The Problem: Subprime Car Loans.

This essay was inspired by an article in Bloomberg titled: Subprime Auto Loans Make Wall Street Rich as Borrowers Struggle By Ann Choi, Paige Smith and Rachel Dottle on September 28th 2026.

The 21 Percent Problem: Subprime Auto Lending in America

The 21 Percent Problem: Subprime Auto Lending in America

How to read this essay. Every linked figure comes from the page linked at that point. Where a fact comes from a source I could not open, or from my own calculation, the text says so. Court and enforcement records, other lending cases, and videos are collected in the source materials, and the companies involved are listed in the addendum. Nothing here is legal or financial advice.

1. The Problem

The problem is that in America the people least able to afford interest are charged the most of it, on the one purchase most of them cannot do without. A borrower with a deep-subprime credit score who buys a used car pays, on average, 21.62 percent a year, while a borrower with excellent credit pays 6.29 percent for the same kind of car (Experian, second quarter of 2026). On a $20,000 loan repaid over sixty months, the difference between 21 percent and 12 percent is roughly $541 a month against $445, and about $12,460 in total interest against about $6,690, a gap of more than $5,700 that comes out of the household budget of someone already living close to the edge (my calculation, ignoring fees).

None of this is hidden, and much of it is lawful, which is part of what makes it hard to name. A lender takes a risk on a borrower whom banks have refused, charges a rate that compensates for that risk, bundles the loan with thousands of others, and sells bonds to investors who are paid a fraction of what the borrower is charged. The borrower, meanwhile, learns that missing a few payments can cost the car, the job that depended on the car, and, after the car is sold at auction, a remaining debt that follows the household for years. This essay asks how large the problem is, who profits from it, who is caught in it, whether anyone has been held to account, and what a country that took the matter seriously would do.

2. The Scale of the Problem

Roughly one American auto account in six is subprime, which by my arithmetic means somewhere between 16 and 19 million accounts and a quarter of a trillion dollars or more of debt. The New York Fed counted $1.713 trillion of auto debt outstanding in the second quarter of 2026, spread across about 108 million open loan and lease accounts, and the Philadelphia Fed puts the subprime share of active accounts at 17 percent. Applying that share to that count gives about 18.4 million accounts, which I present as a range because accounts are not people, since some households hold several, and because the subprime share of balances may differ from its share of accounts. On the same reasoning subprime balances come to roughly $257 billion to $291 billion. These totals are my calculations from published shares, not published counts.

The flow of new lending shows no sign of shrinking. Experian reports that 15.75 percent of vehicles financed in the first quarter of 2026 went to subprime borrowers, up from 14.40 percent a year earlier, and among used cars the share was 20.60 percent. Subprime borrowers hold about 17 percent of accounts yet account for nearly two-thirds of delinquent loans, and Fitch measured 6.74 percent of subprime auto loans at least 60 days delinquent in December 2025, the highest in the thirty-two years of its data. By February 2026 the figure had reached 6.80 percent, against 0.42 percent for prime loans (Wolf Street).

The Philadelphia Fed's authors add a caution that deserves to be repeated: the headline delinquency rate probably overstates how many borrowers are newly in trouble, because delinquent loans now linger longer before charge-off or repossession, partly through the extensions that lenders offer. That is a reason to read the delinquency figures carefully, though it is not a reason to read them as good news, since an extended loan keeps accruing interest on a borrower who cannot pay it down. A national total of repossessions for 2025 is one figure I decline to give, because the sources I found conflate repossession assignments with completed repossessions.

3. The Companies Creating the Problem

Four companies named in the Bloomberg coverage of September 28, 2026 (a page I found in search but could not open) together hold an estimated 2.2 to 3.1 million subprime accounts and earn, by my estimate, between $1.1 billion and $2.6 billion a year in pre-tax profit from them, which is a small share of the whole but a useful window onto how the business is run. The ranges in the table below come from lender filings and rating-agency reports that I gathered earlier in this work and did not re-open for this essay, and they are estimates built from partial disclosure rather than reported figures.

LenderModelEstimated subprime accountsEstimated annual subprime pre-tax profitRecords
Santander ConsumerBank-owned lender, largest of the four in subprime1.0 to 1.3 million$0.4 billion to $0.9 billioncourt records
Exeter FinanceIndependent lender that sells loans to bond investors500,000 to 760,000$0.1 billion to $0.5 billioncourt records
GM FinancialCaptive lender of General Motors390,000 to 550,000$0.2 billion to $0.5 billioncourt records
CarvanaRetailer that originates loans and sells most of them300,000 to 470,000$0.4 billion to $0.7 billioncourt records

The more instructive fact is how the model works rather than who runs it. A dealer arranges most loans and, as Consumer Reports found, may add a point or two to the rate without telling the buyer what the lender offered. The lender then packages the loans into bonds that pay investors roughly 6 to 7 percent, while the borrowers pay 19 to 22 percent, and the difference covers losses, servicing fees, and profit. When a borrower falls behind, the lender chooses between repossession and modification, and by my earlier reading of the filings the four differ sharply, with Exeter modifying about two-thirds of troubled loans and Santander roughly one in five.

These four are not the whole industry, and they are not the worst by every measure. Credit Acceptance, which settled with 41 states this month, and the buy-here-pay-here dealers charge more still, and a Federal Reserve analysis found a weighted average rate of 25.39 percent on those dealers' subprime loans. In September 2025 Tricolor Holdings, a subprime lender and used-car retailer in the Southwest, collapsed into bankruptcy. Blame is easy to assign and harder to apportion, and it is only fair to say that these companies have generally denied wrongdoing, that their settlements were reached without admission of liability, and that lenders can argue with some force that borrowers with damaged credit default more often and that a high rate must cover the resulting losses. The addendum lists the wider field of lenders, banks, credit unions, captives, underwriters and vendors.

4. The People Caught in the Web of the Problem

The borrowers are, for the most part, households for whom a car is the largest asset they own and a loan on it the largest debt they carry. The Chicago Fed notes that a vehicle is the biggest asset for about a quarter of the lowest-wealth households and that auto loans are the largest share of debt for about thirty percent of them, which means a repossession takes away not only transportation but the better part of a family's balance sheet. They score between 300 and 600 on the credit scales that lenders use, and in the second quarter of 2026 they were charged 19.10 percent on average at the subprime tier and 21.62 percent at the deep-subprime tier for a used car (Experian).

Their own accounts, as recorded in interviews by the Center for Responsible Lending, follow a familiar pattern: most were told they qualified for only one loan, most did not fully understand what they signed, some found add-on products on the contract that had been misdescribed, and several described being repossessed without warning, one saying, “We woke up to them taking our car.” Hardship began early, and borrowers took second jobs to keep up the payments. When the car was lost, so in some cases was the job that depended on it. Consumer Reports found that nearly a quarter of the borrowers in its sample of some 858,000 loans paid more than ten percent of their income on the car alone, and that subprime borrowers faced steeper ratios still.

The most extreme cases come from the enforcement record rather than from surveys. New York's attorney general alleged in September 2026 that Credit Acceptance pushed borrowers into loans at interest rates averaging 38 percent, some above 100 percent, with repossession rates near 50 percent (New York Attorney General). Those are allegations settled without an admission, and they concern one company, so they should not be read as typical of the industry.

Service members are a special case that is easy to misread. The CFPB's analysis of some 203,000 service member loans made between 2018 and 2022 found that they were about one percent of originations, paid modestly more than civilians (about 0.3 percentage points more after adjusting for credit score), and were repossessed about half as often (CFPB). The Military Lending Act's 36 percent cap does not apply to a loan taken to buy the car that secures it (CFPB), and a 2023 Fourth Circuit decision, Davidson v. United Auto Credit, held that bundling items such as GAP insurance into the loan does not change that. A March 2026 Senate Banking letter to Ally Financial argues that service members face special vulnerabilities, but I found no reliable count of how many service members hold subprime auto loans, and I would not put one in print.

5. Why Are These People in the Problem? Why Not Cheaper Options?

Most borrowers in this position are not being reckless; they are buying the only transportation that lets them keep a job, in a market where the cheap alternative has largely disappeared and the buyer seldom sees a competing offer. Over ninety percent of American commuters depend on a vehicle, according to the Chicago Fed, which also cites a subsidized-vehicle program in which three quarters of participants who had lacked a car found better jobs or raised their earnings. For a household that has to reach a job, a school, and a clinic, a car is closer to a precondition than a luxury.

The cheap used car, moreover, is vanishing. In the second quarter of 2026 only 17.8 percent of used cars sold for under $15,000, down from 31.6 percent in 2019, and a budget of $10,000 to $15,000 now buys a car that averages nearly nine years old with about 98,000 miles, against five years and 58,000 miles in 2019 (Edmunds). A buyer with little cash cannot absorb an expensive breakdown, and so may sensibly prefer a newer car with a warranty even at a punishing rate, since a failed cheap car threatens the job that the car exists to protect.

The buyer, meanwhile, rarely knows that a better offer exists. Because dealers arrange the great majority of loans and need not disclose the lender's original rate, the customer is typically given one offer and judges it by the monthly payment, and as Consumer Reports observes, buyers tend to negotiate the price of the car rather than the terms of the loan. To this must be added the ordinary human tendency to discount the future when the present is urgent: a person who needs a car this week, to start a job on Monday, will accept a payment that is hard to sustain, and the sales process is arranged so that the monthly payment is discussed and the total interest is not.

As for a scooter, a bicycle, or an electric bike costing perhaps $3,000, these are worth promoting and, for short urban commutes, may fully suffice. I found, however, no rigorous study of subprime borrowers substituting them, and most of the cost comparisons that appear in search results come from vendors. The practical limits are easy to state: distance and safety on fast roads, weather, carrying children and groceries, night shifts, rural geography, and, for anyone who drives for a living, the plain fact that the work requires a car. They belong among the remedies as one option, though not as an answer to the whole problem, and it would be a mistake to suppose that borrowers who chose a car had simply failed to think of a bicycle.

6. What Can Be Done About the Problem?

The short answer is that four things would do most of the work, and the last section of this essay sets them out at length. The first is to stop the lending that is designed to fail, by treating a loan that the lender itself predicts will default as an unfair practice and prosecuting it as one. The second is to give every borrower a second offer before the dealer's, together with the dealer's markup in plain figures. The third is to make expensive loans cheap to leave, through refinancing at credit unions or a public lender after a year of on-time payments, and through modifications that reduce principal rather than extend the term. The fourth is to give households a route to a working car that does not pass through a 21 percent loan at all.

The remainder of the essay takes these in a particular order. It first establishes what is unlawful and what is merely harmful, since the remedy differs, then reviews the cases that have been brought and asks why no one has gone to prison, before returning to the scale of the problem, testing whether it is worsening, and closing with solutions.

7. What Is the Crime in the Problem?

The difficulty in answering is that most of what makes the problem painful is not a crime at all, some of it is a civil wrong that the states have begun to pursue, and only a narrow band of it is criminal fraud. I am not a lawyer and none of this is legal advice, but the record I have read supports three distinctions.

The first is conduct that is lawful and harmful. A rate of 21 percent is legal for a dealer-arranged retail loan in most circumstances, and the Military Lending Act's 36 percent cap does not apply to a loan taken to buy the car that secures it. Dealers may lawfully raise the rate a lender offered by a point or two without telling the buyer. Extending the term of a troubled loan, which keeps interest accruing on a borrower who cannot pay down principal, is a routine and lawful choice. A country could decide that these practices are wrong, but as matters stand they are not offences.

The second is conduct that is unlawful in the civil sense, which is where the states' enforcement has concentrated. The allegations in the settlements I reviewed include lending to borrowers whom the lender's own models predicted would default, ignoring falsified income documents supplied by dealers, financing add-on products the customer did not know about, misleading borrowers about payment options, and, in the CFPB's 2024 review, repossessing cars from borrowers who were current or had an approved extension, misapplying payments so that late fees were charged wrongly, and advertising rates when the lowest offered was more than double the advertised rate (CFPB). These are violations of laws against unfair and deceptive practices, and their remedy has been money and changed practices rather than prison.

The third is criminal fraud, which requires a knowing deception, and the deceptions that have been prosecuted are overwhelmingly deceptions of banks and bond investors rather than of borrowers. The distinction matters for what follows, because it explains both who has been charged and who has not.

8. What Cases Have Been Brought Against the Problem?

The cases that I could confirm from primary or official sources are civil actions by state attorneys general, the SEC, and the CFPB, and every one of them ended in a settlement in which, so far as the documents I read show, the company did not admit liability. The source materials collect forty-five court and enforcement records against subprime auto lenders, with courts, docket numbers where I could find them, links to the filings, and, where a judge actually ruled, an account of how the court ruled and why.

YearCompanyBrought byOutcomeAlleged conduct
2018Santander Consumer recordSEC$1.5 million civil penalty; no individuals chargedFailing to calculate its credit loss allowance correctly over at least eight reporting periods
2019Exeter Finance recordMassachusetts and Delaware attorneys general$5.5 million and $0.6 million; no admissionOriginating loans it “knew or should have known were unfair” and mishandling servicing and collections
2020Santander Consumer record34 statesAbout $550 million, chiefly about $433 million of loan forgiveness and $65 million of restitution; no admissionLending to borrowers predicted to default, ignoring falsified income documents, and deceptive servicing
2023–2024Toyota Motor Credit and Wells Fargo recordsCFPBA consent order against Toyota Motor Credit with a $12 million civil penalty and about $46 million of redress (the CFPB's 2024 supervisory summary cites a $60 million penalty; the figures differ across the records, see the source materials); Wells Fargo finedWithholding add-on refunds, credit reporting errors, and illegal fees
2026Credit Acceptance records41 statesAbout $700 million: over $630 million of debt eliminated for more than 55,000 consumers, $60 million restitution, $15.5 million in penalties; no admission statedUnaffordable loans at average rates of 38 percent, concealed costs, unwanted add-ons, and misrepresentation to investors

The 2026 settlement is the most instructive, because its remedy concedes the practical point that some loans were never meant to be repaid: the company must forgive 95 percent of the debt when a high-risk loan defaults within 12 to 18 months, stop suing defaulted borrowers, and disclose the risk before the sale (Minnesota Attorney General; the California consent judgment is the only filed instrument I could open). Headlines give the total as between $694 million and $710 million depending on how debt relief is counted. I read state press releases and one consent judgment rather than every state's filed judgment, so the final terms may differ from what I describe. Notably, I found no judge's ruling on the merits in the federal case that the CFPB and New York brought against Credit Acceptance in 2023; the CFPB withdrew in 2025 (see the docket entry).

9. Did Anyone Go to Prison for Causing the Problem?

Not, so far as I could find, for the lending itself: I found no case in which an executive of a subprime auto lender was imprisoned for the terms, pricing, or servicing of loans made to borrowers, though a search that finds nothing is not proof that nothing exists. What I did find are prosecutions in which the victims were banks and investors.

James Collins, the former chief executive of Honor Finance, a subprime auto lender in Evanston, Illinois, was sentenced in March 2025 to four years in federal prison for mail and bank fraud (the press release gives an internally inconsistent date, so I state only the month). He gave a bank false information about a $200 million credit line and the securitized loan trusts behind it, and he misappropriated about $5.3 million through a shell company; the losses were about $67 million, most of them the bank's. The release does not state the judge's reasons for the sentence. See the record.

In December 2025 federal prosecutors in New York unsealed an indictment against the leaders of Tricolor Holdings, a subprime auto lender and used-car retailer that collapsed into bankruptcy in September 2025 (ABC News). The founder and chief executive, Daniel Chu, faces an expanded eight-count indictment that includes a continuing financial crimes enterprise charge (National Law Review), and his lawyer says he is innocent. According to a June 2026 report, the former chief operating officer pleaded guilty on June 25, 2026 and agreed to cooperate, and a trade report puts the trial date at October 19. The alleged scheme was to pledge the same collateral to several lenders at once and to disguise delinquent loans as current, and the lenders that lost money included JPMorgan Chase, Fifth Third, and Barclays. Sources differ on which other former executives pleaded guilty and on the names in the case, and the SEC filed a parallel civil action in August 2026. These are allegations against Chu, who is presumed innocent until proven guilty. See the record.

The pattern is clear even in this small sample. Executives have gone to prison, or may, for deceiving the institutions that fund subprime lending, and I found none charged for what the loans did to the people who signed them. For comparison, executives of payday lenders have gone to prison for consumer-facing lending schemes (Scott Tucker, sentenced to 200 months, and Charles Hallinan, to 168 months); those cases are in the source materials for review, though they involved rates of several hundred percent and the evasion of state usury law through sham tribal arrangements.

10. If Not, Why Not?

The reasons are, in my analysis and not a finding of any court, partly legal and partly a matter of who has power. Most of the harmful conduct is lawful, so there is nothing to prosecute: a high rate, a dealer markup, and an extended term are not offences, and the Military Lending Act does not reach the purchase loan. Where the conduct is unlawful, it is unlawful under statutes that prohibit unfair and deceptive practices, whose remedies are civil, and the states that have enforced them have preferred settlements that deliver relief to borrowers quickly to trials that might deliver a verdict years later. A settlement of $630 million of debt forgiveness helps tens of thousands of people now, and it is not surprising that attorneys general took it, but it also means that no court has found facts and no individual has been named.

Criminal fraud, moreover, requires proof that a particular person knowingly made a false statement to a victim, and the statements that are hardest to defend are the ones made to investors. The New York attorney general alleged that Credit Acceptance misrepresented loan compliance when selling securities, and yet the allegation was resolved civilly, while the prosecutions that reached prison or indictment were built on lies to banks with large legal departments and on documents that could be checked against bank records. The victims of the lending itself are dispersed, poor, and harmed one loan at a time, which is the least favourable posture for a prosecutor with limited resources, and the SEC's 2018 action against Santander Consumer named the company and no individual.

There is also a difficulty of design. A lender that sets a rate high enough to cover its expected losses, discloses that rate on a contract the borrower signs, and profits when the loan fails is doing something that the law was not written to catch, and the Credit Acceptance settlement shows that the states are beginning to treat a loan predicted to fail as an unfair practice in itself. Whether prosecutors will go one step further, and treat an executive who knowingly builds a business on such loans and misstates their performance to investors as a criminal, is a question that only a case can answer, and I found none brought.

11. A Reminder of the Scale of the Problem

Sixteen to nineteen million accounts is a number that slides past the mind, so it helps to put it beside things that can be pictured. My central estimate of 18.4 million is nearly the whole population of New York State, which the Census Bureau put at 20.0 million in 2025. It is more than the 17.1 million people who live in the fourteen least populous states combined, from Wyoming, Vermont, and Alaska through Idaho and Nebraska, and it comes close to the 19.2 million who live in the fifteen least populous once New Mexico is added. Those fifteen states send thirty senators to Washington (population figures from Census estimates for July 1, 2025).

A second image is the cars themselves. If each of 18.4 million vehicles is taken to be fifteen feet long, an assumption of mine, they would form a line of about 52,000 miles, which is a little more than twice the distance around the Earth at the equator. Along that line, going by the delinquency rate Fitch reported for subprime auto bonds, about one car in fifteen would belong to a borrower at least sixty days behind.

The caution that attaches to the count applies here too. These are accounts, not people, and the estimate is a range whose low end of 16 million would still exceed the 15.1 million people in the thirteen least populous states. What the analogy is meant to show is that this is not a niche of the credit market but a population the size of a large state.

12. Is the Problem Getting Worse? The Evidence from 2022 to 2026

The share of auto balances newly seriously delinquent has risen every year since 2022 Line chart of the annualized flow into 90-plus-day auto delinquency: 1.81 percent in Q2 2022, 2.41 in Q2 2023, 2.88 in Q2 2024, 2.99 in Q3 2025 and 3.00 in Q2 2026, with total auto debt of $1.50 trillion rising to $1.71 trillion. The share of auto balances newly seriously delinquent has risen at every reading since 2022 1.81% to 3.00% of balances newly 90+ days late (up 66%), while total auto debt grew 14% Q2 2022: 1.81% newly 90+ days delinquent; auto debt $1.50TQ2 2023: 2.41%; auto debt $1.58TQ2 2024: 2.88%; auto debt $1.63TQ3 2025: 2.99%; auto debt $1.66TQ2 2026: 3.00%; auto debt $1.71T 1.81%2.41%2.88%2.99%3.00% Q2 2022Q2 2023Q2 2024Q3 2025Q2 2026 debt $1.50Tdebt $1.58Tdebt $1.63Tdebt $1.66Tdebt $1.71T Annualized flow into 90+ days delinquency, all auto borrowers; total auto debt shown beneath each quarter.
Sources: New York Fed Household Debt and Credit releases for Q2 2022, Q2 2023, Q2 2024, Q3 2025 and Q2 2026. Q3 2025 stands in for Q2 2025, whose release I did not open.

On the measure that the chart shows, the answer is yes, though it comes with two qualifications that a careful reader should have. The rate at which auto balances newly become seriously delinquent has risen at every reading I could open, and it has flattened only in the last, from 2.99 to 3.00 percent, so it would be premature to say that the deterioration has stopped. The same direction appears in other measures: Fitch's rate of subprime auto loans 60 days or more delinquent was 6.74 percent in December 2025, up 59 basis points on the year before and the highest in its records, which implies about 6.15 percent in December 2024, and Experian's subprime share of new financing rose from 14.40 percent to 15.75 percent between the first quarters of 2025 and 2026. Balances have grown too, from $1.50 trillion to $1.71 trillion.

The first qualification is that the starting point was unusually favourable. Repossessions fell from nearly 1.7 million in 2019 to about 1.1 million in 2021 as lenders offered pandemic accommodations, and 2022 was still a year of recovery from that leniency (ConsumerAffairs, citing Cox Automotive), so part of the rise is a return toward normal. The second is the Philadelphia Fed's warning that delinquent loans now linger longer before charge-off, which inflates delinquency rates without necessarily meaning that more people are newly in trouble. Neither qualification, in my reading, changes the direction of the evidence, though they should temper any claim about how fast the problem is growing, and the New York Fed measure covers all auto borrowers rather than the subprime segment alone.

13. Solutions for a Country That Works for Its People

The measures below are chosen for what would work rather than for what could pass a divided Congress, and where another country has tried something I say so, and where nobody has I say that the idea is untested. They fall into four groups: stopping the harm at its source, making credit cheaper, providing cars outside the loan market, and making the industry accountable and visible.

Stop the harm at its source.

  1. Make loans designed to fail a crime. A federal statute could make it a felony for a lender's officers knowingly to originate or securitize loans that the lender's own models predict will default, or to misstate loan performance to investors, and could require an officer's signed certification of the loan-quality claims in every auto bond. The Santander and Credit Acceptance settlements allege exactly this conduct civilly, and no court has yet tested it as a crime.
  2. Require an ability-to-repay test and cap payments. Consumer Reports found that lenders verified income on only 4 percent of loans, and nearly a quarter of borrowers paid more than a tenth of their income on the car. A hard limit on payment-to-income, checked against verified income, is, as I understand it, how home mortgages have been regulated since 2014, and there is no reason a car that depreciates faster than a house deserves looser rules.
  3. Set a ceiling on the rate. Canada made any rate above 35 percent APR a criminal rate on January 1, 2025 (BLG), with exemptions for some commercial and payday lending. I would set a hard ceiling at the 36 percent that Congress already gives service members, with a rebuttable presumption of unfairness above 18 percent, so that the lender must show why a higher rate is warranted.
  4. End the hidden dealer markup. Britain's regulator has confirmed a redress scheme covering 12.1 million car finance agreements, at an estimated cost of £9.1 billion and an average payout of £829, largely over discretionary commission arrangements that let brokers adjust the interest rate (FCA). That is the same markup American buyers cannot see. A normal country would pay dealers a flat fee that does not vary with the rate, and would require the lender's original offer to be printed on the contract.
  5. Settle the debt when the car is taken. When a lender repossesses and sells a car, the borrower should not then owe the shortfall. The Credit Acceptance settlement already requires the company to stop suing high-risk defaulters and to forgive 95 percent of certain balances, and generalizing that rule would force lenders to bear the losses they price into the rate, which is what the rate is supposedly for.
  6. Require principal-reducing modifications, automatic add-on refunds, and no remote disabling of cars. An extension that keeps interest accruing is a way of keeping a failing loan on the books, and a modification that cuts principal after two missed payments would cost lenders something and change their incentives. Unused add-ons should be refunded automatically when a loan ends early, and the kill switches that the Center for Responsible Lending's interviewees described should be banned.

Make credit cheaper.

  1. Offer a second-chance refinance after twelve on-time payments, through credit unions or a public lender, supported by a federal loss-reserve fund so that lenders will take these borrowers. Credit unions already run refinance campaigns, and Navy Federal reports that its refinancing members cut their monthly payment by $74 on average (America's Credit Unions). On a $20,000 loan over sixty months, moving from 21 percent to 12 percent saves about $5,770 in interest.
  2. Create a public auto lender, through the postal system, state banks, or a federal charter, that borrows at government rates and prices loans at funding cost plus honestly measured losses. No such lender exists for cars in the United States, so this idea is untested here, and its central question is how much of today's spread between the 19 to 22 percent that borrowers pay and the roughly 6.74 percent that bond investors receive is real loss. That question could be answered if lenders were required to publish their loan-level performance, which is the last measure below.

Provide cars outside the loan market.

  1. Scale up the vehicle programs that work. Vehicles for Change has placed more than 7,000 inspected donated cars with low-income households since 1999, selling them for about $850 to $950 on a 12-month loan of $80 to $90 a month with a six-month warranty (Consumer Reports; Klein, 2024). The evidence for its effect on employment is a small interview study, so a national program should be built with a control group from the start.
  2. Pilot an earn-a-car scheme. A city or state that holds surplus fleet vehicles could let a household earn a paid-off used car through weekend service such as planting trees, with the value of the labor and the car fixed in advance. I found no tested precedent, so it should be treated as an experiment, designed so that the work requirement does not exclude people with disabilities, caregiving duties, or night shifts.
  3. Fund the alternatives that do work for some trips, such as e-bike purchase subsidies for short urban commutes and better transit, while recognizing that for many borrowers, in rural areas and in gig work, they cannot replace the car.

Make the industry accountable and visible.

  1. Publish loan-level performance and lender scorecards, so that the public, and not only the lenders, can see rates, defaults, modifications, and repossessions by lender, and so that the question of how much of the spread is loss can be answered from data rather than argued from advocacy.
  2. Fund enforcement and use it. State attorneys general have recovered roughly $700 million in a single settlement this month, and the CFPB's supervisory findings show where else to look. Adequate budgets for both, and a policy of naming responsible individuals rather than settling with the company alone, would change what lenders expect the consequences of these practices to be.
  3. Give buyers three days to change their minds. As I understand it, the federal cooling-off rule generally does not cover vehicles bought at a dealership, and a short right to cancel a financed used-car purchase would give borrowers time to obtain the second offer that education alone cannot supply. This is my recommendation, and I have not tested it against evidence.

If I had to choose only three, I would choose settling the debt when the car is taken, the ability-to-repay test with a rate ceiling, and criminal accountability for loans designed to fail, since together they change what lenders gain from a loan that fails, which is the root of the problem. That ordering is my judgment and not a finding, and reasonable people who accept the diagnosis could rank the measures differently.

Sources cited in this essay

For court and enforcement records, other predatory-lending cases, and videos, see the source materials. For the wider list of lenders, see the addendum.

Limits. The lender-level account and profit ranges in section 3 come from filings I gathered earlier and did not re-open; the national account and balance totals are my own calculations; the legal analysis is not legal advice; I could not open the Bloomberg article and read state press releases and one consent judgment rather than every filed judgment for the Credit Acceptance terms. Summaries of court records were produced from fetched pages, so check any holding against the linked opinion before quoting it.

Source materials · Addendum: who lends


There are other solutions possible. America could make cheap, tiny but high quality and safe cars for 8K. It may still not solve the problem of a family of 4, but it would solve it for a significant amount of the 20M people. Alternatively, we could allow for those cheap cars to be imported. I realize that imports strengthen our adversaries.

Imagine asking a wage earner or their spouse to work on their weekends planting trees, that pays off the car and provides a small stipend for food. It would be a win for the environment, a “free” car for the family and also would pay for increasing grocery costs.

Those are the types of solutions a rich and wealthy country like America ought to do, instead of letting predatory lenders create a negative spiral of depression economics.


Source Materials: Court Records, Other Cases and Videos

Source materials

How reliable is this page? Each link is tagged. “Opened” means the page was fetched and read during research. “Search result only” means the address appeared in search results but was not opened, so treat it as a lead. The case summaries were written from summaries of fetched pages, not from a full reading of each opinion, so the holdings and the reasoning must be checked against the linked judgment or order before anyone quotes them. Where a case ended in a settlement or consent order there was no judge’s reasoning, and the entry says so plainly. Missing docket numbers mean I could not find them, not that none exist.

Contents: sources cited in the essay · 45 court and enforcement records against subprime auto lenders · 48 other predatory-lending cases · videos, advertisements and films · what is missing

Sources cited in the essay

These are the pages behind the figures in the essay, in the order they were first used.

Court and enforcement records against subprime auto lenders

Each record shows the court, docket number and judge where I found them, and then how the case ended and why. The categories run from criminal prosecutions through state and federal civil enforcement, regulatory orders, class actions and appellate opinions. The most useful thing to know before reading them is that the great majority ended by consent or settlement, so there is no adjudicated finding that any lender broke the law. The reasoned judicial rulings are in the appellate opinions, the class-action rulings and the criminal cases, and those are flagged in each summary.

Santander Consumer USA

Multistate settlement (33 states plus DC per NY AG; press coverage says 34) with Santander Consumer USA Inc.

Court
State courts / AG settlements (per-state consent judgments); court and docket for each not confirmed
Docket
not located in this research
Date
2020-05-19
Brought by
State attorneys general (Illinois and Virginia among leads; NY, CA, TN, GA, MI, NH, DC participating)
Outcome
About $550M in relief (potentially up to $780M), including about $433M in deficiency balance waivers, up to $45M in title transfers/balance waivers on lowest-quality loans, plus buyback-related waivers; requires ability-to-pay underwriting and stronger dealer income/expense verification; credit report deletions.
Admission
no

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument). Allegations: used credit scores to identify high-default borrowers and gave them risky loan terms; did not police dealers who inflated income/expenses; deceptive servicing.

In the Matter of Santander Consumer USA Inc. (SEC administrative proceeding, Exchange Act Rel. 34-84829)

Court
SEC administrative proceeding
Docket
34-84829
Date
2018-12-17
Brought by
SEC
Outcome
$1.5M civil penalty. Findings: for at least eight reporting periods (IPO Jan 2014 through second restatement in late 2016) it failed to calculate credit loss allowance for impaired loans per GAAP, and had deficient internal accounting controls.
Admission
no

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument). Settled without admitting or denying findings (findings are the SEC's own, in an order entered on consent).

In the Matter of Santander Consumer USA Inc. (CFPB consent order)

Court
CFPB administrative
Docket
not located in this research
Date
2018-11-20
Brought by
CFPB
Outcome
Restitution of about $9.29M and $2.5M civil penalty over failure to clearly disclose terms of S-GUARD GAP product and loan extensions (extra interest accrues before principal resumes).
Admission
no

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument). No formal admission per CFPB summary.

United States v. Santander Consumer USA Inc. (Servicemembers Civil Relief Act)

Court
U.S. District Court, N.D. Tex.
Docket
not located in this research
Date
2015-02-25
Brought by
U.S. Department of Justice
Outcome
At least $9.35M; 1,112 unlawful repossessions of servicemembers' vehicles (2008-2013); $10,000 plus lost equity for 760 direct repossessions and $5,000 for 352 secondary repossessions; must check DoD database before repossession.
Admission
no

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument). DOJ press release only; complaint and settlement filed together and subject to court approval; I did not locate the entered decree.

DEKA Investment GmbH v. Santander Consumer USA Holdings Inc. (Steck v. Santander Consumer USA Holdings on docket)

Court
U.S. District Court, N.D. Tex., Dallas Division
Docket
3:15-cv-02129-K
Judge
Ed Kinkeade
Date
Settlement notice Sept 2020; fairness hearing Jan 12, 2021
Brought by
Shareholder class (IPO Jan 2014 purchasers)
Outcome
$47M cash settlement of securities claims over IPO statements.
Admission
no

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument). Final approval order not located; I did not find opinions on motion to dismiss or class certification. Settlement reflects no finding of liability.

In re Santander Consumer USA Holdings Inc. Stockholders' Litigation

Court
Delaware Court of Chancery
Docket
2022-0689 (from Justia URL)
Judge
Vice Chancellor Lori W. Will
Date
2025-03-31
Brought by
Stockholders (take-private at $41.50/share)
Outcome
$162.5M settlement approved; incentive award to Elliott set at $500,000 rather than $1.625M requested.
Admission
no

How the court or agency ruled, and why. Merger fairness case, not subprime lending conduct. Court applied the Raider factors (time/expertise spent and benefit to class), found Elliott's work exceptional (1,630 hours; about 6.5% premium recovered) but benchmarked to El Paso ($450,000), citing conflict-of-interest risk in large incentive awards.

Henson v. Santander Consumer USA Inc.

Court
Supreme Court of the United States
Docket
16-349
Judge
Justice Neil Gorsuch (unanimous, 9-0)
Date
2017-06
Brought by
Private borrowers
Outcome
Affirmed 4th Circuit (817 F.3d 131): Santander, collecting defaulted auto loans it bought, is not a 'debt collector' under the FDCPA.
Admission
unknown

How the court or agency ruled, and why. The FDCPA defines debt collectors as those who collect debts 'owed or due another.' Petitioners said 'owed' implies debts formerly owed to someone else. The Court read 'owed' as describing a present condition (like 'burnt toast'), noted Congress distinguished originators from purchasers elsewhere but not here, and refused to rewrite text on speculation about Congress's aims. Result: buy-and-collect debt buyers fall outside this FDCPA definition.

Massachusetts AG v. Santander Consumer USA Holdings (subprime auto loan funding)

Court
Suffolk Superior Court (Massachusetts)
Docket
not located in this research
Date
2017-03-29
Brought by
Massachusetts Attorney General
Outcome
$22M total ($16M relief to 2,000+ consumers, $6M to the state); described as first-in-the-nation subprime auto funding settlement. Alleged it funded loans knowing dealer-reported incomes were inflated.
Admission
no

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument).

Massachusetts AG v. Santander Consumer USA (deficiency notices)

Court
Suffolk Superior Court (Massachusetts)
Docket
not located in this research
Date
2022-02-18
Brought by
Massachusetts Attorney General
Outcome
$5.56M; 1,000+ borrowers; debt relief and credit repair over inadequate explanation of deficiency calculations after repossession.
Admission
no

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument).

Exeter Finance

Massachusetts AG and Exeter Finance LLC (Assurance of Discontinuance)

Court
Suffolk Superior Court (Assurance of Discontinuance filed, per MA AG release)
Docket
not located in this research
Date
2019-04-08
Brought by
Massachusetts Attorney General
Outcome
$5.5M ($4.675M consumer relief; $825,000 to state); deficiency waivers, credit trade-line deletion requests, independent trustee.
Admission
no

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument). Alleged: financing subprime loans without basis to believe borrowers could repay in the normal course; servicing and collection issues.

Delaware DOJ and Exeter Finance LLC (Cease and Desist agreement)

Court
Delaware DOJ administrative agreement (not a court judgment)
Docket
not located in this research
Date
2019-04-08
Brought by
Delaware Attorney General Kathleen Jennings
Outcome
$550,000 relief fund, deficiency waivers, credit report corrections; trustee to distribute restitution.
Admission
no

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument). Alleged Exeter facilitated origination of loans it knew or should have known violated state law.

Rivera v. Exeter Finance Corp.

Court
U.S. Court of Appeals, 10th Circuit
Docket
20-1031
Judge
Tymkovich, Briscoe, Murphy
Date
2020-11-23
Brought by
Private plaintiff (TCPA robocalls)
Outcome
Affirmed denial of class certification.
Admission
unknown

How the court or agency ruled, and why. Rivera offered a list of 482 names without defining the class. Court: a list is not fatal but the plaintiff bears the burden of proposing a proper class definition meeting Rule 23; the court need not do it. Procedural; no finding on whether calls violated the TCPA.

Johnson v. Exeter Finance LLC

Court
U.S. District Court, E.D. Va., Richmond
Docket
3:23-cv-833-HEH
Judge
Henry E. Hudson
Date
2024-09-27
Brought by
Private borrower
Outcome
FDCPA claim dismissed; state claims remanded to state court.
Admission
unknown

How the court or agency ruled, and why. Court held defendants held a valid security interest and present right to possession after default, so the FDCPA repossession claim failed. With the federal claim gone, it remanded state claims, noting a novel Virginia question about 'prompt' return of personal property from repossessed cars.

GM Financial / AmeriCredit

DOJ subpoena to GM Financial on subprime auto securitizations

Court
No court proceeding located
Docket
not located in this research
Date
2014-08-04
Brought by
U.S. Department of Justice
Outcome
Subpoena disclosed Aug 2014. I found no public resolution.
Admission
unknown

How the court or agency ruled, and why. No proceeding or ruling; outcome unknown.

United States v. GM Financial Company, Inc. (SCRA)

Court
U.S. District Court, N.D. Tex.
Docket
not located in this research
Date
2022-10-05
Brought by
U.S. Department of Justice
Outcome
$3,534,171 to servicemembers plus $65,480 civil penalty; 71 unlawful repossessions and 1,000+ mishandled lease terminations; credit repair and training.
Admission
no

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument). Per DOJ press release; entered order not located.

Massachusetts AG and GM Financial (Assurance of Discontinuance)

Court
Suffolk Superior Court (AOD filed, per MA AG)
Docket
not located in this research
Date
2022-03-07
Brought by
Massachusetts Attorney General
Outcome
More than $1.8M restitution to 2,000+ residents (unpaid interest on delayed GAP refunds; insufficient post-repossession information).
Admission
no

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument).

AmeriCredit Financial Services Inc. (GM Financial) v. Bell

Court
Missouri Court of Appeals, Eastern District, Division Four
Docket
ED112095 (consolidated with ED112097, ED112098, ED112167, ED112187, ED112188)
Judge
Robert M. Clayton III (author); Torbitzky and Wright concurring
Date
2024-10-29
Brought by
Borrower counterclaim against lender's deficiency suit
Outcome
Reversed trial court: GM Financial's repossession notices complied with UCC Article 9; no damages; deficiency claim not barred.
Admission
unknown

How the court or agency ruled, and why. Borrower defaulted on a $15,469 loan (2008); car repossessed and sold privately; $8,251.80 deficiency sought. Trial court found two notice violations and barred any deficiency. Appeals court found none: notice adequately identified a private sale 'after 10 days'; 'may reduce or increase' language on proceeds was not misleading; post-default interest was allowed; a $92 overestimate on a $14,507 redemption figure was immaterial; letterhead sufficed to authenticate. Note: pro-lender ruling.

Carvana

Jennings v. Carvana LLC

Court
U.S. District Court, E.D. Pa.
Docket
21-5400
Judge
Judge Smith
Date
2022-09-30
Brought by
Private plaintiffs (breach of contract, PA UTPCPL: late titles)
Outcome
Motion to compel arbitration and motion to dismiss denied.
Admission
unknown

How the court or agency ruled, and why. Pennsylvania requires an installment sale to be in a single document (the RISC). Plaintiffs also signed a separate purchase agreement and arbitration agreement not referenced in the RISC; the court held those were subsumed by the RISC, which had an integration clause. Because the rule applies to all contract terms, it is a generally applicable contract defense not preempted by the FAA. Claims pleaded plausibly.

Jennings v. Carvana LLC; Harvin v. Carvana LLC

Court
U.S. District Court, E.D. Pa.
Docket
21-5400; 23-2068
Judge
Judge Perez
Date
2026-01-06
Brought by
Private plaintiffs
Outcome
Motion to dismiss UTPCPL claims granted in part, denied in part.
Admission
unknown

How the court or agency ruled, and why. Dismissed the 'double-dipping' theory ($590 delivery charge allegedly embedded in price and charged again) as speculative. Kept: (1) confusing disclosure (contract lists delivery 'N/A' but total exceeds components by $590), and (2) aggregation of government fees (registration, title, plates, lien) into one line. Plausibility standard at pleading stage; no merits finding.

Illinois Secretary of State and Carvana (licensing settlement)

Court
Administrative (no court)
Docket
not located in this research
Date
2023-01-24
Brought by
Illinois Secretary of State Alexi Giannoulias
Outcome
Carvana forfeited $250,000 bond, admitted violating title/registration law, accepted enhanced inspections and risk of summary suspension.
Admission
yes

How the court or agency ruled, and why. Administrative settlement, not a judicial ruling. Per news reports, titles sometimes took 4-6 months against a 20-day requirement and out-of-state temporary tags were used improperly. Note: this is the Secretary of State, not the AG.

Michigan Department of State v. Carvana LLC (Novi dealership)

Court
Administrative
Docket
not located in this research
Date
2022-10-07
Brought by
Michigan Department of State
Outcome
Summary suspension of Novi dealership after 18-month probation ($2,500, May 2021) and probation extension ($5,000, Feb 2022). Cited 112 late title/registration filings, employees destroying title documents on three vehicles, and 127 probation violations.
Admission
unknown

How the court or agency ruled, and why. Administrative; state's allegations, no court ruling.

Texas DMV citations against Carvana

Court
Administrative
Docket
not located in this research
Date
reported 2021
Brought by
Texas Department of Motor Vehicles
Outcome
More than 30 violations and over $10,000 in fines since 2019; open investigations reported.
Admission
unknown

How the court or agency ruled, and why. News report only (CBS Texas); I did not locate the underlying orders.

United Association National Pension Fund v. Carvana Co. (consolidated securities litigation)

Court
U.S. District Court, D. Ariz.
Docket
2:22-cv-02126
Judge
Michael T. Liburdi
Date
2025-02-03 (reconsideration ruling)
Brought by
Shareholders
Outcome
Motions to dismiss denied in 2024; motion for reconsideration denied Feb 2025; case proceeds to discovery.
Admission
unknown

How the court or agency ruled, and why. Complaint alleges Carvana misled investors about sustainable sales growth and compliance with state title/registration laws. I only have a plaintiffs' law firm summary of the ruling; the court's detailed reasoning was not read.

Credit Acceptance

41-state attorney general settlement with Credit Acceptance Corporation (New Jersey and Maryland co-leading; consent judgments)

Court
Filed in state courts per state; the only filed instrument I opened is California's proposed judgment in Superior Court of California, County of Alameda (People of the State of California v. Credit Acceptance Corporation). Other states' courts and NY federal filing not confirmed.
Docket
not located in this research
Date
2026-09-17 (parties signed; effective date Nov 2, 2026)
Brought by
41 state AGs / consumer protection offices (NJ, MD co-lead; NY, CA participating)
Outcome
About $694M-$710M headline depending on source: $60M cash restitution, about $388M debt relief (repossessed vehicles), about $246M debt relief (non-repossessed), about $15.5M to AGs; 5-7 years of injunctive terms (income verification, term caps, price cap at 109% of retail book value, 95% deficiency waiver and no collection suits on early-default high-risk loans, add-on product disclosure/cancellation, dealer monitoring).
Admission
no

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument). The California proposed judgment states it is entered 'without trial or adjudication of any issue of fact or law, and without finding or admission of wrongdoing or liability of any kind'; the judge's signature line was blank in the copy I opened (awaiting approval). Allegations only: originating loans it knew or should have known borrowers could not afford (using proprietary scores predicting collections), and failing to stop dealers packing vehicle service contracts and GAP.

Consumer Financial Protection Bureau and People of the State of New York v. Credit Acceptance Corporation

Court
U.S. District Court, S.D.N.Y.
Docket
1:23-cv-00038
Judge
Jennifer H. Rearden
Date
2023-01-04 filed; CFPB withdrawn 2025-04; NY claims settled 2026-09 (see notes)
Brought by
CFPB and New York Attorney General
Outcome
Filed Jan 4, 2023 (not 2020). CFPB withdrew as plaintiff April 2025 (unopposed motion filed Apr 24, granted Apr 29 per CFPB page). NY AG continued as sole plaintiff. Motion to dismiss was fully briefed and pending at the time of the CFPB withdrawal; I found no ruling on it. NY reported settling as part of the Sept 2026 deal; I did not find the NY consent judgment or docket entry.
Admission
no

How the court or agency ruled, and why. No merits ruling located. The only reasoned rulings I found are the Aug 2023 stay order and the withdrawal order (separate entries). Alleged: deceptive and abusive practices (unaffordable loans, hidden costs, add-on products, misleading securitization disclosures). Orrick and CA release say NY is concurrently settling federal litigation; the terms/entry are not verified.

CFPB v. Credit Acceptance Corp., stay order

Court
U.S. District Court, S.D.N.Y.
Docket
1:23-cv-00038 (23 Civ. 00038)
Judge
Jennifer H. Rearden
Date
2023-08-07
Brought by
CFPB and NY AG (plaintiffs); Credit Acceptance moved to stay
Outcome
Motion to stay granted pending U.S. Supreme Court decision in the CFPB funding-structure case (CFSA).
Admission
unknown

How the court or agency ruled, and why. Judge Rearden stayed the case because the Supreme Court was about to decide whether the CFPB's funding mechanism is constitutional, which could affect whether the Bureau could sue at all. She reasoned that (1) waiting would avoid wasted discovery, (2) the federal and New York state claims overlap so New York was not significantly delayed, and (3) the stay would be short (until the end of the 2024 Term). She rejected the claim that delay harms consumers because preservation duties protect evidence. This was a procedural ruling, not a merits ruling.

CFPB unopposed motion to withdraw as plaintiff, CFPB v. Credit Acceptance

Court
U.S. District Court, S.D.N.Y.
Docket
1:23-cv-00038
Judge
Jennifer H. Rearden (presiding judge on the docket)
Date
2025-04-24 filed; 2025-04-29 granted (per CFPB page)
Brought by
CFPB
Outcome
CFPB dropped out; NY AG continued for New York consumers only. No ruling on the pending motion to dismiss located.
Admission
no

How the court or agency ruled, and why. No reasoned opinion; withdrawal granted on consent. Company statement (press release) says it believes the case never should have been brought; that is advocacy, not a finding.

Commonwealth of Massachusetts (AG Healey) v. Credit Acceptance Corporation

Court
Suffolk Superior Court (Massachusetts)
Docket
not located in this research
Date
2020-08-31 filed; 2021-09-01 settlement announced
Brought by
Massachusetts Attorney General
Outcome
$27.2M settlement with debt relief and credit repair for 3,000+ borrowers. Complaint alleged unfair loans borrowers could not repay, hidden finance charges over the 21% usury cap, faulty repossession notices, and misleading investors about securitized loan pools.
Admission
no

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument). I found no reported ruling on motions before settlement.

Davis et al. v. Credit Acceptance Corporation et al.

Court
U.S. District Court, E.D. Mich.
Docket
2:22-cv-10367
Judge
George Caram Steeh
Date
2022-08-24
Brought by
Private plaintiffs (TILA, Odometer Act, Magnuson-Moss claims)
Outcome
Motion to dismiss and compel arbitration denied.
Admission
unknown

How the court or agency ruled, and why. The contract had an arbitration clause with a 30-day opt-out requiring signatures of all buyers. Plaintiffs' lawyer sent an opt-out letter on their behalf. Judge Steeh held that under Michigan agency law an authorized attorney's signature binds the principal, so the opt-out was valid and arbitration could not be compelled; insisting on personal signatures was an overly technical reading.

Criminal prosecutions: Honor Finance and Tricolor, and dealership fraud

United States v. James Collins (also Robert DiMeo, Michael Walsh)

Court
U.S. District Court, N.D. Ill.
Docket
not located in this research
Judge
Franklin W. Valderrama (sentencing)
Date
Indictments 2020-05-15 (mail fraud) and 2022-12-16 (bank/securities fraud); sentenced March 2025
Brought by
U.S. Attorney, N.D. Ill.
Outcome
Collins pleaded guilty to mail fraud (stipulated bank fraud); 4 years prison; about $67M restitution. Schemes 2015-2018: falsified data to a bank on a $200M credit line, put known-delinquent ineligible loans into a securitization trust, and a GPS/warranty commission diversion of about $5.3M via LHS Solutions. Co-defendant DiMeo: one day imprisonment and 12 months supervised release per SEC release.
Admission
yes

How the court or agency ruled, and why. Sentencing described in DOJ press release; I did not read a sentencing transcript, so the judge's stated reasons are not captured. Docket number not verified (searches returned several unrelated 'United States v. Collins' dockets). DOJ release gives sentencing date as 'Wednesday, March 7, 2025', an inconsistent weekday; treat exact date as unverified.

SEC v. James R. Collins and Robert F. DiMeo

Court
U.S. District Court, N.D. Ill.
Docket
1:21-cv-05040
Date
2021 complaint; final judgment 2025-12-03
Brought by
SEC
Outcome
Final judgments: Collins $550,332 disgorgement and interest; DiMeo $198,731; permanent antifraud injunctions and officer/director bars; criminal restitution of $67,243,790.94 credited.
Admission
unknown

How the court or agency ruled, and why. Judgments entered after criminal case; reasoning not in the litigation release. Concerned false statements in a $100M subprime auto-backed securities offering.

United States v. Daniel Chu et al.

Court
U.S. District Court, S.D.N.Y.
Docket
1:25-cr-00579 (from CourtListener search-result title; docket page blocked by robots.txt)
Judge
Sources conflict: Kevin Castel (June 2026 news) vs. Lewis J. Liman (FDIC-OIG summary for pleas); unresolved
Date
Charged 2025-12-17; superseding indictment unsealed 2026-06-24; arraignment 2026-06-30
Brought by
U.S. Attorney, S.D.N.Y.
Outcome
Pending. Chu pleaded not guilty; superseding indictment reportedly eight counts including continuing financial crimes enterprise (18 U.S.C. 225). Trial reported for Oct 19, 2026 (alternate Feb 1, 2027). Cooperating co-defendants reported pleaded guilty (Kollar, Seibold; COO Goodgame per NY Daily Record June 2026, though FDIC-OIG and NIADA said he was charged and pleaded not guilty in Jan). Alleged: double-pledged collateral; about $2.2B pledged vs about $1.4B real; delinquent loans made to look current; Chapter 7 filed Sept 2025.
Admission
unknown

How the court or agency ruled, and why. Not decided. Charges are allegations. Confirm current trial status directly with the docket.

SEC v. Daniel Chu, Jerome Kollar, Ameryn Seibold

Court
U.S. District Court, S.D.N.Y.
Docket
not located in this research
Date
2026-08-18 (per SEC page)
Brought by
SEC
Outcome
Pending civil complaint seeking injunctions, disgorgement, penalties, officer/director bars. SEC complaint PDF appeared in search results.
Admission
unknown

How the court or agency ruled, and why. Allegations only.

United States v. Mohamad Jihad Fakih

Court
U.S. District Court, M.D. Fla.
Docket
not located in this research
Judge
Virginia M. Hernandez Covington
Date
2026-03-02 sentenced
Brought by
U.S. Attorney, M.D. Fla.
Outcome
54 months; $378,886.96 forfeiture; wire fraud conspiracy using straw buyers, plus attempted export of a stolen Rolls-Royce.
Admission
yes

How the court or agency ruled, and why. Fraud against lenders, not by a subprime lender; included as dealer/auto-loan fraud context. Verdict Aug 21, 2025.

United States v. Cartier et al. (fake-dealership auto loan scheme)

Court
U.S. District Court, N.D. Ga. (inferred from U.S. Attorney Pak)
Docket
not located in this research
Date
undated in article
Brought by
U.S. Attorney, N.D. Ga.
Outcome
Seven defendants sentenced (4 yr 9 mo for Cartier down to probation); about $1.7M loss to lenders, mostly credit unions.
Admission
yes

How the court or agency ruled, and why. Fraud against lenders; tangential. Court and case number not confirmed.

Other lenders: Toyota, Wells Fargo, Ally, United Auto Credit, CPS, Westlake, DriveTime

CFPB/DOJ resolution: discriminatory dealer markup

Court
CFPB administrative order; DOJ complaint and proposed consent order in C.D. Cal.
Docket
not located in this research
Date
2016-02-02
Brought by
CFPB and DOJ
Outcome
$19.9M to affected borrowers (+ up to $2M); markup caps cut to 1.25% (loans up to 5 years) or 1%; no civil penalty.
Admission
no

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument). Investigation did not find intentional discrimination; discretionary pricing produced disparities.

In the Matter of Toyota Motor Credit Corporation, 2023-CFPB-0015

Court
CFPB administrative
Docket
2023-CFPB-0015
Date
2023-11-17 (order); terminated 2025-05-12 per order title
Brought by
CFPB
Outcome
$12M civil penalty; about $46M+ redress. Findings: hard-to-cancel add-ons, unrefunded unearned GAP/CLAH, inaccurate credit reporting on 27,507+ lease accounts.
Admission
no

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument). Consented 'without admitting or denying' findings.

In the Matter of Wells Fargo Bank, N.A., 2018-BCFP-0001

Court
CFPB administrative (parallel OCC)
Docket
2018-BCFP-0001
Date
2018-04-20
Brought by
CFPB and OCC
Outcome
$1B total ($500M credited to OCC) covering force-placed collateral protection insurance on auto loans and mortgage rate-lock fees; later modified Dec 2022 and terminated.
Admission
unknown

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument).

In the Matter of Wells Fargo Bank, N.A., 2022-CFPB-0011

Court
CFPB administrative
Docket
2022-CFPB-0011
Date
2022-12-20
Brought by
CFPB
Outcome
$3.7B overall; auto servicing failures 2011-2022 across 11M+ accounts including misapplied payments, erroneous fees, wrongful repossessions, unrefunded GAP; about $1.3B auto redress per summary, plus civil penalty.
Admission
no

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument). Consented 'without admitting or denying' findings, admitting only jurisdictional facts.

CFPB/DOJ Ally consent order and United States v. Ally Financial (ECOA dealer markup)

Court
CFPB 2013-CFPB-0010; DOJ filing in U.S. District Court, E.D. Mich.
Docket
2:13-cv-15180 (from DOJ filing title in search result)
Date
2013-12-20
Brought by
CFPB and DOJ
Outcome
$80M damages plus $18M civil penalty for discriminatory dealer markups (April 2011-Dec 2013).
Admission
unknown

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument). CFPB findings recited in order; press summary did not describe a formal admission.

Massachusetts AG and United Auto Credit Corporation (assurance)

Court
Suffolk Superior Court
Docket
not located in this research
Date
2021-05-24
Brought by
Massachusetts Attorney General
Outcome
$250,000 plus waiver of about $336k and $162k in deficiencies, release of about $31k in judgments, about $14k restitution; changes to collections, garnishment, and dealer complaint tracking.
Admission
no

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument). UACC denied all allegations. Alleged: bought contracts from two problematic dealers, weak dealer oversight, overbroad releases in voluntary surrender agreements.

Davidson v. United Auto Credit Corporation

Court
U.S. Court of Appeals, 4th Circuit (appeal from E.D. Va., Judge Leonie Brinkema)
Docket
21-1697
Judge
Richardson (author), Thacker; Wilkinson dissenting
Date
2023-04-12
Brought by
Private plaintiff (servicemember; Military Lending Act)
Outcome
Affirmed dismissal: loan financing a car plus GAP fits the MLA's exemption for credit 'for the express purpose of financing the purchase' of a vehicle.
Admission
no

How the court or agency ruled, and why. Majority: 'express purpose' means specific, not sole. The exemption is a checklist; once its elements are met (credit is to finance the vehicle purchase, secured by it) additional financed items like GAP do not remove it. Dissent (Wilkinson): GAP is a standalone product, so the exemption requires the loan to be solely for the car; warned of a slippery slope (car washes, parking). Pro-lender ruling; no post-decision history located.

United States (FTC) v. Consumer Portfolio Services, Inc.

Court
U.S. District Court, C.D. Cal.
Docket
(FTC matter no. 112-3010 per release; court docket not given)
Date
2014-05-28
Brought by
FTC (DOJ filed complaint)
Outcome
Consent decree: $2M civil penalty; $3.5M+ refunds/adjustments for 128,000 consumers; forbearance on 35,000 accounts; data integrity program and 10 years of assessments.
Admission
no

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument). Alleged FTC Act, FDCPA, FCRA violations (harassing calls, fees not owed, false repossession threats, caller ID manipulation). Judge not identified.

In the Matter of Westlake Services, LLC and Wilshire Consumer Credit, LLC

Court
CFPB administrative
Docket
not located in this research
Date
2015-10-01
Brought by
CFPB
Outcome
$4.25M civil penalty and about $44.1M redress ($25.8M cash, rest balance reductions) for false caller IDs, false threats, unlawful third-party disclosure, unauthorized loan modifications, missing APR disclosures.
Admission
no

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument). Consent order; no formal admission.

In the Matter of DriveTime Automotive Group, Inc. and DT Acceptance Corp.

Court
CFPB administrative
Docket
not located in this research
Date
2014-11-19
Brought by
CFPB
Outcome
$8M civil penalty; end workplace calls on request, stop furnishing unconfirmed repossession data, correct credit reports, monthly audits.
Admission
unknown

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning (consent/settlement instrument). Consent order; admission not detailed in the summary I opened.

Notes on the auto-lender records

  • Credit Acceptance CFPB/NY suit was filed Jan 4, 2023 in S.D.N.Y. (not 2020); the 2020 suit was Massachusetts AG in Suffolk Superior Court.
  • I could not find a ruling on Credit Acceptance's motion to dismiss in the S.D.N.Y. case; it was pending at CFPB withdrawal. The only reasoned rulings found are the Aug 2023 stay order and the withdrawal order.
  • Sept 2026 Credit Acceptance figures vary by source ($694M, $700M, $709.5M, $710M) because of different counting of debt relief and payments.
  • Web summaries were produced by a fetch tool that summarizes pages; quoted holdings should be checked against the linked opinion PDFs before publication.
  • Tricolor sources conflict on judge and on which executives pleaded guilty.
  • Blocked or not fetchable: CourtListener dockets (robots.txt) for Chu and Collins, NJ AG site (403), Civil Rights Litigation Clearinghouse (403).

Other predatory-lending cases

The essay is about auto loans, but the legal material on other forms of predatory lending is a useful comparison. The payday and tribal-lending prosecutions matter most, because they show prosecutors treating a high-interest consumer lender as a criminal enterprise, which has not happened to any large subprime auto lender. Several entries are included for contrast and are labelled as such in the notes at the end of this section.

Payday and tribal lending (18)

United States v. Tucker and Muir (trial and sentencing) (Scott Tucker and Timothy Muir (AMG/Tucker payday enterprise))

Court
U.S. District Court, S.D.N.Y. (Judge P. Kevin Castel)
Docket
not located in this research
Date
Verdict 2017-10-13; sentenced 2018-01-05
Brought by
U.S. Attorney, S.D.N.Y.
Outcome
Jury convicted both on all 14 counts (RICO conspiracy, RICO, wire fraud, money laundering, Truth in Lending Act violations) after a five-week trial. Tucker sentenced to 200 months; Muir to 84 months.
Criminal or civil
criminal
Sentence
Tucker 200 months + 3 yrs supervised release; Muir 84 months + 3 yrs; forfeiture ordered

How the court or agency ruled, and why. Jury verdict, so no reasoned opinion on the merits. The government's theory was that the 'tribal' ownership was a sham used to evade state usury laws while the business ran from Kansas, and that TILA disclosures hid true costs (a $500 loan that disclosed a $650 payoff actually cost $1,925 by the court's account at sentencing). Judge Castel, per DOJ's release, described the scheme as one 'to extract money from people in desperate circumstances.'

United States v. Tucker (Muir), appeal (Scott Tucker and Timothy Muir)

Court
U.S. Court of Appeals, 2d Cir. (Leval, Pooler, Parker)
Docket
18-181
Date
2020-06-02
Brought by
United States (appellee)
Outcome
Convictions affirmed on all counts.
Criminal or civil
criminal
Sentence
Sentences not disturbed (see pt-01)

How the court or agency ruled, and why. The court rejected the argument that tribal involvement exempted the loans from state usury law, treating the tribal arrangement as a sham: Tucker kept capital, control and operations in Kansas and paid the tribes about 1% of revenue. Because the jury convicted on the RICO conspiracy count under a correct willfulness standard, any error in the willfulness instruction on the substantive RICO counts did not affect substantial rights (plain-error review). Sufficient evidence supported TILA counts: the disclosure box showed only the cost of the opt-out path while borrowers were defaulted into renewals with far higher finance charges (often exceeding 600% annualized).

FTC v. AMG Services, Inc. (district court judgment) (AMG Services / Scott Tucker and affiliated companies)

Court
U.S. District Court, D. Nev. (Chief Judge Gloria M. Navarro)
Docket
2:12-cv-01532
Date
2016-10-04
Brought by
Federal Trade Commission
Outcome
Summary judgment for FTC; roughly $1.3 billion judgment; ban from consumer lending.
Criminal or civil
civil

How the court or agency ruled, and why. The court found Tucker ran the operation and was individually liable for violating Section 5 of the FTC Act through false statements about loan costs and deceptive fee structures, and it banned prohibited practices such as conditioning credit on preauthorized electronic transfers. (Monetary relief was later undone by the Supreme Court; see pt-04.)

AMG Capital Management, LLC v. FTC (AMG Capital Management, LLC (Tucker entities))

Court
U.S. Supreme Court (Justice Breyer, unanimous), on cert from the 9th Cir. (910 F.3d 417)
Docket
19-508
Date
2021-04-22
Brought by
Tucker entities (petitioners) against the FTC
Outcome
9-0: Section 13(b) of the FTC Act does not authorize equitable monetary relief such as restitution or disgorgement.
Criminal or civil
civil

How the court or agency ruled, and why. The Court read 'permanent injunction' as prospective relief only, noted that Section 13(b)'s structure focuses on injunctions, and reasoned that Congress would not have enacted Section 19 (which expressly authorizes conditioned, limited monetary relief two years later) if 13(b) already implicitly allowed the same relief without those limits. Consequence: the FTC lost its principal tool for returning money to borrowers, though DOJ's forfeiture from the criminal case (see pt-05) later funded more than $500M in victim payments.

SDNY non-prosecution agreements and forfeiture return to victims (Modoc Tribe of Oklahoma; Santee Sioux Nation of Nebraska (tribal corporations used by Tucker); Miami Tribe of Oklahoma)

Court
U.S. Attorney's Office, S.D.N.Y. (agreements, not a court ruling)
Docket
not located in this research
Date
2018-06-26
Brought by
U.S. Attorney, S.D.N.Y.
Outcome
Two tribes forfeited $3M under non-prosecution agreements; DOJ said more than $500M in recovered funds (including $48M from the Miami Tribe, and sums from Tucker and U.S. Bancorp) would go to the FTC for victim redress.
Criminal or civil
criminal

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning. The tribal corporations acknowledged, per DOJ, that Tucker used his agreements with them to evade state usury laws.

United States v. Hallinan and Neff (trial verdict) (Charles M. Hallinan and Wheeler K. Neff)

Court
U.S. District Court, E.D. Pa.
Docket
not located in this research
Date
2017-11-27
Brought by
U.S. Attorney, E.D. Pa.
Outcome
Jury found both guilty on all counts: RICO conspiracy, conspiracy to commit mail/wire fraud and money laundering, mail and wire fraud; Hallinan also nine counts of international money laundering.
Criminal or civil
criminal
Sentence
See pt-07

How the court or agency ruled, and why. Jury verdict; no reasoned opinion in the sources opened. Prosecutors alleged the pair used 'rent-a-tribe' and straw-lender structures to collect unlawful debt (loans over 700% APR) totaling about $688M in 2008-2013, and also defrauded roughly 1,400 plaintiffs in a related lawsuit.

United States v. Hallinan (sentencing) (Charles M. Hallinan ('godfather of payday lending'))

Court
U.S. District Court, E.D. Pa.
Docket
not located in this research
Date
2018-07-06
Brought by
U.S. Attorney, E.D. Pa.
Outcome
168 months in prison; $2.5M fine; $64M forfeiture judgment plus forfeiture of accounts, cars and a Villanova mansion.
Criminal or civil
criminal
Sentence
168 months; $2.5M fine; $64M forfeiture

How the court or agency ruled, and why. Sentencing, per DOJ: Hallinan ran payday businesses from 1997-2013 and hid his role behind straw lenders including a federally insured bank and tribal entities. Judge not named in the release opened.

United States v. Neff (sentencing) (Wheeler K. Neff (Delaware attorney for Hallinan))

Court
U.S. District Court, E.D. Pa.
Docket
not located in this research
Date
2018-05-25
Brought by
U.S. Attorney, E.D. Pa.
Outcome
8 years in prison (release also reports a $50,000 fine per its title).
Criminal or civil
criminal
Sentence
96 months

How the court or agency ruled, and why. Sentencing; no reasoned opinion in sources opened. Per DOJ, Neff drafted contracts and structures that falsely claimed tribal ownership and immunity, enabling over $490M in unlawful debt collection in 2008-2013.

United States v. Hallinan (forfeiture appeals) (Charles Hallinan (forfeiture; Linda Hallinan as third-party claimant))

Court
U.S. Court of Appeals, 3d Cir. (Krause, Bibas, Rendell)
Docket
21-1362, 21-2623, 22-1953 & 22-2328
Date
2023-07-19
Brought by
United States (appellee)
Outcome
Affirmed in part; dismissed in part for lack of jurisdiction.
Criminal or civil
criminal

How the court or agency ruled, and why. Third parties cannot challenge the forfeiture order against a defendant; only the defendant can. Linda Hallinan failed to show she was a bona fide purchaser or held a superior property interest, because the criminal enterprise acquired the assets before transferring them to her. Subpoena-related appeals were not final orders. The opinion is about forfeiture of proceeds, not the merits of the lending scheme.

CFPB v. CashCall, Inc. (district court, true lender) (CashCall, Inc.; WS Funding; Delbert Services; J. Paul Reddam (Western Sky loans))

Court
U.S. District Court, C.D. Cal.
Docket
2:15-cv-07522
Date
2016-08-31
Brought by
Consumer Financial Protection Bureau
Outcome
Summary judgment for the CFPB on liability; CEO held personally liable.
Criminal or civil
civil

How the court or agency ruled, and why. The court held CashCall, not Western Sky (a tribally licensed entity), was the true lender, because CashCall funded and bought every loan before any payment was collected, guaranteed Western Sky's minimum payments, set underwriting criteria and collected the payments, so it bore all the economic risk. The tribal choice-of-law clauses were not enforced because the borrowers' home-state interests outweighed tribal sovereignty; loans above state caps were void or uncollectible under state law, so collecting on them was deceptive.

CFPB v. CashCall, Inc. (appeal) (CashCall, Inc. and J. Paul Reddam)

Court
U.S. Court of Appeals, 9th Cir. (Owens, Nelson, Miller)
Docket
18-55407, 18-55479
Date
2022-05-23
Brought by
CFPB (appellee/cross-appellant)
Outcome
Liability affirmed; civil penalty and denial of restitution vacated and remanded.
Criminal or civil
civil

How the court or agency ruled, and why. (1) CFPB's enforcement authority stood because the case was filed under a lawfully appointed director (Collins v. Yellen). (2) The tribal-law clause was unenforceable: the Cheyenne River Sioux Tribe had no substantial relationship to the parties, borrowers applied and paid from their home states, and CashCall was the real economic actor behind a shell lender, so state law governed and the loans were void. (3) Demanding payment on legally unenforceable debts is deceptive even if the unenforceability comes from state law. (4) CashCall's conduct became reckless after Sept. 2013 when counsel warned of regulatory risk, supporting tier-two penalties, and the CEO could not rely on advice of counsel. (5) The district court erred in requiring intent to defraud or lack of benefit of the bargain for restitution; on remand restitution may be measured by net revenue, potentially exceeding profit.

Commonwealth of Pennsylvania v. Think Finance, Inc. (Think Finance, Inc.; Kenneth Rees; National Credit Adjusters and others)

Court
U.S. District Court, E.D. Pa. (Judge Joyner)
Docket
14-7139
Date
2019-11-18 (summary judgment ruling); AG settlement reported 2019
Brought by
Pennsylvania Attorney General
Outcome
All pending summary judgment motions denied (both sides), so no liability finding; a Pennsylvania settlement was later reported (see northcentralpa link, search-result-only).
Criminal or civil
civil

How the court or agency ruled, and why. The court found genuine factual disputes on the state RICO-type (Corrupt Organizations Act), Fair Credit Extension Uniformity Act, UTPCPL and Dodd-Frank claims, mainly over how personally involved Rees was and whether the defendants agreed to and knew of unlawful acts. The opinion describes loans averaging 250-390% APR to Pennsylvanians, against a 6% cap for unlicensed lenders, arranged with tribes as fronts, but it did not resolve tribal immunity as a dispositive issue.

Gingras v. Think Finance, Inc. (Think Finance, Inc. and related entities; Chippewa Cree tribal lending entities (Plain Green))

Court
U.S. Court of Appeals, 2d Cir. (Leval, Hall, Chin)
Docket
16-2019 (consolidated with 16-2132, 16-2135, 16-2138, 16-2140)
Date
2019-04-24
Brought by
Borrower class (private)
Outcome
Affirmed denial of motions to compel arbitration and to dismiss; RICO and other claims proceeded.
Criminal or civil
civil

How the court or agency ruled, and why. Tribal sovereign immunity does not bar suits against tribal officials for prospective injunctive relief over off-reservation violations of state and federal law (an Ex parte Young analogue). The arbitration clauses were unenforceable and unconscionable because they required tribal law only, disclaimed state and federal protections, and offered an illusory review mechanism given tribal courts' unchecked discretion.

Williams v. Big Picture Loans, LLC (Big Picture Loans, LLC; Ascension Technologies (Lac Vieux Desert Band))

Court
U.S. Court of Appeals, 4th Cir. (Gregory, Agee, Diaz)
Docket
18-1827
Date
2019-07-03
Brought by
Five Virginia borrowers (private)
Outcome
Reversed the district court; the two lenders held to be arms of the tribe entitled to sovereign immunity. A useful counterexample to Tucker and Gingras.
Criminal or civil
civil

How the court or agency ruled, and why. Applying the five-factor Breakthrough test (creation under tribal law, purpose, tribal control, tribal intent to share immunity, and financial relationship), the court found each factor favored immunity: the entities were formed by tribal council resolution, served economic development goals, were co-managed by tribal officials, and produced roughly 10% of the tribe's general fund. The court said immunity covers commercial activity and does not depend on judging the business's respectability. Borrowers alleged loan rates about 50 times Virginia's limit.

Otoe-Missouria Tribe of Indians v. New York State Department of Financial Services (Otoe-Missouria Tribe and tribal online lenders vs. New York DFS)

Court
U.S. Court of Appeals, 2d Cir. (Sack, Lynch, Lohier)
Docket
13-3769
Date
2014-10-01
Brought by
Tribal lenders (plaintiffs) suing the state regulator
Outcome
Denial of preliminary injunction affirmed; New York's efforts to choke off tribal lenders' bank access could continue.
Criminal or civil
civil

How the court or agency ruled, and why. Tribal sovereignty limits on state power depend on where the regulated conduct occurs. The lenders failed to show the transactions took place on tribal land; loans went to New York residents who never left New York and were repaid by automatic debits from New York bank accounts, so state usury law could reach the conduct. The record was thin on where servers, staff and underwriting sat. The claim of discriminatory targeting also failed on that record because DFS's letters to banks addressed all payday lenders.

FTC v. Payday Financial LLC et al. (settlement) (Martin A. Webb and companies (Western Sky Financial, Payday Financial, others))

Court
U.S. District Court, D.S.D.
Docket
not located in this research
Date
2014-04-04 (approved); announced 2014-04-11
Brought by
Federal Trade Commission
Outcome
Settlement: $967,740 ($550,000 civil penalty plus $417,740 surrendered gains); ban on suing consumers in tribal court and on unlawful garnishment.
Criminal or civil
civil

How the court or agency ruled, and why. Settlement; no adjudicated findings and no judicial reasoning. The FTC alleged the defendants tried to garnish wages without court orders and sued borrowers nationwide in Cheyenne River Sioux tribal court, which lacked jurisdiction.

Madden v. Midland Funding, LLC (Midland Funding LLC and Midland Credit Management (debt buyer of Bank of America/FIA credit card debt))

Court
U.S. Court of Appeals, 2d Cir. (Leval, Straub, Droney)
Docket
14-2131
Date
2015-05-22
Brought by
Saliha Madden (private class plaintiff)
Outcome
Reversed the National Bank Act preemption ruling and remanded. The Supreme Court later declined review (per a search result). It is the case usually cited in true-lender and rent-a-bank debates.
Criminal or civil
civil

How the court or agency ruled, and why. Non-bank debt buyers cannot claim National Bank Act preemption of state usury law merely because they bought the debt from a national bank. Preemption reaches non-banks only when they act for the bank or when applying state law would significantly interfere with the bank's exercise of its powers, and applying New York usury law to a buyer would not stop banks from selling debt. Note: this was a credit-card debt case, not a payday loan, and the holding was later legislatively and administratively contested (OCC/FDIC valid-when-made rules).

In re ACE Cash Express (CFPB consent order 2014-CFPB-0008) (ACE Cash Express, Inc.)

Court
CFPB administrative proceeding
Docket
2014-CFPB-0008
Date
2014-07-10
Brought by
CFPB
Outcome
Consent order: $5M consumer refunds and $5M civil penalty.
Criminal or civil
civil

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning. CFPB alleged false threats of lawsuits and prosecution, harassing calls, and pressuring overdue borrowers into new payday loans they could not afford.

Mortgage and financial-crisis era (10)

Multistate settlement with state attorneys general and regulators (Household International (Household Finance, Beneficial))

Court
State AG/regulator settlement (court filing not identified)
Docket
not located in this research
Date
2002-10-11
Brought by
About 20 state AGs and regulators (per NY AG release)
Outcome
$484M nationwide restitution (New York share over $37M); five-year independent monitor; caps on fees and end of single-premium credit insurance on real estate loans.
Criminal or civil
civil

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning. States alleged omitted taxes and closing costs, undisclosed rates above 21% on second loans, repeated refinancings and fees within short periods (flipping), and single-premium credit insurance.

Multistate settlement (Ameriquest Mortgage Company / ACC Capital Holdings)

Court
State AG/regulator settlement (court filing not identified)
Docket
not located in this research
Date
2006-01-23
Brought by
State AGs and regulators (48 states per NY AG release as summarized)
Outcome
$325M ($295M consumer restitution, $30M to states) and lending reforms with an independent monitor.
Criminal or civil
civil

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning. States alleged discount points that benefited sales staff, concealed costs, inaccurate estimates, falsified income documentation, pressure on appraisers, and closing loans before corporate approval.

Multistate settlement (stipulated judgment and injunction following California AG suit of 2008-06-30) (Countrywide Financial Corporation)

Court
Court filings by the participating states (identifying details not confirmed)
Docket
not located in this research
Date
2008-10-06
Brought by
11 state AGs led by California and Illinois (CA, AZ, CT, FL, IL, IA, MI, NC, OH, TX, WA)
Outcome
Up to $8.68B nationally in loan modifications, rate cuts and principal reductions ($3.5B in California).
Criminal or civil
civil

How the court or agency ruled, and why. Stipulated judgment; no adjudicated findings and no judicial reasoning. States alleged misrepresented terms, payment increases and affordability in subprime and pay-option ARM loans.

United States ex rel. O'Donnell v. Countrywide Home Loans, Inc. (the 'Hustle' / High Speed Swim Lane case), district court (Bank of America, N.A. (as successor to Countrywide); Rebecca Mairone)

Court
U.S. District Court, S.D.N.Y. (Judge Jed S. Rakoff)
Docket
not located in this research
Date
2014 penalty ruling (following a jury verdict)
Brought by
United States (FIRREA civil fraud), on a relator's claims
Outcome
Jury found civil liability for fraud on Fannie Mae and Freddie Mac; Rakoff imposed $1,267,491,770 on Bank of America and $1,000,000 on Mairone.
Criminal or civil
civil

How the court or agency ruled, and why. Countrywide's 2007-2008 HSSL program moved approval authority from experienced underwriters to loan specialists, dropped quality checklists, cut funding time from 45-60 to 15 days and removed pay penalties for poor loans. On penalty, Rakoff read FIRREA's 'pecuniary gain or loss' as gross amounts (deterrent and punitive purpose), then scaled the penalty to 42.81% of the $2.96B maximum because about 67% of the loans proved acceptable. (The CourtListener page opened returned a summary that misidentified the court as C.D. Cal.; the 2d Cir. page identifies S.D.N.Y. and Rakoff.)

United States ex rel. O'Donnell v. Countrywide Home Loans, Inc., 822 F.3d 650 (appeal) (Countrywide Home Loans, Bank of America, Rebecca Mairone)

Court
U.S. Court of Appeals, 2d Cir. (Raggi, Wesley, Droney)
Docket
15-496, 15-499
Date
2016-05-23
Brought by
United States / relator (appellee)
Outcome
The Second Circuit reversed the fraud judgment and the $1.2B+ penalty (the outcome is stated in search-result titles from Davis Polk and Dorsey; the Leagle page I opened set out only the question presented).
Criminal or civil
civil

How the court or agency ruled, and why. The question was when a breach of contract can also be fraud. As reported by secondary sources in the search results, the court held that a party's failure to perform a promise is not fraud unless it had no intention to perform when it made the promise, so the sale of loans that later failed representations was a breach, not FIRREA-actionable fraud. Confirm against the opinion text before quoting.

Commonwealth v. Fremont Investment & Loan, 452 Mass. 733, 897 N.E.2d 548 (Fremont Investment & Loan)

Court
Massachusetts Supreme Judicial Court
Docket
not located in this research
Date
2008-12-09
Brought by
Massachusetts Attorney General
Outcome
Preliminary injunction restricting foreclosures on loans meeting four criteria affirmed.
Criminal or civil
civil

How the court or agency ruled, and why. Originating a home loan the lender should recognize the borrower is unlikely to repay can be an unfair practice under Chapter 93A. The court rejected an industry-standard defense and found the covered loans (short-teaser ARMs with initial rate 3+ points below the fully indexed rate, debt-to-income above 50% at the indexed rate, and loan-to-value at or near 100%) were 'doomed to foreclosure' absent perpetual house-price appreciation, citing federal guidance from 2001 warning against ignoring repayment ability. The fetched summary names Judge Botsford; check the opinion for the author and panel.

Bank of America Corp. v. City of Miami, 581 U.S. ___ (Bank of America, Wells Fargo (and others) sued by the City of Miami)

Court
U.S. Supreme Court (Breyer, J.; 5-3 in part, Thomas concurring in part and dissenting in part)
Docket
15-1111
Date
2017-05-01
Brought by
City of Miami (plaintiff) under the Fair Housing Act
Outcome
City is an 'aggrieved person' with standing; proximate cause needs a direct relation to the injury, foreseeability alone is not enough. Eleventh Circuit vacated and remanded.
Criminal or civil
civil

How the court or agency ruled, and why. Miami alleged the banks steered minority borrowers into riskier loans on worse terms, concentrating foreclosures and cutting tax revenue. The Court held the FHA's zone of interests reaches the city but that lower courts must define proximate cause more tightly.

United States v. Wells Fargo (fair lending settlement) (Wells Fargo Bank)

Court
U.S. District Court, D.D.C. (per fetched DOJ page; docket not stated)
Docket
not located in this research
Date
2012-07-12
Brought by
U.S. Department of Justice
Outcome
Consent settlement: $184.3M compensation for wholesale borrowers, $50M in down-payment assistance, and review of retail lending.
Criminal or civil
civil

How the court or agency ruled, and why. Settlement of allegations; the DOJ page opened did not state whether Wells Fargo admitted fault. No judicial reasoning. DOJ alleged about 4,000 minority wholesale borrowers were steered into subprime loans and about 30,000 were charged higher fees or rates based on race or national origin (2004-2009).

HSBC national mortgage settlement (HSBC (Household mortgage servicing and origination))

Court
Settlement of DOJ, CFPB, HUD and 49 state AGs plus DC (court filing not identified)
Docket
not located in this research
Date
2016-02-05
Brought by
DOJ, CFPB, HUD, state AGs; Federal Reserve separate $131M fine
Outcome
$470M ($100M cash, $370M customer relief) for abusive foreclosure, origination and servicing practices.
Criminal or civil
civil

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning. This came from an American Banker article (opened); it concerns servicing and foreclosure conduct more than origination.

United States v. Farkas (trial, sentencing, post-trial) (Lee B. Farkas (Taylor, Bean & Whitaker; Colonial Bank; Ocala Funding))

Court
U.S. District Court, E.D. Va. (Judge Leonie M. Brinkema)
Docket
1:10-cr-200 (as stated in the fetched CourtListener page: '1:10-CR-200')
Date
Sentenced 2011-06-30
Brought by
DOJ
Outcome
Convicted on 14 counts (conspiracy, bank, wire and securities fraud); 30 years and about $38.5M forfeiture; restitution of about $3.5B joint and several.
Criminal or civil
criminal
Sentence
360 months; 3 years supervised release; forfeiture about $38.5M

How the court or agency ruled, and why. Caveat: this is a fraud on banks and investors (about $2.9B in losses that contributed to Colonial Bank's failure), not predatory lending to borrowers. Farkas took over $1.4B from Colonial's mortgage warehouse division and about $1.5B from Ocala Funding using fake and double-sold mortgage assets. The CourtListener opinion I opened is a 2015 recusal ruling by Brinkema, denying recusal under 28 U.S.C. 455(a).

Credit cards (8)

In re Capital One Bank (USA) (CFPB consent order) (Capital One Bank (USA), N.A.)

Court
CFPB administrative proceeding
Docket
2012-CFPB-0001
Date
2012-07-18
Brought by
CFPB (with OCC, and a separate FDIC action)
Outcome
Consent order: about $140M in consumer refunds (widely reported total with penalties: $210M). Consent order terminated Sept. 2016.
Criminal or civil
civil

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning. Vendors allegedly pressured consumers at card activation into buying add-on products (payment protection, credit monitoring) with misleading descriptions.

In re Discover Bank (FDIC and CFPB consent orders) (Discover Bank)

Court
FDIC and CFPB administrative proceedings
Docket
not located in this research
Date
2012-09-24
Brought by
FDIC and CFPB
Outcome
About $200M refunds to 3.5M+ consumers and a $14M civil penalty.
Criminal or civil
civil

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning. Deceptive telemarketing of add-on products (Payment Protection, Credit Score Tracker, Identity Theft Protection, Wallet Protection): calling products free, enrolling consumers without consent, and hiding eligibility limits.

In re American Express (CFPB, FDIC, Fed, OCC orders) (American Express (Centurion Bank, FSB, Travel Related Services))

Court
Administrative proceedings
Docket
2012-CFPB-0004 (from the consent order PDF filename in search results)
Date
2012-10-01
Brought by
CFPB, FDIC, Federal Reserve, OCC
Outcome
$85M refunds to about 250,000 consumers; $27.5M total civil penalties across agencies ($14.1M CFPB, $3.9M FDIC, $9M Fed, $0.5M OCC).
Criminal or civil
civil

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning. Findings included an unfulfilled $300 bonus promotion, late fees exceeding what the CARD Act allowed, age-based credit scoring for applicants over 35, failure to report disputes to credit bureaus, and deceptive debt-collection statements.

In re Bank of America / FIA Card Services (CFPB and OCC) (Bank of America, N.A. and FIA Card Services, N.A.)

Court
Administrative proceedings
Docket
not located in this research
Date
2014-04-09
Brought by
CFPB and OCC
Outcome
$727M in consumer relief plus a $20M CFPB penalty; the OCC assessed a separate $25M penalty.
Criminal or civil
civil

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning. Deceptive marketing of 'Credit Protection' products (2010-2012) and unfair billing for identity-protection products (about 1.9M accounts).

FTC v. CompuCredit Corp. and Jefferson Capital Systems, LLC (with parallel FDIC action) (CompuCredit Corporation and Jefferson Capital Systems, LLC)

Court
U.S. District Court, N.D. Ga.
Docket
FTC matter no. 062-3212 (agency file number; court civil action number not confirmed)
Date
Complaint 2008-06-10; stipulated orders 2008-12-19
Brought by
Federal Trade Commission
Outcome
Stipulated orders; at least $114M in consumer redress from CompuCredit.
Criminal or civil
civil

How the court or agency ruled, and why. Settlement; no adjudicated findings and no judicial reasoning. FTC alleged deceptive marketing of subprime credit cards (hidden fees, misleading credit-limit and cost claims) and deceptive debt collection by Jefferson Capital.

New York Attorney General settlement with First Premier Bank (First Premier Bank (South Dakota))

Court
New York AG assurance (no court ruling identified)
Docket
not located in this research
Date
2007-08-15
Brought by
New York Attorney General (Andrew Cuomo)
Outcome
$4.5M in consumer refunds and $105,000 in penalties and costs; ban on charging fees before activation and on deceptive advertising in the subprime market.
Criminal or civil
civil

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning. AG alleged mailers falsely said consumers were 'pre-approved' for up to $2,000 at '9.9% APR Fixed' with 'no processing fee' while charging $178 in upfront fees and rates that could more than double.

In re Goldman Sachs Bank USA and Apple Inc. (CFPB consent orders) (Apple Inc. and Goldman Sachs Bank USA (Apple Card))

Court
CFPB administrative proceeding
Docket
not located in this research
Date
2024-10-23
Brought by
CFPB
Outcome
Over $89M ordered (redress and penalties); the consent order was terminated Sept. 22, 2025 per the CFPB page, with Apple's $25M penalty fulfilled.
Criminal or civil
civil

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning. CFPB cited customer-service breakdowns and misrepresentations affecting hundreds of thousands of Apple Card users (transaction disputes and payment handling). Not a subprime product; included as a recent credit-card enforcement action.

Federal Reserve and FDIC orders on card misclassification (Discover Bank / Discover Financial Services)

Court
Administrative proceedings
Docket
not located in this research
Date
2025-04-21
Brought by
Federal Reserve Board and FDIC
Outcome
$250M in civil penalties ($100M Fed, $150M FDIC) and at least $1.23B restitution to merchants; tied to the Capital One merger.
Criminal or civil
civil

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning. About 5 million consumer cards were classed as commercial cards for roughly 17 years, raising interchange fees paid by merchants. Harm fell on merchants rather than cardholders, so it is a comparison point rather than predatory lending to consumers.

Criminal loan sharking (7)

Perez v. United States, 402 U.S. 146 (Perez (Chicago-area loan shark))

Court
U.S. Supreme Court (Douglas, J.; 8-1, Stewart dissenting)
Docket
not located in this research
Date
1971-04-26
Brought by
United States (prosecution under Title II of the Consumer Credit Protection Act, now 18 U.S.C. 891-896)
Outcome
Conviction affirmed; Congress may reach local loan sharking under the Commerce Clause.
Criminal or civil
criminal

How the court or agency ruled, and why. Applying a class-of-activities test, the Court held that when Congress regulates a category of conduct that in the aggregate affects interstate commerce, courts may not carve out individual instances as trivial. Congress had found loan sharking supplies organized crime with its second most lucrative source of revenue. Perez had loaned money to a butcher and used threats of violence against him and his family to collect. Justice Stewart dissented that the statute reaches wholly local crime.

United States v. Madori, 419 F.3d 159 (Steven Madori)

Court
U.S. Court of Appeals, 2d Cir.
Docket
not located in this research
Date
2005-08-19
Brought by
United States
Outcome
Conviction for extortionate extension and collection of credit (18 U.S.C. 892, 894) affirmed.
Criminal or civil
criminal

How the court or agency ruled, and why. The government need not prove explicit threats. Implicit threats suffice if borrower and lender share an understanding that force may be used; the court wrote that Congress could not have meant to punish only 'those loan sharks foolish enough to make the terms explicit.' Evidence included 150% annual interest in cash, recorded veiled threats ('red zone', 'danger zone'), and Madori's involvement from the start.

United States v. Gjeli et al. (trial and sentencing) (Ylli Gjeli, Fatimir Mustafaraj, Gezim Asllani, Rezart Telushi (Philadelphia Albanian organization))

Court
U.S. District Court, E.D. Pa. (Judge William H. Yohn Jr.)
Docket
not located in this research
Date
Verdict 2014-12-15; sentenced 2015-06-11
Brought by
U.S. Attorney, E.D. Pa. / DOJ
Outcome
Guilty on all counts (racketeering conspiracy, collection of unlawful debt, extortionate credit extensions; gambling for two). Gjeli 168 months; Mustafaraj 147 months.
Criminal or civil
criminal
Sentence
Gjeli 168 months; Mustafaraj 147 months

How the court or agency ruled, and why. Jury verdict; no opinion at this stage. The group made 125 loans totaling $1.78M at 104-395% annual interest (2011-2013) and used threats and violence to collect, alongside an illegal sports-betting website with $2.9M gross profit.

United States v. Mustafaraj (Gjeli), 867 F.3d 418 (Fatmir Mustafaraj and Ylli Gjeli)

Court
U.S. Court of Appeals, 3d Cir. (Jordan, Krause, and District Judge Stearns)
Docket
15-1892 and 15-2521
Date
2017-08-11
Brought by
United States
Outcome
Convictions and prison terms affirmed; forfeiture orders remanded after Honeycutt v. United States barred joint-and-several forfeiture.
Criminal or civil
criminal
Sentence
Mustafaraj 147 months; Gjeli 168 months (affirmed)

How the court or agency ruled, and why. The court affirmed the racketeering and extortionate-credit convictions and rejected Sixth Amendment challenges to sentencing enhancements (use of an axe supported an enhancement despite acquittal on firearm counts). The forfeiture remand followed the Supreme Court's requirement of individual culpability for forfeiture.

United States v. Ragano (sentencing) (John Ragano ('Bazoo'))

Court
U.S. District Court, E.D.N.Y. (Judge Hector Gonzalez)
Docket
not located in this research
Date
2025-03-19
Brought by
U.S. Attorney, E.D.N.Y.
Outcome
37 months for extortionate collection of credit after a four-day jury trial (Oct. 2024), consecutive to a 57-month sentence, plus $3,000 forfeiture.
Criminal or civil
criminal
Sentence
37 months (served after 57-month sentence)

How the court or agency ruled, and why. Jury verdict and sentencing; no opinion in the source. Ragano lent a victim $150,000 in 2021 at about $1,800 a week, kept collecting through 2023 despite a 2022 plea and supervision, and threatened the victim at an auto parts yard in July 2023.

United States v. Semplice (sentencing) (Paul Semplice (Gambino family))

Court
U.S. District Court, E.D.N.Y. (Judge Pamela K. Chen)
Docket
not located in this research
Date
2019-03-22
Brought by
U.S. Attorney, E.D.N.Y.
Outcome
28 months for loansharking conspiracy.
Criminal or civil
criminal
Sentence
28 months

How the court or agency ruled, and why. Sentencing; no opinion in the source. Semplice ran extortionate loans at up to 54% annual interest; on one $200,000 loan he collected $9,000 monthly, keeping $8,000. He admitted striking a borrower.

United States v. Pietranico and Sarcinella (sentencing) (Dominick Pietranico and Joseph Sarcinella (Genovese family))

Court
U.S. District Court, S.D.N.Y. (Judge P. Kevin Castel)
Docket
not located in this research
Date
2014-05-13
Brought by
U.S. Attorney, S.D.N.Y.
Outcome
Five months each, one year supervised release, fines and forfeiture.
Criminal or civil
criminal
Sentence
5 months each

How the court or agency ruled, and why. Sentencing; no opinion in the source. They provided protection to a cooperating witness and extended an extortionate loan above 100% annual interest, within a waste-hauling scheme.

Title loans and pawn (3)

In re TitleMax (CFPB consent order, Military Lending Act) (TitleMax / TMX Finance entities)

Court
CFPB administrative proceeding
Docket
not located in this research
Date
2023-02-23
Brought by
CFPB
Outcome
Consent order: $5.05M consumer redress and $10M civil penalty. CFPB's first action against a nonbank title lender under the MLA.
Criminal or civil
civil

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning. CFPB found at least 2,670 title loans to covered service members at rates nearly three times the 36% MLA cap (Oct. 2016-Sept. 2021), altered borrowers' information to hide military status, and charged fees on about 15,000 loans for insurance that provided no real coverage.

CFPB v. FirstCash, Inc. (pawn loans to service members) (FirstCash, Inc. and Cash America West, Inc.)

Court
U.S. District Court, N.D. Tex.
Docket
4:21-cv-01251
Date
Filed 2021-11-12; settled 2025-07-11
Brought by
CFPB
Outcome
Settlement entered July 2025: $4M civil penalty, full consumer redress, and an obligation to offer MLA-compliant products or screen for MLA-covered borrowers.
Criminal or civil
civil

How the court or agency ruled, and why. Settlement; no adjudicated findings and no judicial reasoning. CFPB alleged over 3,600 pawn loans to covered borrowers (June 2017-May 2021, in AZ, NV, UT, WA) at APRs frequently above 200% against the 36% cap, forced arbitration clauses, missing disclosures, and violation of a 2013 order against predecessor Cash America.

In re Cash America International (CFPB consent order) (Cash America International, Inc.)

Court
CFPB administrative proceeding
Docket
2013-CFPB-0008
Date
2013-11-20
Brought by
CFPB
Outcome
Up to $14M in refunds and $5M civil penalty; the order was terminated July 22, 2025. The file number 2013-CFPB-0008 is inferred from a search-result URL (files.consumerfinance.gov/f/2013-cfpb_0008_stipulation.pdf), not stated in a fetched page.
Criminal or civil
civil

How the court or agency ruled, and why. No adjudicated findings; no judicial reasoning. Robo-signing of debt collection court filings and illegally overcharging service members; this was the CFPB's first enforcement action against a payday lender.

Other (2)

FTC v. Progressive Leasing (stipulated order) (Progressive Leasing (rent-to-own / lease-to-own))

Court
U.S. District Court, N.D. Ga.
Docket
not located in this research
Date
2020-04-20
Brought by
Federal Trade Commission
Outcome
$175M in consumer refunds and conduct rules.
Criminal or civil
civil

How the court or agency ruled, and why. Settlement; no adjudicated findings and no judicial reasoning. FTC alleged 'same as cash' marketing hid that consumers would pay about twice retail price; over 15,000 complaints in 15 months.

CFPB v. Navient (proposed order, 2024) and state AG settlement (Jan. 2022) (Navient)

Court
Federal court not identified in the CFPB page opened
Docket
not located in this research
Date
2024-09-12 (CFPB); 2022-01 (state AGs, $1.85B per NASFAA/search results)
Brought by
CFPB; state attorneys general
Outcome
CFPB proposed order: ban from federal Direct Loan servicing, $100M redress, $20M penalty (the headline reads $120M). State AG settlement reported at $1.85B in debt cancellation and payments.
Criminal or civil
civil

How the court or agency ruled, and why. Settlements; no adjudicated findings and no judicial reasoning. CFPB alleged steering borrowers into forbearance, misapplied payments and misleading cosigner-release statements.

Notes on the other cases

  • Attribution: summaries of fetched pages were produced by a small model; docket numbers, panels and holdings are as reported there and should be spot-checked against the linked opinion before publication.
  • 'verified: opened' means the page was fetched in this session; it does not mean every detail was independently confirmed. 'search-result-only' means the URL appeared in search results only.
  • Blocked: nbcnews.com (robots.txt disallowed the fetch). No circumvention attempted. The FTC press release on the $1.3B AMG judgment was opened instead.
  • One fetch (an FTC 2008 CompuCredit press release URL) returned 404 and is not cited; the CompuCredit press-release link in cc-05 is search-result-only.
  • Category caveats: Madden (pt-17) concerns credit-card debt bought from a national bank; Farkas (mg-10) is fraud on banks, not predatory lending; Discover 2025 (cc-08) harmed merchants; Apple Card (cc-07) is not subprime. They are included for comparison and labelled as such.
  • Countrywide 'Hustle': the penalty phase (mg-04) and the reversal (mg-05) are separate entries. The reversal outcome rests on search-result titles and the Leagle summary of the question presented; read the opinion (822 F.3d 650) before stating its reasoning in print.
  • Where a result was a settlement or consent order, the entry says 'no adjudicated findings; no judicial reasoning'. Judicial reasoning exists in pt-02, pt-04, pt-09 to pt-15, pt-17, ls-01, ls-02, ls-04, mg-04 to mg-07, mg-10 (recusal only).

Videos, advertisements and films

The point of this section is to let you watch the resemblance for yourself: the friendly, no-questions-asked pitch to people with damaged credit sounds much the same in a 2005 mortgage commercial and a 2025 “bad credit, no problem” car ad. Please read the label on each item. Only a handful of these pages were opened, YouTube blocked my automated fetches, and several items are unofficial uploads or tangential, so this is a viewing list to be checked, not a verified archive. Films other than the usual financial-crisis titles are included where I found them.

Advertising from the subprime mortgage boom (8)

Subprime and buy-here-pay-here auto advertising (16)

News segments and investigations (12)

Hearings (4)

Documentaries (14)

Series episodes (2)

Feature films (1)

Notes on the media list

  • WebFetch on YouTube returned HTTP 429 (rate limited) and instructed not to retry; all YouTube items are search-result-only.
  • Bloomberg (402), C-SPAN and NPR (robots.txt) blocked fetching; not circumvented.
  • Search tool is US-only; no URLs were constructed.

What is missing

These are the things I looked for and could not find or verify. Any of them may exist.

  • Filed consent judgment courts and docket numbers for most of the 41 states in the Credit Acceptance settlement; NY federal consent judgment.
  • Santander 2020 per-state consent judgment courts/dockets; Santander CFPB $4.75M penalty matter seen in search results not reviewed.
  • Any Deka motion-to-dismiss or final approval order.
  • Docket number for U.S. v. Collins (N.D. Ill.) and SEC Tricolor case number.
  • No Flagship or America's Car-Mart government action found; no GM Financial DOJ subpoena resolution found; no state AG action against Carvana found (only Secretary of State/DMV administrative actions).
  • Few reported appellate opinions on MLA, deficiency, UDAP for subprime auto beyond Davidson, Henson, Bell, Rivera.
  • Current status of Chu trial after Sept 2026.
  • Not covered: Colorado or other state true-lender settlements (e.g., Avant/Marlette), Think Finance bankruptcy and CFPB v. Think Finance, CFPB rent-a-bank litigation.
  • No DOJ Military Lending Act cases beyond the CFPB actions; the SCRA settlements surfaced in search but were not opened.
  • No state AG title-loan court rulings found that could be verified; search results were consumer journalism.
  • Recent (2024-2026) subprime credit-card enforcement is thin: only Apple/Goldman (2024) and Discover (2025) verified; nothing on First Premier, Credit One or Mission Lane, and the fate of the CFPB late-fee rule litigation was not confirmed.
  • 18 U.S.C. 891-894: only Perez (S. Ct.) was found as a Supreme Court opinion; other circuit opinions (Martorano 1st Cir., Pacione, 3d Cir. 151892p) appeared only as search results. Recent loan-sharking cases verified are mostly organized-crime prosecutions; no consumer-lender prosecution under section 894 was found.
  • Docket numbers are missing for most district-court criminal matters (S.D.N.Y. Tucker/Muir, E.D. Pa. Hallinan/Neff, E.D. Pa. Gjeli, E.D.N.Y. Ragano and Semplice, S.D.N.Y. Pietranico); CourtListener/PACER lookups needed.
  • No buy-here-pay-here or state AG debt-collection cases; Hallinan/Neff judge names not confirmed; the Ninth Circuit FTC v. AMG opinion (910 F.3d 417) was not opened.
  • Household/Ameriquest/Countrywide: no state court filings or consent judgments were opened, only press releases.
  • No genuine Countrywide, Household/Beneficial, Fremont, WMC, New Century or Option One TV ad found.
  • No Frontline 'Car Loan Trap' episode found; it may not exist.
  • No Full Frontal, 60 Minutes, Vice or NYT subprime auto video found.
  • No Senate/House hearing video specifically on subprime auto found.
  • No Carvana bad-credit ad, Credit Acceptance/Santander/Exeter/GM Financial ads found; The Card Counter, Generation Wealth legitimacy not checked.
  • Total is under 50 items and mostly unopened.
Nothing on this page is legal or financial advice. Summaries were compiled from fetched web pages on September 28, 2026 and may contain errors; check every holding, date and figure against the linked primary record before relying on it or quoting it.
Addendum: Who Lends. Companies in Subprime Auto Finance

Addendum: who lends

What this is and is not. This is the most complete sourced list I could assemble in one session. It is not exhaustive, and I would not call it verified. The only lender-level ranking of subprime originations I could find is Experian’s from the fourth quarter of 2016, which is ten years old, so the market shares below show who mattered then, not now. Many headquarters and parent-company fields are marked as unverified. A company appearing here is not an accusation of wrongdoing: the list includes banks and credit unions whose subprime exposure I could not confirm, and it separates those in the “unconfirmed” list.

An essay on who is in the room

The popular picture of subprime auto lending is a sketchy independent finance company that writes 22 percent loans to people the banks turned away. That picture is right about part of the market and wrong about most of it. Experian’s fourth-quarter 2025 lender-type shares of all vehicle financing put banks at 29.29 percent, captive finance arms of the automakers at 27.55 percent and credit unions at 19.56 percent, which means three sectors most people think of as respectable account for about three-quarters of the market before a single finance company is counted. Equifax reported in February 2024 that roughly 16 percent of credit union auto balances were subprime, so the not-for-profit sector is in this business too, at scale and with lower rates. Those figures are as reported in the sources listed under the directory, and they describe all vehicle financing by lender type, not subprime alone.

The independents are where the highest rates, the sharpest collections and most of the enforcement history sit. In Experian’s 2016 list, Santander Consumer USA held 16.8 percent of subprime originations and Credit Acceptance 10.6 percent, followed by World Omni, Gateway One, Westlake and AmeriCredit, the last of which is now GM Financial. The captives lend to riskier buyers for a reason that has little to do with charity: a finance arm that can approve a buyer with a 580 score sells a car the bank would not have financed. The dealers stand in front of all of it, marking up the rate and packing add-ons, and the loans then travel to the bond market.

That last step is the part that makes the system durable. Underwriters such as the banks named below package the loans into asset-backed securities that pay investors far less than the borrower pays the lender, and the difference funds the servicers, the repossession vendors and the profit. Because investors, underwriters, rating agencies and servicers all earn fees on volume, few participants have a reason to slow the machine until it visibly breaks, which is what Tricolor did in 2025. Twenty million Americans are in the subprime segment because a whole industry, not one villain, has a use for them. The directory that follows lists the participants I could source, by role.

Rankings and market-share sources

Top 20 finance companies by subprime originations market share (Experian) (Q4 2016)

Source

  1. 1 Santander Consumer USA 16.8%
  2. 2 Credit Acceptance 10.6%
  3. 3 World Omni Financial 6.8%
  4. 4 Gateway One Lending & Finance 6.8%
  5. 5 Westlake Financial Services 6.4%
  6. 6 AmeriCredit Financial Services 4.8%
  7. 7 Exeter Finance 2.6%
  8. 8 Regional Acceptance 2.5%
  9. 9 American Credit Acceptance 2.3%
  10. 10 Flagship Credit Acceptance 2.2%
  11. 11 Consumer Portfolio Services 2.1%
  12. 12 Prestige Financial Services 1.1%
  13. 13 Lobel Financial 1.0%
  14. 14 Lincoln Automotive Financial Services 0.9%
  15. 15 United Auto Credit 0.7%
  16. 16 Financial Institution Lending Option 0.7%
  17. 17 Reliable Credit Association 0.6%
  18. 18 Global Lending Services 0.6%
  19. 19 First Investors Financial Services Group 0.6%
  20. 20 Nicholas Financial 0.6%

Experian: share of all vehicle financing by lender type and subprime share (Q4 2025)

Source

  1. Subprime 15.31% of financing (14.54% Q4 2024); new 6.61%; used 22.47%
  2. Banks 29.29%
  3. Captives 27.55%
  4. Credit unions 19.56%

Experian: lender-type share of all financing (Q3 2025)

Source

  1. Banks 28.9% (+310 bps)
  2. Captives 26.2% (-262 bps)
  3. Used: banks 29.7%, credit unions 28%
  4. New: captives 52.9%, banks 27.2%

IBISWorld-cited largest subprime auto lenders (low-quality secondary summary, no shares) (2026)

Source

  1. Santander Consumer USA
  2. Credit Acceptance
  3. Toyota Financial Services

Directory by role

Independent subprime finance companies

Santander Consumer USA (Santander Drive / Chrysler Capital)

Parent
Santander Holdings USA (SHUSA) / Banco Santander; current ownership structure not verified in this pass
Headquarters
Dallas, TX (Consumer Reports: Texas; Dallas per directory context not opened)
Role in subprime
Long described as the largest US subprime auto financer; Drive Auto Receivables (SDART) shelf; Chrysler Capital retail program; Nissan SignatureAccess pass-through partner.
Key figures
Rank #1 in Experian Q4 2016 subprime list at 16.8%. SDART 2026-1: $1.585B offered notes, substantially all sub-prime, WA FICO ~604, WA APR ~18.25%, lead underwriters Citigroup, RBC, Santander, Wells Fargo Securities.
Enforcement notes
$550M multistate AG settlement (2020); CFPB $4.75M settlement (per AFN title); Mississippi AG settlement July 2021 (Consumer Reports); repossessed 84,249 vehicles in 2021 per CR.

Credit Acceptance Corporation

Parent
Public (NASDAQ: CACC)
Headquarters
Southfield, MI
Role in subprime
Dealer-lender model buying/advancing on subprime loans; ABS issuer.
Key figures
Rank #2 in Experian Q4 2016 subprime list at 10.6%. Reported average APR >38% in AG settlement.
Enforcement notes
$700M settlement with 40 jurisdictions announced Sept 17 2026 (NY AG; $630M+ debt relief, $60M restitution, $15.5M penalties); CFPB/NY AG suit filed Jan 2023, CFPB dropped 2025; earlier $550M with 34 AGs (2020) and MA settlement 2021. NOTE: Hoodline reports $694M/41 states - discrepancy with NY AG release.

Westlake Financial Services

Parent
Westlake Services LLC (privately held)
Headquarters
Los Angeles, CA
Role in subprime
Large private subprime auto lender and ABS issuer, focus on used-vehicle and Hispanic-market lending (focus claim not verified here).
Key figures
Rank #5 in Experian Q4 2016 list at 6.4%; multiple consecutive ABS offerings above $1B (AR); named by S&P (June 2026) as driving subprime/nonprime ABS growth.
Enforcement notes
DOJ SCRA settlement, Sept 28 2022, >$225,000.

Exeter Finance

Parent
Warburg Pincus (acquired 2021 from Blackstone; reportedly exploring $2.5-3B sale/IPO)
Headquarters
Irving, TX
Role in subprime
Subprime/nonprime retail installment lender via ~15,000 dealers; EART ABS shelf.
Key figures
Rank #7 in Experian Q4 2016 at 2.6%. ~650,000 customers (PEW); ProPublica: >500,000 active loans worth ~$10B. ESART 2026-1 (an Exeter-associated shelf) joint bookrunners Deutsche Bank (structuring), Citi, Mizuho; co-managers AmeriVet, Barclays, Citizens.
Enforcement notes
MA and DE AG settlements totaling ~$6M (2019); ProPublica reports GA and LA investigations; ex-Santander executives lead firm (ProPublica).

Consumer Portfolio Services (CPS)

Parent
Public (NASDAQ: CPSS)
Headquarters
Principal offices Las Vegas, NV; operations Irvine, CA (10-K)
Role in subprime
Subprime auto contract purchaser/servicer; ABS shelf name not verified.
Key figures
2025 contract purchases ~$1.64B; 4 securitizations totaling ~$1.73B in 2025; 107 term securitizations of ~$22.4B since 1994; rank #11 in Q4 2016 at 2.1%.
Enforcement notes
10-K legal proceedings section not read in detail.

American Credit Acceptance (ACA)

Parent
Parent not verified in this pass
Headquarters
Spartanburg, SC
Role in subprime
Subprime auto lender; American Credit Acceptance Receivables Trust ABS rated by KBRA.
Key figures
Rank #9 in Experian Q4 2016 at 2.3%.
Enforcement notes
None found in this pass.

Global Lending Services (GLS)

Parent
Parent not verified in this pass
Headquarters
Greenville, SC (also Atlanta, Phoenix per directory)
Role in subprime
Deep-subprime auto lender; Nissan SignatureAccess partner; GLS Auto Receivables Issuer Trust ABS.
Key figures
Rank #18 in Q4 2016 at 0.6%; ABS deals ~$666.5M (ASR title).
Enforcement notes
None found.

First Investors Financial Services

Parent
Stellantis (acquired for $285M, announced Sept 2021) - now captive-style arm
Headquarters
Houston, TX
Role in subprime
Subprime/nonprime auto lender turned Stellantis captive; First Investors Auto Owner Trust ABS still rated by KBRA (2026-2).
Key figures
Rank #19 in Q4 2016 at 0.6%; $1.17B managed portfolio Oct 2015.
Enforcement notes
None found.

Flagship Credit Acceptance

Parent
Perella Weinberg-affiliated PE (per 2015-16 slideshow); current owner not verified in this pass
Headquarters
HQ not verified in this pass
Role in subprime
Subprime auto lender; sold to dealers incl. military program.
Key figures
Rank #10 in Q4 2016 at 2.2%; $2.7B total assets then.
Enforcement notes
None found.

Tricolor Holdings

Parent
Founder Daniel Chu (charged); Chapter 7 filed Sept 2025
Headquarters
Texas (Dallas HQ not verified in this pass)
Role in subprime
Deep-subprime auto retailer/lender serving Hispanic borrowers ('underserved communities'); Tricolor Auto Securitization Trust ABS; collapsed on alleged double-pledging.
Key figures
~$2.2B receivables implicated; prosecutors allege >$800M bogus. Past ABS: $328M (Mar 2025), $217M (June 2025).
Enforcement notes
Executives charged with fraud (Dec 2025, CNBC blocked - search-result-only); investor suits vs JPMorgan/Barclays/Fifth Third (Mar 2026) and vs Wilmington Trust and Vervent.

Honor Finance

Parent
Former principals James Collins and Robert DiMeo
Headquarters
HQ not verified in this pass
Role in subprime
Subprime auto lender; Honor Automobile Trust Securitization 2016-1 ($100M).
Key figures
$100M ABS at issue.
Enforcement notes
SEC complaint Sept 23 2021 alleging misleading investors (ineligible loans, unconsented extensions, forgiven payments).

Southern Auto Finance Company (SAFCO)

Parent
Parent not verified in this pass
Headquarters
HQ not verified in this pass
Role in subprime
Subprime auto finance company listed in the trade directory (role only as listed).
Key figures
Listed in 2025 SubPrime Auto Finance News directory; no volume figure.
Enforcement notes
None found.

Vehicle Acceptance Corporation

Parent
Parent not verified in this pass
Headquarters
HQ not verified in this pass
Role in subprime
Subprime auto finance company listed in the trade directory (role only as listed).
Key figures
Listed in 2025 SubPrime Auto Finance News directory; no volume figure.
Enforcement notes
None found.

Skopos Financial

Parent
Parent not verified in this pass
Headquarters
HQ not verified in this pass
Role in subprime
Subprime auto finance company listed in the trade directory (role only as listed).
Key figures
Listed in 2025 SubPrime Auto Finance News directory; no volume figure. $340M outstanding balance Oct 2015 (American Banker slideshow).
Enforcement notes
None found.

PrimaLend Capital Partners

Parent
Parent not verified in this pass
Headquarters
Texas
Role in subprime
Dealer-floorplan/receivables lender to subprime buy-here-pay-here dealers; Chapter 11 Oct 2025.
Key figures
Banks owed at least $161.7M.
Enforcement notes
Bankruptcy Oct 2025.

Captive finance arms of automakers

GM Financial (incl. AmeriCredit Financial Services)

Parent
General Motors
Headquarters
Fort Worth, TX (not opened)
Role in subprime
Captive with historic subprime arm (AmeriCredit); GM Financial/Santander described as moving toward prime in 2015.
Key figures
AmeriCredit rank #6 in Q4 2016 at 4.8%.
Enforcement notes
DOJ subpoena over subprime auto loan securities (Aug 2014).

Ford Credit / Lincoln Automotive Financial Services

Parent
Ford Motor Company
Headquarters
Dearborn, MI (not opened)
Role in subprime
Lincoln Automotive Financial Services appears in Experian Q4 2016 subprime list; Ford Credit appears in ABS data set of below-prime loans (CUCollector).
Key figures
Lincoln Auto Fin rank #14 at 0.9% (2016).
Enforcement notes
None found.

Toyota Motor Credit Corp.

Parent
Toyota Motor Corp.
Headquarters
Plano, TX (not opened)
Role in subprime
Captive named by IBISWorld/Review for Loans among market participants (weak source).
Key figures
Subprime share not verified in this pass.
Enforcement notes
CFPB $60M action Nov 20 2023 (add-on products, credit reporting); earlier CFPB/DOJ ECOA resolution.

Nissan Motor Acceptance Company (NMAC)

Parent
Nissan Motor Co.
Headquarters
Franklin, TN (not opened)
Role in subprime
Launched SignatureAccess sub-prime dual pass-through program March 16 2021 routing declined applicants to Santander Consumer and GLS.
Key figures
Target FICO under 620.
Enforcement notes
None found.

Hyundai Capital America

Parent
Hyundai Motor / Hyundai Capital
Headquarters
Irvine, CA (not opened)
Role in subprime
Captive; subprime share not verified in this pass.
Key figures
No figure.
Enforcement notes
CFPB July 26 2022: $13.2M redress + $6M penalty (FCRA furnishing).

Stellantis Financial Services US / Chrysler Capital

Parent
Stellantis N.V. (plus Santander partnership)
Headquarters
Detroit / Dallas (not opened)
Role in subprime
Stellantis extended Santander contract in 2022 and bought First Investors to add captive lending.
Key figures
First Investors purchase $285M.
Enforcement notes
None found.

Banks

Capital One Auto Finance

Parent
Capital One Financial
Headquarters
McLean, VA (not opened)
Role in subprime
Began auto lending 1998 via Summit Acceptance (subprime); in 2014-15 third-largest US auto lender and #2 in used; services Carvana/DriveTime-type subprime loans per interview snippet.
Key figures
Third-largest auto lender behind Wells Fargo and Ally (2014-15).
Enforcement notes
NY district attorney subpoena Nov 2014.

Ally Financial

Parent
Ally Financial Inc.
Headquarters
Detroit, MI (not opened)
Role in subprime
Large auto lender with nonprime segment; forward-flow agreement with Carvana extended/increased Oct 2025.
Key figures
1.734M retail contracts originated Dec 31 2015 (slideshow); ~12% subprime per IBISWorld via Review for Loans (weak source). Carvana originations $3.4B Q3 2025.
Enforcement notes
Subpoena 2014-15 era (Yahoo).

Wells Fargo Auto Finance

Parent
Wells Fargo & Co.
Headquarters
San Francisco, CA (not opened)
Role in subprime
Once largest auto lender; trimmed originations 2017; VW/Audi captive-like deal cited by Experian analysts as bank growth driver in 2025.
Key figures
Listed in subprime directory 2025.
Enforcement notes
None found in this pass.

Chase Auto Finance (JPMorgan Chase)

Parent
JPMorgan Chase & Co.
Headquarters
New York, NY (not opened)
Role in subprime
Auto lender; Tesla partnership per AFN; also Tricolor warehouse lender and ABS underwriter.
Key figures
Q3 2025 $170M charge-off tied to Tricolor (Banking Dive).
Enforcement notes
Defendant in March 2026 investor suit over Tricolor.

PNC Bank

Parent
PNC Financial Services
Headquarters
Pittsburgh, PA (not opened)
Role in subprime
Scaled back auto finance business (AFN title); subprime exposure not verified in this pass.
Key figures
No figure.
Enforcement notes
None found.

Mechanics Bank

Parent
Parent not verified in this pass
Headquarters
HQ not verified in this pass
Role in subprime
Exited indirect auto lending Feb 2023.
Key figures
No figure.
Enforcement notes
None found.

Santander Bank / SHUSA

Parent
Banco Santander
Headquarters
Boston/New York (not opened)
Role in subprime
Parent group of Santander Consumer; publishes quarterly fixed-income decks with auto segment data (not read).
Key figures
See Santander Consumer entry.
Enforcement notes
See Santander Consumer entry.

Credit unions

TTCU Federal Credit Union

Parent
Member-owned
Headquarters
HQ: Oklahoma
Role in subprime
Runs a second-chance / credit-rebuilder auto program for low or no-score borrowers.
Key figures
Target FICO under 640 or no score (America's Credit Unions blog).
Enforcement notes
None found.

Peach State Federal Credit Union

Parent
Member-owned
Headquarters
HQ: HQ not verified in this pass
Role in subprime
Runs a second-chance / credit-rebuilder auto program for low or no-score borrowers.
Key figures
Target FICO under 640 or no score (America's Credit Unions blog).
Enforcement notes
None found.

Valley Oak Credit Union

Parent
Member-owned
Headquarters
HQ: HQ not verified in this pass
Role in subprime
Runs a second-chance / credit-rebuilder auto program for low or no-score borrowers.
Key figures
Target FICO under 640 or no score (America's Credit Unions blog).
Enforcement notes
None found.

WyHy Federal Credit Union

Parent
Member-owned
Headquarters
HQ: Wyoming
Role in subprime
Runs a second-chance / credit-rebuilder auto program for low or no-score borrowers.
Key figures
Target FICO under 640 or no score (America's Credit Unions blog).
Enforcement notes
None found.

SCU Credit Union

Parent
Member-owned
Headquarters
HQ: Massachusetts (per source)
Role in subprime
Runs a second-chance / credit-rebuilder auto program for low or no-score borrowers.
Key figures
Target FICO under 640 or no score (America's Credit Unions blog).
Enforcement notes
None found.

Alive Credit Union

Parent
Member-owned
Headquarters
HQ: Florida
Role in subprime
Runs a second-chance / credit-rebuilder auto program for low or no-score borrowers.
Key figures
Target FICO under 640 or no score (America's Credit Unions blog).
Enforcement notes
None found.

Camino Federal Credit Union (Montebello)

Parent
Member-owned
Headquarters
HQ: CA
Role in subprime
Runs a second-chance / credit-rebuilder auto program for low or no-score borrowers.
Key figures
Target FICO under 640 or no score (America's Credit Unions blog).
Enforcement notes
None found.

Credit Union of Southern California

Parent
Member-owned
Headquarters
HQ: CA
Role in subprime
Runs a second-chance / credit-rebuilder auto program for low or no-score borrowers.
Key figures
Target FICO under 640 or no score (America's Credit Unions blog).
Enforcement notes
None found.

SnoCope Credit Union

Parent
Member-owned
Headquarters
HQ: WA (not verified)
Role in subprime
Runs a second-chance / credit-rebuilder auto program for low or no-score borrowers.
Key figures
Target FICO under 640 or no score (America's Credit Unions blog).
Enforcement notes
None found.

Credit unions (sector aggregate)

Parent
NCUA-chartered
Headquarters
n/a
Role in subprime
Sector share data, not one lender.
Key figures
Equifax: ~16% of CU auto balances subprime (Feb 2024), CU subprime +13% YoY vs banks +8.1%, captives +3.9%. Experian Q4 2025: CU 19.56% of total financing; used-vehicle CU share 28% (Q3 2025).
Enforcement notes
n/a

Retailers and buy-here-pay-here operators

Carvana

Parent
Public (NASDAQ: CVNA); founder family Garcia also controls DriveTime
Headquarters
Tempe, AZ (not opened)
Role in subprime
Online used-car retailer; sells loans via ABS (Carvana Auto Receivables Trust) and forward-flow to Ally; Bridgecrest/Capital One serviced loans.
Key figures
CART 2024-P3 pool: avg score 695, 44% subprime; Q3 2025 originations $3.4B.
Enforcement notes
CT AG $1.5M settlement Jan 14 2025 (titles/registration, not lending); Hindenburg Research report on DriveTime link.

America's Car-Mart

Parent
Public (NASDAQ: CRMT)
Headquarters
Arkansas (Rogers founded; HQ city not verified in this pass)
Role in subprime
BHPH dealer-lender; ~94 stores after closing 42; finance receivables >$1.5B in 2025; ~$2.5B ABS borrowings 2022-25.
Key figures
Half of loans modified; defaulted on Silver Point $300M loan June 2026; seeking rescue financing (July 2026).
Enforcement notes
None found.

Fintech and marketplace lenders

Lendbuzz

Parent
Public/IPO filed Sept 12 2025 (per AFN)
Headquarters
Boston, MA
Role in subprime
AI underwriting for thin/no-file and subprime auto borrowers; ABS issuer.
Key figures
$1.1B originations (AFN).
Enforcement notes
None found.

AutoUSA

Parent
Parent not verified in this pass
Headquarters
HQ not verified in this pass
Role in subprime
Lead generator that entered the subprime lead business (AR title).
Key figures
No figure.
Enforcement notes
None found.

Underwriters and investors in auto securitizations

Barclays

Parent
Barclays PLC
Headquarters
London / New York (not opened)
Role in subprime
Tricolor warehouse lender and ABS underwriter; co-manager on 2026-1 Exeter-linked ESART deal.
Key figures
$150M loss Q3 2025 tied to Tricolor. Banks collectively advanced >$1B to Tricolor 2020-2025 (Octus).
Enforcement notes
Defendant in Tricolor investor suit.

Citigroup

Parent
Parent not verified in this pass
Headquarters
HQ not verified in this pass
Role in subprime
Joint bookrunner ESART 2026-1; lead underwriter SDART 2026-1
Key figures
Deal-level role only; no league table rank.
Enforcement notes
None found.

RBC Capital Markets

Parent
Parent not verified in this pass
Headquarters
HQ not verified in this pass
Role in subprime
Lead underwriter SDART 2026-1
Key figures
Deal-level role only; no league table rank.
Enforcement notes
None found.

Wells Fargo Securities

Parent
Parent not verified in this pass
Headquarters
HQ not verified in this pass
Role in subprime
Lead underwriter SDART 2026-1
Key figures
Deal-level role only; no league table rank.
Enforcement notes
None found.

Santander US Capital Markets

Parent
Parent not verified in this pass
Headquarters
HQ not verified in this pass
Role in subprime
Lead underwriter on SDART 2026-1 (affiliate of sponsor)
Key figures
Deal-level role only; no league table rank.
Enforcement notes
None found.

CastleOak Securities

Parent
Parent not verified in this pass
Headquarters
HQ not verified in this pass
Role in subprime
Co-manager SDART 2026-1 Class A
Key figures
Deal-level role only; no league table rank.
Enforcement notes
None found.

Ramirez & Co.

Parent
Parent not verified in this pass
Headquarters
HQ not verified in this pass
Role in subprime
Co-manager SDART 2026-1 Class A
Key figures
Deal-level role only; no league table rank.
Enforcement notes
None found.

Silver Point Capital

Parent
Private
Headquarters
Greenwich, CT (not opened)
Role in subprime
Private credit lender to America's Car-Mart ($300M five-year loan; SOFR+7.50% with warrants).
Key figures
$300M.
Enforcement notes
None found.

Castlelake

Parent
Private
Headquarters
HQ not verified in this pass
Role in subprime
Investor buying auto loans from Pagaya (up to $500M, search-result title).
Key figures
$500M.
Enforcement notes
None found.

Servicers, technology and repossession vendors

Bridgecrest (Bridgecrest Acceptance Corp / DT Acceptance)

Parent
DriveTime Automotive Group (Ernest Garcia II ~75%)
Headquarters
Mesa, AZ
Role in subprime
Subprime lender/servicer for DriveTime and Carvana loans; Bridgecrest Lending Auto Securitization Trust and DT Auto Owner Trust ABS.
Key figures
Serviced ~$21.4B receivables late 2024 (only $5.8B owned by parent); avg APR ~22% on recent securitizations.
Enforcement notes
CFPB $8M penalty Nov 2014 (LegalClarity).

PassTime GPS

Parent
Parent not verified in this pass
Headquarters
Littleton, CO
Role in subprime
Maker of GPS trackers and starter-interrupt devices for subprime dealers/lenders/credit unions (Encore, TRAX, Elite, Select).
Key figures
No volume figure.
Enforcement notes
None found.

Rating agencies

KBRA

Parent
Kroll
Headquarters
New York, NY (not opened)
Role in subprime
Rates Bridgecrest, GLS, American Credit Acceptance, First Investors, Carvana, and rated 7 Tricolor deals (downgraded to default Dec 9 2025).
Key figures
7 Tricolor deals.
Enforcement notes
Reputational scrutiny after Tricolor.

S&P Global Ratings

Parent
S&P Global
Headquarters
New York, NY (not opened)
Role in subprime
Rates Exeter, Santander, Carvana, CarMax subprime/nonprime deals; publishes ABS tracker; rated other Tricolor deals.
Key figures
n/a
Enforcement notes
None found.

Unconfirmed: names that came up but could not be tied to subprime auto lending

  • Navy Federal CU, PenFed CU, VyStar CU: only product pages surfaced; subprime program not verified
  • Bank of America, U.S. Bank, TD Bank, Truist: no sourced subprime exposure found
  • Regional Management Corp (consumer installment lender; auto role not verified; distinct from Regional Acceptance)
  • Kia/Kia Finance, Volkswagen Credit, BMW, Mercedes, VW Credit: no subprime evidence
  • AutoNation Finance, Sonic/EchoPark, Group 1: subprime role not confirmed
  • Open Lending, Upstart, myAutoloan: search hits only; auto-subprime role not verified
  • Spireon, Ituran, other GPS device makers: not verified
  • DBRS Morningstar: rating role in subprime auto not verified (only newswire pages surfaced)
  • Hindenburg Research report on Carvana/DriveTime (allegations, not verified)

Notes and gaps

  • Web fetch summaries were produced by a small model; figures, dates and 2026 events are as reported by those fetches and should be spot-checked.
  • Discrepancy: Credit Acceptance settlement reported as $700M/40 jurisdictions by NY AG vs $694M/41 states by Hoodline.
  • Discrepancy: Tricolor investors' holdings ~$230M (Banking Dive) vs ~$270M subordinated notes (Octus).
  • Experian Q4 2016 ranking is old; no more recent lender-level subprime ranking was found.
  • ESART 2026-1 was described by the fetch summarizer as 'near-prime'; tie to Exeter's subprime shelf not confirmed.
  • No current (2023-2026) lender-level subprime origination ranking found
  • No ABS underwriter league table content retrievable (asreport pages use Flourish charts)
  • Full Experian/Equifax/Philadelphia Fed lender-type cross-tabs for subprime not retrieved
  • Most HQ/parent fields marked not verified
  • Blocked: cnbc.com (403), doj.nh.gov (403); paywalled Auto Finance News and Asset Securitization Report bodies
Nothing on this page is legal or financial advice. Summaries were compiled from fetched web pages on September 28, 2026 and may contain errors; check every holding, date and figure against the linked primary record before relying on it or quoting it.
Previous
Previous

The Central Contradiction.

Next
Next

I think it was the eye!"