 15%, or even 10% would save consumers tens of billions of dollars a year — without reducing access to credit or cutting rewards. According to the study, banks earn “excess” profits across all risk tiers and could simply trim their marketing budgets to absorb a cap.
I did say banks could trim their massive marketing budgets. But the real reason they can absorb the caps is that the profits are absolutely massive. They would absorb a cap mostly by taking less profit, but trimming the marketing budget would help too (though much less so).
It’s a provocative claim, but the study rests on a false premise, flawed methodology and a misunderstanding of how banking works and how firms behave in competitive markets. In reality, the credit card industry is highly competitive, profitability is far lower and more cyclical than the paper incorrectly assumes, and history shows that interest rate caps harm the very consumers they aim to protect by restricting credit access and eroding benefits.
I’m not sure what history books they are looking at. Usury caps are a longstanding norm and have existed for most of the history of human civilization. In fact, there are caps in place now. Currently, for a variety of reasons that are too complicated to explain but mostly attributable to federal laws pertaining to credit offered to active-duty servicemembers, credit card banks feel they can’t charge more than 36%. And credit unions are already capped under federal law at 18%. Up until about 1978, most states had lower interest rate caps that applied to credit cards, often as low as 10% or 12%. Even now, many states have laws that would prohibit the interest rates credit card companies currently charge—those laws were preempted in 1978 by the Supreme Court, after which the rates on credit card jumped dramatically. So, it’s just a question of where we draw the lines.
At the heart of the Vanderbilt study is a paradoxical belief: the credit card industry is both insufficiently competitive and yet spends “too much” on marketing. Those views are hard to reconcile. In genuinely uncompetitive markets, firms typically don’t compete with advertising to win customers. Robust marketing is itself evidence of market competition.
If the card market is competitive, then what accounts for the study’s finding that profits are excessive and that issuers could simply absorb the impact of a cap without negatively impacting consumers?
Before we get to the question, let’s start with the bizarre point that marketing is proof that markets are competitive. This is just untrue. As one example, the Bell system had a legal monopoly on telephone service for most of the 20th century, and it engaged in advertising and marketing.
Now, to the question: There is an answer to this, and it’s in the paper. There is some competition in the credit card market. But that doesn’t mean they have competed the rates down to low profits as you would expect from reading an Econ 101 textbook. The reason is pretty simple—credit card banks do not compete based on the rates; they compete based on rewards. And the reason for that is it’s impossible for consumers to shop based on interest rates.
Give it a try, look at any credit card shopping website, like this one. Which is the cheapest card for you? You can’t tell because rates are set based on your individual credit score and so rates cannot be advertised. They advertise ranges, but the ranges are so broad that you can’t tell which card is cheapest for you. You learn your interest rate after you apply. NerdWallet doesn’t even bother to show the ranges anymore.
I’m not saying this is some grand conspiracy to avoid competition. There are reasons why banks price cards based on your individual risk level. But what it does mean is banks simply don’t compete based on interest rates. That is why the profits are massive despite the competition that does exist, and that is why we need an interest cap.
The answer lies in the Vanderbilt study’s methodology, particularly the period of analysis and the way in which profitability for each risk tier is calculated.
- Flawed definition. The study equates any positive accounting return with economic viability, but lenders must earn enough to cover not just their cost of funds and the cost of providing the service (which includes losses when clients default on their obligations), but also enough to cover their cost of equity and to maintain the regulatory capital they are required to hold against unsecured credit card loans. Once equity requirements and risk premiums are included, the true “hurdle rate” (that is, the rate at which lending in a particular risk tier is profitable) is much higher than the Vanderbilt study assumes, and many of the FICO tiers that appear “profitable” in the Vanderbilt model would not, in fact, be economically viable under a rate cap.
This is admittedly the best sounding critique. It sounds so smart. But there are hidden assumptions here. First, the “cost of equity” is not an actual cost the bank is paying. Cost of equity is just the rate investors want to earn in returns (dividends and stock value growth) based on the risk-level of the investment. The cost of equity of bonds is lower than bank stocks, which is lower than average stocks, which is lower than risky stocks. Investors believe it is a good investment if the “return on equity,” which is a measure of profit, at least matches the cost of equity.
This is an important analysis in different markets like, for example, utilities, where the companies need to be able to attract equity investors to fund new infrastructure. But we’re talking about banks. Banks don’t fund loans with capital from equity investors. They lend deposits to create capital. This is a really important point. Bank executives might want all of their activities to be profitable enough to exceed investors’ cost of equity estimates, to keep the value of their stock options soaring. But they don’t need to do that in order to maintain business viability or safety and soundness of the bank because they don’t need investor capital to lend. A bank could be perfectly safe and profitable with a dropping stock price.
In addition, because banks are not legally allowed to use deposits to invest in anything other than loans, there is no opportunity cost. It doesn’t matter if credit card returns are lower than bonds or the stock market or anything else. Banks will lend if they can make a marginal profit because the alternative is just sitting on the deposits—a 1% return is better than a 0% return. This is actually what’s happening right now—banks have much more deposits than they have lending demand.
The one thing banks do need to do is cover the cost of capital. Some banks do not lend their own deposits, but rather, borrow deposits from other banks at the Federal Funds Rate (FFR) that the Fed sets. My report accounted for that cost by looking at interest revenue spreads, which is the spread over the FFR.
By the way, nonbank lending is different. Nonbank lenders do need high enough returns to attract private capital because they don’t have access to deposits. If a nonbank lender cannot show high enough returns to investors to get them to invest, the nonbank will run out of capital to lend. This is why bankers often cite studies about the impact of usury caps on payday lending—usury caps are more likely to impact supply from nonbank lenders than bank lenders. But credit cards are bank loans.
Let me debunk their point in a different way. Banks’ average Return on Assets (ROA) across all activities is consistently around 1.1%. But the ROA for credit cards is 6.24%, reaching as high as 10.82% for subprime customers, but still a high 2.56% for even super-prime customers. Investors are consistently happy with 1.1% returns from banks, so the credit card profits are much higher than they need to be.
Separately, the study calculates return on assets using only outstanding credit card balances. That approach is reasonable for measuring the profitability of the lending function of cards, but it does not properly account for the large population of “transactor” accounts that pay their balances in full each month. These customers aren’t borrowers at all but still generate interchange revenue for the bank. Transactors are a major part of every issuer’s portfolio and are a key market demographic, especially for premium rewards products. The study’s profitability metric effectively excludes these accounts from the industry’s asset base, which mechanically inflates the profitability of high-risk borrowers (who tend to revolve more). This framing overstates the ability of issuers to absorb a rate cap without cutting credit access or benefits for higher-risk borrowers.
This is just wrong. The report is based on the Average Daily Balance, which includes transactor account balances that are regularly paid at the end of the month.
- Cycle timing bias. The study examines accounts opened from 2015 to 2017 and tracks them for six years, a period characterized by historically low charge-offs and COVID-era paydowns. However, due to the CARD Act’s restrictions on repricing existing balances, issuers must account for the risk of loss over the account’s lifetime when setting interest rates, including during periods of high financial stress. The study’s conclusions are thus based on a historically calm window of performance, thereby exaggerating “lifetime” profitability (especially for high-risk accounts that tend to see charge-off rates rise during cyclical downturns).
First, I picked this period because the Fed released data for that period. It’s the only data I could find. I would love it if someone did a similar analysis based on a longer period, or more recent data.
Second, the argument is that banks must price credit cards so high that they are still able to make profits in the rarest economic crisis. It’s not that they need to build profits to cover losses in the downturns. It’s that they need to set prices now to be able to make profits during the downturns. Here is how I describe this argument in the paper:
“Looking back at the last 40 years, charge-off rates for credit cards have fluctuated fairly minimally, except during the Great Recession where they doubled for a few years. During that recession and again during the COVID-19 outbreak, ROA for credit card banks barely dipped negative for a couple quarters. In the long run, these brief dips did not jeopardize overall profitability at any credit tier. Thus, this need appears overblown, especially because banks are required to maintain a capital conservation buffer to weather brief financial stress.”
In other words, we don’t need to set prices high now to ensure bank profits in future downturns. We require banks to have reserves for these moments. Having said that, as I wrote in the paper, I do think we should grant a regulator authority to raise the cap in a crisis if needed.
- Misunderstanding of marketing costs. To account for credit card marketing costs, the Vanderbilt study assumes that marketing spend accounts for 20 percent of each card issuer’s operating expenses and allocates these costs equally across risk tiers. That is implausible. A large portion of card marketing consists of television ads for premium travel cards and glossy mailers touting lucrative sign-up bonuses — in other words, campaigns aimed at high-FICO, high-spend customers (not subprime revolvers). By allocating marketing costs equally across risk tiers, the study attributes too much advertising expense to low-FICO tiers and too little to the high-FICO tiers. This misallocation exaggerates the industry’s ability to offset the effect of a rate cap on higher-risk accounts simply by trimming advertising.
I’m not really sure why these guys are so triggered by the marketing stuff. It’s fair to say that the marketing estimates were more back-of-envelope because I didn’t actually have data on marketing costs. I just used Capital One’s ratio. But, in my defense, Cap One is a bank that specializes in customers with subprime credit scores. So I don’t think it’s necessarily true that banks spend more on advertising to high-credit-scored customers than low-credit-scored customers. Keep in mind, banks earn 10.82% ROA from subprime customers but only 2.56% from super-prime customers. If you were in the marketing department considering who to mail the glossy mailers to, who would you send them to?
Having said that, the marketing cost is somewhat of a sideshow. The vast majority of the reason that banks can afford the price caps is the absolutely astronomical profits. Banks can find some money in the marketing budgets too, but that’s not the bulk of it.
As illustrated above, the Vanderbilt study’s ROA estimates for each risk tier are significantly overstated, particularly for accounts with low FICO scores. Indeed, according to the Federal Reserve’s most recent annual Report to Congress on the Profitability of Credit Card Banks, the pretax ROA has averaged 4.2 percent over the last 10 years and was just 3.33 percent in 2023, which is roughly half the ROA used in the Vanderbilt study. After taxes, the credit card industry’s returns are likely closer to 2.5 percent — a profitable business, but a far cry from the study’s conclusion that profits are excessive.
Either they don’t know what they are talking about, or this was intentionally misleading. My report was about ROAs from credit cards. The Fed’s reports are about the ROAs for entire banks that specialize in credit cards but also do other things. So the ROAs in the Fed’s reports are lower because, even if a bank specializes in credit cards, they still have less profitable activities like mortgages and auto loans and commercial loans bringing down the total average ROA to 4.2%.
But the question you should ask yourself is, why do banks with significant credit card portfolios have total ROAs of 4.2% when average bank returns are 1.1%? That is almost 4x more profit!
What would actually happen under a rate cap?
Banks and credit unions that issue credit cards are not public utilities. They are businesses in competitive markets. When lending becomes less profitable, the market should expect to see less of it. As such, when the government steps in to restrict the price of credit, the result is that credit availability falls and other fees rise to offset the lost revenue. These changes are typically most acute for those at the margins of the economy — those who, ironically, are the very people the price caps are intended to help.
This is the crux of the issue. The classic story economists tell about price controls is that when government lowers a price below the competitive or profit-maximizing price, it increases demand for the thing but depresses supply because private capital flows to more profitable ventures. Supply shortages are the result. The bank lobbyists use this Econ 101 model to argue that interest caps will always lower supply of credit because it lowers profits.
To be clear, I agree with the model in general. It just doesn’t apply to bank loans because banks do not use private capital to lend. As I described above, they lend out deposits. And banks are not allowed to invest deposits in anything other than loans. That means the “capital”—deposits—has nowhere else to flow to. In fact, there are a lot more lendable deposits in banks than there is demand for loans. That means even a loan that produces just 1% expected return is the most profitable place to invest the deposits. In short, we should not expect to see less lending when lending becomes less profitable. We should only expect to see less lending once lending becomes unprofitable. As my report shows, you can cut the rates pretty deep before you get there.
But we should consider banks public utilities because they get so much help from the government. They are backed by the federal government through FDIC insurance. They get dividends from the Federal Reserve. They get access to Federal Reserve bank accounts with higher savings rates than any bank offers the rest of us. Banks put your deposits in their own Fed bank account, which pays more interest than they pay you and is a form of constant government subsidization of the banks. Then they get periodically bailed out by the Federal government, even if their crisis is one of their own making. And they have special national charters that exempt them from state laws. They even create our money. These are not just businesses in a competitive market. If banks aren’t public utilities, I’m not sure what is.
Previous efforts to impose price caps on lenders bear this out:
- After the 2009 CARD Act limited repricing tools, consumer credit card pricing and availability shifted markedly relative to small business cards, which were not covered by the law. Purchase APRs on consumer cards increased significantly, annual fees became more common and rose in size, credit lines fell and originations tightened, especially for higher-risk borrowers. For additional information about these effects, see ABA’s 2023 letter to the CFPB.
The research on the CARD Act shows that it lowered overall costs of credit cards. Prices have more recently increased, but that was well after the CARD Act’s implementation period.
This study argues that the CARD Act’s limits on various penalty fees undermined lenders’ ability to manage risk, and as a result, credit card access contracted. First, this is not about interest rate caps, but risk-management fees. Interest rates are higher for higher-risk customers, but this study is about fees that deter risky behavior after an account is opened, which interest rates do not do. Second, there is no reason to believe the reduction in lending was due to the CARD Act, rather than a response to the Great Recession. The fact that annual subprime account openings for general purpose cards increased from 11 million in 2013 to 26 million in 2022 suggests that the CARD Act isn’t preventing opening subprime accounts and probably never did. For further evidence that it was the recession, not the CARD Act, that reduced credit lines, look at the fact that total credit lines decreased substantially between 2008 and 2010, but then stabilized, even though the CARD Act didn’t go into effect until February 2010. If anything, the reduction in access slowed and stabilized right after the CARD Act. Controlling for inflation, available credit lines now are about the same as they were pre-Great Recession and pre-CARD Act.
- After debit interchange was capped via the 2010 Durbin Amendment, consumers experienced a sharp decline in the availability of debit card rewards programs and free checking accounts, as well as higher minimum balance requirements and fees. These effects disproportionately affected lower-income consumers, who struggled to meet the more stringent minimum balance requirements and were less able to afford higher fees.
I don’t know why we are talking about interchange fees on deposit accounts. That’s not a credit card. It isn’t even a loan. The fact that this is exhibit B is telling. It shows how weak their argument is.
The debates on the Durbin Amendment are also more mixed than this summary suggests. One of the studies these arguments often cite was recently revised to show that less than 50% of the lost interchange revenue was recouped elsewhere. But more importantly, the profit margins of debit cards pre-Durbin weren’t even close to current margins on credit cards.
- In Chile, a 2013 rate cap resulted in more than 80 percent of consumers ending up worse off, including 200,000 families that were cut out of the credit market entirely. As typically happens when price caps are instituted, poor and less-educated families bore the brunt of the burden.
Exhibit C is . . . Chile? You should take this as a sign that the arguments against caps are very thin. Presumably, if the case was so ironclad, they’d have more than this.
The study says that lending volumes went down when Chile implemented a deep rate cap. It acknowledges that one of the reasons was a reduction in lending applications, which makes sense. One of the things a rate cap can do is put more money back in people’s pockets, so they don’t have to borrow more. That kind of reduction in lending is unquestionably good.
To be clear, the report also suggests there were customers who lost access who wanted credit. But, importantly, what it doesn’t do is say whether the cap made certain customers unprofitable, or just less profitable, which again, is the real debate here.
These examples show the same pattern: when a major revenue source is constrained, the firm will adjust its credit standards, fees, and product features (including rewards). Contrary to the Vanderbilt study’s assertions, lenders cannot simply absorb the impact of a rate cap. A rate cap will reduce interest revenue, reduce credit access among higher-risk consumers and reduce credit card spending. This will result in lower interchange revenues and fewer cardholder rewards.
The impulse to cap credit card APRs reflects understandable concerns about household debt burdens. Consumers struggling with credit card debt should inquire with their card issuers about debt refinancing opportunities and personalized payment plans. As both economic theory and a century of experience demonstrate, price caps are not the answer. The policy prescriptions recommended by the Vanderbilt paper would make lending to high-risk populations unviable and ultimately cause more harm than good.
Lots of hyperbole here. Apparently, a cap on junk fees for credit cards, a cap on deposit account fees, and a study from Chile, all from the last 20 years, is “a century of experience” relevant to interest rates in the U.S. credit card industry. We don’t have to stretch that far for relevant evidence here—in the United States, right now, credit unions have to follow an 18% interest cap, and they still offer cards to subprime customers. According to the Bank Policy Institute, 41% of Navy Federal Credit Union’s cardholders are subprime, which is a high ratio. Of course, 18% is not 10%, but the bankers aren’t arguing that 18% is more reasonable than 10%. They are arguing that interest caps always lead to lost access by subprime customers no matter where they are set, and the experience of the 7th largest subprime credit card lender in the US suggests otherwise.
The bottom line is this. Credit card lending is bank lending, not nonbank lending. Because banks lend out deposits, and a lot of deposits are just sitting in banks making 0% returns right now, banks will continue to lend out deposits at even a very minimal profit. Right now, credit cards make massive profits in every risk-tier because the rates are extremely high, and banks also make billions from interchange fees. That is why we can make deep cuts in an interest rate cap, without reducing access to subprime customers.
###
Brian Shearer is the Director of Competition and Regulatory Policy at the Vanderbilt Policy Accelerator. This blog was also published at In Debt, a Protect Borrowers Substack, and at Vanderbilt Policy Accelerator’s Substack.