By Bassem Banafa, Aaron Greenspan and Christine Richard
Credit Acceptance Corporation (NASDAQ: CACC) is a Michigan-based lender and servicer of deep subprime auto loans. Its motto, "We change lives," reflects the second chance Credit Acceptance says it gives to consumers whose credit histories make it difficult for them to get credit elsewhere. Credit Acceptance reports that it originated over 300,000 loans in 2016 and over 250,000 loans in 2017 through Q3.
Wall Street has embraced Credit Acceptance's business model as a unique way to profit from the extraordinarily high finance charges and fees imposed on subprime borrowers while offloading much of the loan performance risk to dealers. The company's shares have gained more than 2,000% over the last 10 years, outpacing Apple and Google. Its CEO is among the highest-paid executives in the nation, having agreed to a mostly-stock pay package that today may be worth as much as $114 million.
Credit Acceptance's loan performance disclosures belie a deeply flawed business model. In Detroit, Michigan thus far in 2017, nearly one in eight of all available civil lawsuits filed in the city involve Credit Acceptance suing borrowers. Overall, 72% of these lawsuits resulted in the company garnishing the wages or tax refunds of borrowers. In essence, Credit Acceptance is a new kind of hybrid: a debt collector that originates its own loans—a combination that has proved extraordinarily profitable for investors as the business of lending to troubled borrowers has surged since the financial crisis. The risks of that model have yet to be reckoned with. These include:
Part I of this Reality Check examines the regulatory risks facing Credit Acceptance Corporation. Part II examines financial risks, as even in the absence of a regulatory crackdown, the company relies upon a fragile and complex financing structure.
Credit Acceptance Corporation is America's second-largest lender for deep subprime borrowers: those with a bad credit history or no credit history.1 According to the company's Q2 2017 Form 10-Q, 95% of loans had a FICO score of 650 or below or no score at all.2 The company, formerly known as Auto Finance Corporation, was incorporated on August 23, 1972 by Donald A. Foss, a used car salesman who built an empire of used car lots after starting with a single dealership in Detroit. The company's initial stated purpose was to "[lend] money to purchase used automobiles" and to "buy, sell and generally deal with commercial paper"—to borrow from capital markets and lend those funds to buyers of used vehicles.
Today, Credit Acceptance has a market capitalization of over $6 billion. On November 1, 2008, its stock closed at $13.10 per share. Nine years later, it trades at over $286, a 2,083% increase. By comparison, Apple, Inc. stock was worth $13.24 on November 1, 2008. Nine years later, it is worth a mere $170.15, having increased by only 1,185%. Whatever business Credit Acceptance is in, it has been growing at nearly twice the rate of the most valuable company on Earth, and the company responsible for the iPhone, the most profitable technology product in history.3
Foss, "the world's richest used car salesman,"4 sold some of his shares in Credit Acceptance in 2017 for a total of $128 million shortly after his retirement from the company. According to Bloomberg, "He's still the largest shareholder in Credit Acceptance with a $700 million stake," approximately half of his $1.3 billion net worth.5
Perceived Market Advantages
Solid Earnings History
Credit Acceptance has enjoyed an industry reputation as a solid player with "an outstanding track record in a very difficult business."6 As recently as November of 2016, Zacks Investment Research called CACC, "an intriguing choice for investors right now" owing to "solid earnings estimate revision activity over the past month, suggesting analysts are becoming a bit more bullish on the firm's prospects in both the short and long term."7,8 As of the writing of this report, Credit Acceptance does have a solid price-to-earnings ratio of 16.03, which is still lower than that of its much smaller competitor, Nicholas Financial, Inc. (NASDAQ: NICK).9 The company has also beat analysts' earnings forecasts in every quarter of 2017 thus far. It may therefore be tempting to view Credit Acceptance as a value play, even with its stock at stratospheric levels relative to its pre-2008 average price, which hovered below $10 per share.
An Underwriting System That Creates Smarter Loans
When a typical buyer purchases a used car from a dealer that works with Credit Acceptance, the loan is structured through the company's database driven website: the Credit Approval Processing System (CAPS). Credit Acceptance charges dealers $599 per month to access CAPS.10
CAPS allows dealers to upload a record of all their inventory and receive feedback from Credit Acceptance on potential profits across an entire lot. For each deal, a dealer can input information about a borrower (monthly income, co-signer information, bank data, etc.) and deal structure (down payment, interest rate, and loan term) and receive a response within 30 seconds with finalized profit/loss information for every vehicle available. Credit Acceptance markets CAPS to dealers as offering "Guaranteed Credit" and describes its system as "[t]he most powerful selling tool you can offer!" CAPS is integral to Credit Acceptance's claim that it approves "everyone."11
Dealers pay a $9,850 enrollment fee to begin using CAPS. Alternatively, a dealer can avoid the up-front fee by "agreeing to allow [Credit Acceptance] to retain 50% of their first accelerated Dealer Holdback payment."12,13 Either way, Credit Acceptance deducts the additional $599 CAPS monthly fee from dealer profits.
A Growing Dealer Network
Over time, the company has been able to sign up more and more car dealers nationwide. In Q4 2002, the company originated loans from 555 active dealers. During Q3 2017, Credit Acceptance originated loans from 7,737 active dealers.14
We were moved to ask a fundamental question about Credit Acceptance: is it a lender, or a debt collector? In other industries, these roles have been separated into different roles and companies by market forces. Lenders want to loan to high-quality borrowers so that they can collect on their loans, ideally in full plus interest, and are reluctant to settle for pennies on the dollar. On the flip side, debt collection agencies, who end up purchasing bad loans from lenders for such steeply reduced prices, are driven solely by the opportunity to collect: the cheaper the debt (and thus, the lower the average borrower quality), the greater the opportunity.
While publicly emphasizing its lending role, Credit Acceptance marries both sides of this equation into one entity. Yet as was demonstrated when Glass-Steagall came tumbling down, hybrid financial creatures born of risk-averse and aggressively risk-loving components are immediately dominated by their risk-loving side. In other words, Credit Acceptance is a debt collector on steroids, and thinks of itself as such.
Clause 2.02(a) of its Dealer Servicing Agreement states, "If Credit Acceptance accepts assignment of a Contract for administration, servicing and collection, Credit Acceptance's duties shall consist of holding the Contract Files, collecting payments due under the Contracts as set forth in subsection (b) of this Section 2.02 and applying the amounts collected..." Clause 2.02(b) continues, "Credit Acceptance shall use reasonable efforts to collect all payments called for under the terms and conditions of the Contracts as and when the same shall become due," after which the Agreement discusses Credit Acceptance's option to repossess vehicles. The market hype that it has found a magical lending formula or a way to ensure more accurate loan forecasting is just that: hype.15 The most important question about Credit Acceptance Corporation is how well its loans actually perform. It is a question that the company has gone to great lengths to avoid answering.16 In the words of Don Foss, "Well, you know, when you go public, the one thing I didn't realize is that you actually tell everybody what you do."17
How Credit Acceptance's "Portfolio Program" Works
The Credit Acceptance program is poorly understood according to one dealer we spoke with. "Very few general managers at the store level understand the program. There's no one dealer who understands it in its entirety."
Historically through 2015, about 90% of Credit Acceptance's reported unit volume has come from its "Portfolio Program," in which Credit Acceptance loans money to its "Dealer-Partners"—commonly called used car dealerships—in exchange for assignment of the loan and the right to service the loan, which involves a 20% servicing fee. The remaining small but increasing percentage of the company's business comes from buying loans that dealers have already made, which Credit Acceptance refers to as its "Purchase Program."18
The Credit Acceptance formula is so attractive to dealers because they receive an up-front advance from Credit Acceptance for each sale, regardless of whether the borrower pays back their loan. The dealer is also eligible to receive money from later loan payments, which Credit Acceptance calls "dealer holdback," though these payments are contingent on the dealer closing a group of 100 loans, called a "pool," and that dealer's loans—as a group—performing up to a certain standard.
Not every dealer is able to achieve the formation of a pool. While Credit Acceptance's dealer network has grown over the years, the number of loans each dealer produces has declined on average.19 In the future, the company may have trouble locating additional car dealerships interested in signing up as it gets closer to achieving market saturation.
Dealers are to some degree caught between the promise of easy money in the form of the advance and the restrictions set in place by the federal Truth In Lending Act.20 The Act is a powerful regulatory tool because violations trigger specific monetary penalties, outlined in § 1640. Sub-section (a)(2)(A)(iii) is of particular relevance, explaining that damages are "in the case of an individual action relating to an open end consumer credit plan that is not secured by real property or a dwelling, twice the amount of any finance charge in connection with the transaction, with a minimum of $500 and a maximum of $5,000, or such higher amount as may be appropriate in the case of an established pattern or practice of such failures." State laws may also have separate penalties. As explained below, these statutes and the market pressures of the subprime auto industry are often at odds.
Price Increases During the Underwriting Process
The process of underwriting subprime auto loans is fraught with risk: for the lender, who is taking a chance on a borrower with a history of failing to pay back loans, and for the borrower, who is often presented with a take-it-or-leave-it deal because of credit history and lack of options. "Generally, Credit Acceptance is a last-ditch effort," said one dealer who has used the Purchase Program. "If it doesn't work, I'm not selling them the car."
Under the Truth in Lending Act, consumers financing purchases are entitled to receive written standard disclosures about a loan. This information includes the loan amount, monthly payment, loan term and interest rate. Lenders are permitted to charge a higher interest rate, but not price, based on a borrower's credit risk, though most states cap the maximum allowable rate. In the deep subprime market in which Credit Acceptance operates, Credit Acceptance typically imposes the maximum rate allowed by state law, although as of 2015 its Funding Guidelines policy document outlined a company-wide Fixed Interest Rate of 23.99%, where allowable. Across the entire subprime market, unscrupulous dealers look for ways to increase profits by hiding finance charges in the price of the vehicle or in other fees.
Although it is hardly alone amongst subprime lenders,21 there are at least three ways that Credit Acceptance facilitates hidden finance charges, by providing opportunities for or directly advising dealers to do the following:
A Credit Acceptance document entitled "Program Set-Up: The Owner's Guide" instructs dealers on how to successfully incorporate the company's lending program into their business. The process begins with an explanation of how to purchase inventory that will work best under the program. Specifically, it tells dealers that they should be buying inventory below Black Book Wholesale Average prices in order to assure a reasonable profit when selling to "higher-risk customers."
A dealer may have access to the occasional auction bargain, but dealers routinely purchase used cars that are worth less than average because of high mileage, wear and tear, or even a history of accidents—and then mark those vehicles up as if they are of higher quality. October 2017 records from ADESA, "a leader in the vehicle remarketing industry since 1989," show that repossessed Credit Acceptance vehicles sold at auction have frequent notes such as "Structural Damage,; AS IS," "INOP," "WONT START," "WONT STAY RUNNING WITH JUMP" and "BRAKE UNSAFE."
Next, Credit Acceptance advises the dealer to display down payments on car windows, implicitly suggesting that it should hide prices. This strategy ensures that consumers don't know the real price of the vehicle—or the price that the dealer would have accepted from a buyer paying all cash up-front. Some dealers have reported unintended consequences:22
"Sounded great because before we signed up the ONLY question anyone asked was 'how much down.' Well now all of a sudden everyone wants to know the total price. It's hard to look at someone with a straight face and tell them a price that's $2500 over high retail... We were like 1/3 to 1/2 less on each one. Customers are not seeming to respond too [sic] well."
"Unless you're selling for a significant amount over what the market dictates, you can't use Credit Acceptance," said another dealer who doesn't use the Portfolio Program, but has used the Purchase Program. A former Credit Acceptance employee confirmed that CAPS generally allows dealers to price vehicles at $2,500 (or higher) over the clean retail price. This is despite the fact that many cars sold should not qualify for "clean" pricing. "You buy rough," one former Credit Acceptance dealer told us.
Furthermore, when a dealer and borrower sit down to arrange financing, it is possible for the dealer to adjust the sale price even further upward based on the borrower's credit profile—a practice apparently common enough that Credit Acceptance added a long disclaimer directly on the CAPS pricing page:
"Please remember that federal and state law prohibit you from raising the selling price (the cash price) of the vehicle simply because the customer is buying the vehicle on credit. Cash and credit buyers must be offered the same selling price for the vehicle."
Indeed, unlike many software applications that confront users with errors that prevent further progress, CAPS is unusually permissive, typically only displaying warnings when the user inputs questionable data for practically any screen, unless a vehicle's VIN or mileage is completely missing. Other seemingly important problems, such as impossible income amounts or deliberately changed vehicle models that could increase a vehicle's price might result in a warning at best, allowing the process to continue.
Since Credit Acceptance requires car buyers to pursue disputes through arbitration and state court records are only sporadically available, there is limited public data on legal battles between the company, dealers, and borrowers that have arisen as a result of these practices. However, we do get a glimpse of how some borrowers have fought back in a June 2014 lawsuit.
Car buyer Corveta Houston purchased a 2005 Buick Lacrosse from Dealz Auto Trade, Inc. in Cleveland, Ohio in 2013. Ms. Houston alleged that Dealz neglected to inform her that her car had previously been in an accident, and priced it at $7,798, or 289% over the NADA "Rough Trade-In" price, according to her complaint.23 She argued that the price differential constituted a hidden finance charge, which pushed her true APR to over 41%.
Houston further alleged that CAPS itself was part of a civil conspiracy between Credit Acceptance and Dealz to charge an excessive price, violate the Truth In Lending Act, and commit usury. Her lawsuit also contained claims under Ohio's Consumer Sales Practices Act (for overcharging for the car relative to market price and engaging in deceptive practices), Ohio's Retail Installment Sales Act (for charging a usurious interest rate above the statutory maximum of 25%), common law fraud, and the FTC Holder Rule for Derivative Liability.
After Houston's case was filed and removed to federal court, Dealz filed its own counter-claim against Credit Acceptance over the CAPS issue, arguing that it should not be held liable for a product that was entirely Credit Acceptance's responsibility.24 The case was dismissed without a finding of wrongdoing after it settled in 2015.25
The Vehicle Service Contract
When a cash-strapped borrower purchases a high-mileage used car, the borrower can be understandably concerned about looming repair costs.26 It's an issue that should worry lenders as well, given that dealers told us the top reason a borrower stops paying on a car loan is because of mechanical breakdown.
The industry's answer to this situation is the Vehicle Service Contract (VSC), an agreement that the dealer will directly, or indirectly via a licensed repair center, fix certain problems with the car for a given number of years or until a certain number of miles have been driven.27 On Credit Acceptance loans, these contracts are pricey—around $1,500 for two years or 24,000 miles of coverage for most policies. Given that the average mileage on vehicles financed through Credit Acceptance is nearly 100,000 miles, and given the low quality of cars available below average wholesale price, repairs are a near-certainty. About half of Credit Acceptance deals involve VSCs.
Based on our analysis of Retail Installment Contracts (RICs) included as part of legal actions by Credit Acceptance against defaulted borrowers in Alabama, California, Michigan, Mississippi, New York, Ohio, and Virginia, we determined that roughly two-thirds of Credit Acceptance defaulted loans involved a VSC. These findings point to two possible conclusions. The first is that those most likely to default—people with weaker credit profiles—are disproportionately likely to be sold these high-priced contracts. This could be because the vehicles they purchase are disproportionately likely to break, and/or because dealers desire some a way to build an additional, undisclosed credit charge into their loans, despite the violation of federal and state Truth In Lending laws. The second is that these contracts, rather than preventing defaults by providing funds for consumers to cover repair costs, may actually be hastening defaults by fueling borrower frustration over rejected claims.
By adding an expensive VSC to a loan balance, dealers can boost the advances they receive from Credit Acceptance by hundreds of dollars. "Vehicle Service Contracts have been the golden egg for many years," according to a former Credit Acceptance dealer. Although the dealer sells the VSC, in an "Administrator Obligor" arrangement, meaning that the administrator of the contract is obligated to fulfill its terms, the car dealer has no role in preparing the legal terms of the contract or handling claims—it merely serves as a marketing channel. In many states, a special type of insurance license is required to serve as an Administrator Obligor.28
This arrangement is beneficial to car dealers, who have to perform virtually no work in order to earn their commission. They also sometimes have a legal monopoly on the sale of VSCs, which in some states may not be sold direct to consumers outside of a dealership. Dealers in the Credit Acceptance network can simply check a box on the CAPS screen that adds an overpriced VSC to the customer's loan—and adds a boost to the dealer's advance. As of 2015, Credit Acceptance approved VSCs from only four vendors: Wynn's Extended Care, Inc., Old United Casualty Company, Mechanical Protection Plan, and SouthwestRe.29 (Until 2009, Credit Acceptance also had an arrangement with Warrantech.)
This click of a mouse can sometimes turn an unprofitable loan into a profitable one for the car dealer. Accordingly, while VSCs are technically optional and vehicle purchase agreements do clearly state that purchasing a VSC is not necessary to purchase a car, the presence of the VSC line item can be necessary for the dealer to be willing to sell a vehicle at all. As a member of the website "Auto Dealer People" named Mary wrote on March 21, 2010:30
"CAC made it so that you have to put in a warranty in the calculations. (Warranties are optional) but if you take out the warranty you get far less profit and will be unable to make the deal
A former Credit Acceptance employee we interviewed told us that he had advised dealerships that they were going to want to put a warranty on every subprime loan. Simultaneously, VSCs have prevented Credit Acceptance from expanding into a number of large dealerships. "Big stores won't sign with Credit Acceptance because they want to sell their own Vehicle Service Contracts," a former dealer said. "Dealers view these funds as their retirement money."
While some dealers have also argued that increasing the price of a loan by adding a VSC is not always desirable because it increases the overall risk that the consumer will default,31 it is important to remember that the dealer immediately assigns the loan to Credit Acceptance Corporation and gets paid its up-front advance regardless of the loan's future performance. Faced with a choice between a higher up-front profit or a small chance of a long-term payoff, many dealers will choose the former.
On a typical sale, a car dealer earns anywhere from $29532 to $38533 in fixed profit on a 24 Month/24,000 Mile VSC usually priced between $1,380 to $1,580, though some dealers will charge in excess of $2,800 for the same exact product. (Dealers earn about $150 for the sale of GAP insurance coverage, which covers the difference between the loan balance due and insurance payments if a car is deemed a total loss in a collision.) About $100 of that amount goes to the Third Party Administrator, a portion is set aside to pay for claims and overhead, and the rest is passed on as profit to Credit Acceptance. In one notable instance, a San Francisco-area car dealer sold a VSC for $2,530 on a car worth $6,500.34 Since the price of the VSC is always bundled into the financed amount, that VSC represented one-third of the total amount financed.
With this background, it is difficult to consider the VSC as anything other than a veiled vehicle price increase. That is, in fact, how VSCs are marketed, and how some car dealers think of them. "What profitable opportunities did Wynn's Extended Care provide to you?" one car dealership owner, Travis Armstrong, was asked in a September 3, 2014 deposition. "I guess the ability to make my vehicles more valuable by having an extended warranty on them," he replied.35
For Credit Acceptance, the potential Truth In Lending Act liability is only the beginning of the problem posed by VSCs. The real issues begin when consumers actually try to get their used cars repaired under the terms of their contracts.
Playing the VSC Shuffle
Once a buyer has finally managed to drive off the lot with their new used car and VSC, it can sometimes be merely a matter of days before the car breaks down. Customers inevitably return to the dealership where the vehicle was purchased, but often the dealership will tell the customer that simply examining the car will cost several hundred dollars. Even when customers are able to get a vehicle examined, they are faced with VSC providers that resist paying out claims using a variety of suspect tactics.
Car buyer Penny Bradley's case highlights the lengths that VSC providers will go to in order to avoid honoring claims. Even though Bradley paid for her VSC, she was still denied coverage for a variety of reasons, one of which was that she never actually obtained a contract. In the view of the Third Party Administrator, she had merely signed an "application." Only once the provider acknowledges that the customer has a real "contract" can a claim be filed, the Administrator told her.36
Tijuana Johnson, a New Jersey car buyer who purchased a used 2007 Saturn in 2011, was told by her repair shop that Wynn's Extended Service had rejected her claim because it was still covered by the car manufacturer's original warranty—but the manufacturer's warranty program rejected her claim because Wynn's authorized repair shop had already disassembled the engine. Left in a state of limbo, she filed suit, which consumed the next six years and ultimately spanned several state and federal courts.37
What customers may not realize is that Credit Acceptance rates its dealers on a scale of 1 to 6, with lower ratings reflecting better historical loan performance, among a number of other factors. Dealer ratings are key to future profitability, as a dealer with a "2" rating will receive hundreds of dollars more as part of every loan advance relative to a dealer rated "4," even if those dealers originate the exact same loans. Dealers are thus incentivized to do everything possible to protect their ratings.
According to a former Credit Acceptance employee we interviewed, one factor that determines dealer rating is the dealer's "repair ratio," or the dollar amount of VSC-covered repairs as a percentage of the dollar amount of VSC products sold. The employee explained that a high repair ratio would be an indication that dealers are making loans on poor quality vehicles. A dealer's rating might therefore be penalized.
Yet Credit Acceptance also insists that in order to profit from its underwriting system, dealers should buy cars that are sold below Black Book wholesale price. Such purchasing directives come with the obvious risk that "bargain" vehicles might be of low quality and more likely to require repairs. Each dealer must therefore walk an impossibly fine line between profit margin and future repair likelihood to maintain a good rating. Practically speaking, the best way to do so is to refuse claims.
An article from NBC12 in Richmond, Virginia38 describes this scenario playing out:
"The Hopewell woman bought the car from Unique Auto Sales on Midlothian Turnpike in Richmond back on February 25. She also purchased a 24-month, 24,000-mile service contract from a third party.
She said when the dealership delivered the car to her, the check engine light was on.
'I said, 'What's wrong with the car?' and they said, 'Oh, nothing. These things act kind of funny because of the computers and all,'' Lathrop said.
She says the dealership turned the light off and sent her on her way. Seventeen miles later, she stopped to get gas, and the check engine light came on again. Plus, it was hard to start. A diagnostic revealed a timing cam sensor was triggering the light.
She had it towed to an auto shop, where techs told her she needs a new timing belt...which would run about $1,200.
She went back to the dealership but says workers there told her she purchased the car 'as is.' When she pressed to see the manager, she says she was kicked off the property.
She says she also didn't have any luck when she contacted the service contract provider."
The article goes on to state that, "NBC12 also reached out to the service provider, Superior Protection Plan." A corporation search and a trademark search turn up no such company, and no such trademark. Who or what is Superior Protection Plan, and how would any customer (or television investigative journalist) know who is really behind the VSC?
A Google search for Superior Protection Plan yields few clues: the first page of results is full of complaints and scam warnings, as well as advertisements for extended automotive warranties and cookie-cutter websites of questionable origin. No company seems to want to take ownership of the phrase—in fact, Credit Acceptance's Funding Guidelines instruct dealers to avoid listing "Superior Protection Plan" on RICs, in favor of "SouthwestRe." This begs the next question: how did NBC12's journalist even know to contact "Superior Protection Plan" in the first place?
That question is perhaps answered by a lawsuit filed in the Circuit Court of St. Francis County, Arkansas by Carla Walker. Ms. Walker sued Credit Acceptance,39 alleging that the company "used deception and fraud upon Plaintiff in representing (or withholding) the amount she was financing." The lawsuit was removed to federal court in the Eastern District of Arkansas before being remanded back to state court.
Exhibit A to her complaint40 is a copy of a vehicle purchase agreement on a carbon paper form provided for the dealer not by Credit Acceptance, but by paper form company Reynolds and Reynolds.41 Both a VSC and GAP coverage are listed in addition to the vehicle's $13,309.85 base price, and the VSC is described only as "Superior Protection." Additionally, next to "DOCUMENT FEES," there is a discretionary $499.00 charge, set by the dealer regardless of any documents provided. There is no APR listed on the form, let alone Truth In Lending disclosures—only a total amount of $15,478.43. Credit Acceptance started Walker's initial balance at $32,804.84.
Good Documents Are Hard To Find
Walker may not have been given a copy of her standard Credit Acceptance RIC by her dealer, even though she signed it. As Penny Bradley testified,42 it is common practice for dealers to rush customers through paperwork, asking them to sign repeatedly without paying any attention to terms and conditions. In Walker's lawsuit, Credit Acceptance filed its copy of the RIC—but no paperwork explaining her VSC or GAP coverage—as an exhibit43 , noting that Walker agreed to a 28.99% APR and a $17,326.21 finance charge for a vehicle priced at $13,395.00. Only on the RIC was the true VSC provider actually listed: First Automotive Service Corporation, along with "24 Mos. \ 24,000 Miles" as the only terms.
Many Credit Acceptance car buyers never actually see the paperwork associated with their VSC aside from a small box on the RIC, a copy of which they may also never receive. The RIC used for Credit Acceptance deals is written and copyrighted by Credit Acceptance Corporation, which means that dealers simply use a template that Credit Acceptance hands to them; their only job is effectively to fill in the blanks with the help of CAPS. The VSC marketing materials and contracts, on the other hand, are produced by a third-party vendor, and the specific terms of the VSC are only very broadly referred to in the RIC, if at all. In many RICs, as was true for Carla Walker, no terms concerning which parts are covered, repair procedures, acceptable repair shops, etc. are specified. (As of the writing of this report, Ms. Walker's case is ongoing.)
Failure to provide documentation is not just an inconvenience. In some states, it's very likely illegal. Georgia law § 33-39-9 requires insurance providers and their agents to provide access to documentation upon request within 30 days. New Mexico law §§ 59A-12A-4 and 59A-18-30 similarly require written agreements for insurance products.
With so little information, whom does one even call to get a copy of the contract? Even assuming that an average car buyer understands what an "obligor" is—a dubious assumption at best—according to the State of California Department of Insurance, "Some obligors are generally better at honoring their contracts than others. Therefore, the most important aspect of any contract is who the obligor is."44 In some cases, by the time a borrower realizes they never received the loan paperwork (including the car title) to begin with and need them, their subprime used car dealer has gone out of business.45 There's no one to call, and no contract to refer back to.
In Penny Bradley's case, the VSC was provided by Wynn's Extended Care, Inc. Her RIC listed a $1,580.00 charge for the "Cost of Optional Extended Warranty or Service Contact Paid to," but the adjacent line had nothing written in it. On top of the application-versus-contract issue, Wynn's ultimately refused to honor her warranty because it claimed that her used 2007 Honda Civic Hybrid counted as an "electric" car, which was ineligible for coverage. She sued.
In a deposition on August 29, 2014, Ms. Bradley was asked, "Other than Exhibit 2 [the VSC application], did you get any documents that had the name Wynn's on it?" Her answer: "No."46
We also got in touch with Samantha Rajapakse, a buyer of a used 2007 Chevrolet Trailblazer with 94,998 miles. She purchased the vehicle in 2014 from 1 Stop Auto Sales in Memphis, Tennessee for $20,134.24, of which she financed $10,893.36 through Credit Acceptance. Included in the financing was a $1,380.00 24 Month/24,000 Mile VSC. Ms. Rajapakse's RIC states, "Refer to the optional extended warranty or service contract for details about coverage and duration," suggesting the existence of a separate document. This wording was standard on Credit Acceptance RICs as early as 2012.
In 2016, Ms. Rajapakse sued Credit Acceptance Corporation in federal court in the Eastern District of Michigan pro se.47,48 In her complaint, she alleged that she had taken her truck to two Chevrolet dealerships near Memphis, and that neither dealer would honor her VSC. She also attempted to bring her vehicle to certified mechanics, who also would not honor the VSC. Her communications with Credit Acceptance, in writing and by phone, did not lead anywhere. She was unable to ascertain where the VSC would be honored, and Credit Acceptance never made it clear to her who was actually behind the VSC or what its terms were. In her complaint, she alleged that she was given a telephone number for a "warranty company unknown." The only name that offers a clue is printed on her RIC: First Automotive Service Corporation.
A Tangled Corporate Structure Obscures Accountability
First Automotive Service Corporation (FASC) was incorporated in New Mexico on January 18, 2000, after which it was registered as a foreign corporation in both Texas and Pennsylvania on February 3, 2000 and July 19, 2000, respectively.
A Google search for the company reveals that despite having at least three corporate entities, FASC does not appear to have its own dedicated website and is surprisingly hard to find in the real world with any degree of certainty. Instead, the generic-sounding brand "First Automotive"—which an average consumer might not directly associate with their stated VSC provider since two words are missing and brands commonly begin with "First"—appears on the Products sub-section of the Agents section of a website belonging to a different company called Southwest Re.
In turn, Southwest Re is also really a number of different and similarly-named companies. Southwest Re, Inc., Southwest Reinsure, Inc., Southwest Reinsure (NM), Inc., and Southwest Risk Management, Inc. are registered in various states as domestic and foreign corporations, including New Mexico, Florida, Pennsylvania, and Georgia. Twenty-odd other companies, including the three FASC entities, share the same Albuquerque P.O. Box, including Warranty Solutions, Inc. and Superior Autocare, Inc. With this information, it becomes slightly clearer that Southwest Re is actually in the automotive warranty business, but the key piece of information—almost guaranteed to be ignored by customers—is the registration mark symbol alongside the First Automotive logo on the Southwest Re website. A search of the United States Patent and Trademark Office (USPTO) trademark database reveals that trademark application number 86139923 for "FIRST AUTOMOTIVE" in the field of "Third-party extended warranty services, namely, vehicle service contracts on vehicles manufactured by others for mechanical breakdown and servicing," is actually registered to Helios Financial Holdings Corp. According to a 2014 interview with its owner, Jim Smith, Helios is the umbrella organization that owns Southwest Re and its many sister organizations.49 The trademark was only granted registration on August 29, 2014. If a car buyer searched for the link between February 1, 2000 (when Helios claims it began using the mark in commerce, over a year before it even existed) and that date, it would have been nearly impossible to tell who was behind the First Automotive brand.
As of October, 2017, the Southwest Re website—heavy on stock photography of smiling Caucasian families, generic icons and stretched-out graphics—contains a curious statement on a page designated for "Contract Holders." (The link to this page only becomes apparent on the home page after the visitor waits through two different animations that last roughly twenty-five seconds. It is the last link to appear.) Above a button labeled "MAKE REQUEST," the site states, "Perhaps you never received a complete copy of your contract from your dealer, or maybe you've misplaced it over time..." This is a stunning near-admission that indeed, perhaps many of Southwest Re's customers, who undoubtedly have no idea that they are Southwest Re customers, "never received a complete copy of [their] contract from [their] dealer." At the very least, the problem is so frequent that the company felt compelled to highlight it on its own website.
Other aspects of Southwest Re's operations raise serious questions as well. Its introductory video makes clear that the company considers car dealers its customers, and that it exists to help them boost and keep "their" profits—not to help fix cars. The company calls itself "a nationwide provider of premier F&I products and agent and dealer participation programs," selling, "highly profitable turnkey solutions for the F&I office." In this context, car repair seems like an afterthought.
In October 2017, under the heading "Claims Paid," the Contract Holders page stated, "Each month, we cover the cost of many claims. In September, we paid $4,600,000." The Internet Archive's cached copy of the same page from a year prior shows that in September of 2016, the company claims to have paid $4.9 million exactly.50 In July, 2016, the number was once again $4.6 million even.51 Yet in June, 2016, the figure was quite specific: $5,211,649.52 The fact that the statement "we paid" is not qualified by whom received payment, or what for, is enough to raise eyebrows after such a vague, generic assertion as "we cover the cost of many claims."53
Another box with the title "Contract Holders Say" and a generic faceless person icon states immediately below, with no quotation marks or any other kind of attribution, "My 10-year contract with SouthwestRe was a terrific experience." Combined with the site's amateur design, this non-quotation from no actual customer is an enormous red flag. At worst, it is boilerplate text mistakenly left in place by a website designer. (Only slightly more convincing testimonials attributed to individuals with names do appear elsewhere on the Southwest Re website.54 )
The VSC products page is also confusing given that the company nominally sells, or owns a company that sells, these very products. Under the heading "Our Exclusive Brands," there are descriptions of First Automotive and other Southwest Re warranty brands, but no product details. Only on the "Agents" side of the site do descriptions of each VSC product offering appear, with tiny pictures of brochures. Yet the site appears to omit contractual terms of any of the company's VSC agreements or even a complete brochure anywhere for marketing purposes.
For a warranty or insurance company, this is unusual. Standard car insurance policies, such as those offered by GEICO or Progressive, include on-line calculators that make terms and pricing transparent. Policy documents are e-mailed or downloadable instantly or within minutes of a policy becoming effective. There is no mystery as to whether your car has coverage, how much, and who is behind it. Yet with First Automotive, a brand corresponding to FASC, in turn a division of Southwest Re, in turn a division of Helios, everything is for some reason shrouded in mystery.
Contract Holders Denied Access To Documents
Perplexed as to how Samantha Rajapakse could be missing her VSC documents, we inquired with Ms. Rajapakse as to whether she had filled out the "Contract Copy Request" form on the Southwest Re website. Ms. Rajapakse replied to us by e-mail and included the transcript of her related e-mail inquiry to Stephen W. King of King & Murray PLLC, counsel for Credit Acceptance. Mr. King had attached a copy of the completed and signed RIC, but not the VSC agreement referenced therein. If Credit Acceptance had the documents, it was not offering to share.
As Credit Acceptance's lawyer responding to her formal complaint, Mr. King was clearly already aware of her allegations in federal court regarding the VSC. Nonetheless, he encouraged her to "follow the procedures set forth in the RIC / Arbitration Clause," without explaining the omission of the VSC documents. In a follow-up message referring to the RIC, Ms. Rajapakse explained to us what she had been led to believe: "[M]any of the customers like me gets [sic] this on our contract and nothing else. [This] is the only thing they have regarding the warranty."
We placed and recorded a call on Ms. Rajapakse's behalf on October 6, 2017 in order to get to the bottom of the issue. After contacting Southwest Re's toll-free telephone number, we were transferred by a receptionist to the "contract holder's department," where our call was answered by David in "Claims." We informed David that we were attempting to get a copy of a contract for a customer who filled out the form on their website and never got a response. David asked for the last eight digits of the vehicle's VIN number, which we provided, and he successfully located the VSC agreement in his database, which he informed us had already expired. We asked for a copy of the agreement anyway.
"Yeah, um, we're unable to send these out—we don't have these on our website yet," David responded. "So with these, uh, I don't direct people to the website; we're working on it with Credit Acceptance as a peer to get these added to our website, but we don't have a copy of these yet." This was confusing, but David went on. "So, I have a digital copy of ones that I have in my system but they are secured in there and I can't remove them to send them out anywhere."
"The contract holder is entitled to a copy of their contract," we pointed out.
"Yeah. I'm aware. I'm unable to send one out as a contract Administrator," David replied. "We don't have these on our website so the only other thing to do would be to direct them back to the selling dealer."
"The contract isn't with the dealer, it's with you," we pressed, attempting to clarify.
We went back and forth several times with David trying to establish how he could look up Ms. Rajapakse's VSC agreement but simultaneously not be able to provide it to her. "I have a digital copy embedded into my system," was the best clarification David could offer. David's supervisor, Jeff Walker (who refused to provide his own last name to us, because he said he was not required to)55 , also insisted that he could not and would not provide us with any documents, not even a blank contract with general terms, which he claimed to have. Mr. Walker referred us to Credit Acceptance, having denied that Ms. Rajapakse had a contract with his company at all. He stated that his company was merely the VSC "Administrator," even though the VSC "Company" listed on the RIC is clearly First Automotive Service Corporation. "We do not hold the contract," he stated. "The bottom line is, I can't send you a copy of the contract, I can't send you a blank page of the contract—I can't send you anything. I don't have the ability to. You have to contact Credit Acceptance."
When we asked about the purpose of the Contract Copy Request form on the Southwest Re website, Mr. Walker responded curtly, "'Cause it's there to help out customers, is there anything else I can help you with, sir?" In the end, we were unable to obtain a copy of Ms. Rajapakse's VSC agreement from Southwest Re.
The complex corporate structure of the Helios/Southwest Re family of companies, along with numerous associated brands, make it far too difficult to ascertain basic facts about a large proportion of the VSC products sold with Credit Acceptance car loans. Furthermore, Southwest Re's explicit referral back to Credit Acceptance raises the question as to whether Southwest Re deliberately acts as an alter-ego of Credit Acceptance in regulated markets where Credit Acceptance may not want to risk its own legal exposure. Our call brought into question the very legitimacy of FASC VSC agreements in the first place: selected by Credit Acceptance to begin with: overpriced, rarely honored, valid for short periods, and based upon phantom terms and conditions frequently unavailable for contract holders to read.
If in fact Credit Acceptance is routing its customers to unusable VSC products, the ramifications could be significant. The company could have liability for tens or hundreds of thousands of fraudulently or improperly-structured loans, and could be forced to refund both principal and interest payments and/or waive late fees on those payments. Especially if there emerges any kind of proof that Credit Acceptance knew that the VSC products it had chosen (and thanks to CAPS, priced) for its customers were not designed to actually pay for necessary repair work, the resulting damage could force the company to record significant unexpected losses. After all, the numbers indicate that a substantial portion of buyers are uncomfortable with buying a used car that has no service contract of any kind.
As of September 26, 2017, James B. Smith agreed to sell the Southwest Re group of companies to one of Canada's largest insurance companies, Industrial Alliance Insurance and Financial Services, Inc., also known as iA Financial Group.56
Trouble with Insurance Regulators
Despite FASC's refusal to assist, we were able to obtain copies of FASC VSC documents both from the few number of lawsuits where VSC contracts were in the possession of the car buyer and attached to the complaint, and from the Nevada Department of Insurance57 going all the way back to January 2007. FASC documents featured various brand names such as "XTENDED CARE APPLICATION," "HENDRICK AUTOGUARD," "QUANTUM PROTECTION," and the elusive "SUPERIOR PROTECTION PLAN," each with corresponding logos at the top. Since none of these brands match the FASC dealer marketing materials we saw or were ever registered as a trademark with the USPTO, it would have been difficult for a car buyer to pin down each brand's backer without prior knowledge. Even today, internet searches for these brands reveal many confused borrowers.
In the unlikely event that a Credit Acceptance car buyer is able to actually get their hands on their own FASC agreement, one particular disclaimer sticks out:
THIS CONTRACT IS NOT AN INSURANCE POLICY
At best, this is a confusing claim, because four insurance commissioners in Wisconsin, Utah, Missouri and New York have cited FASC, a licensed VSC provider, for violations of each state's respective insurance laws. On January 25, 2007, the Wisconsin Department of Insurance fined FASC $10,000.00 for "failing to file its financial statements." In August of the same year, the Wisconsin Department of Insurance again fined FASC for "failing to comply with an Order of the Commissioner."58 On July 17, 2007, FASC was fined $15,000 by the State of New York because "Respondent acted as a service contract provider without first obtaining an approval of a registration to do so from the Superintendent."59 Within 30 days of each violation, FASC should have reported the violations to the Insurance Commissioner of the State of Utah and the State of Missouri's Department of Insurance, Financial Institutions and Professional Registration—jurisdictions where it was licensed—but it didn't. Instead, it had to pay Utah a $250 "monetary penalty" in March, 200860 and Missouri a $1,000 "voluntary forfeiture" in January, 200961.
In addition, a Texas state court judge issued a permanent injunction against FASC on January 12, 2004 in response to a lawsuit filed by the Texas Department of Licensing and Regulation (TDLR).62 TDLR issued a press release the same day, stating that it had obtained,
"a permanent injunction ordering a New Mexico service contract provider to stop offering or selling extended warranties for automobiles in Texas for which it lacks proof of adequate financial security. Judge Suzanne Covington of the 201st District Court in Austin Monday ordered First Automotive Service Corp. (FASC) of Albuquerque to stop providing, offering, advertising, selling or marketing service contracts unless it can show that the policies will be properly backed."63
The injunction arose out of an investigation into a bankrupt company called Warranty Gold Ltd., which ultimately surrendered its license, and for which FASC once issued VSC contracts. TDLR obtained the injunction on the basis that FASC was lying to customers about its VSC contracts being insured by a company called Dealers Assurance Company (DAC). DAC said it had never heard of, let alone insured, FASC. Soon after, Southwest Re purchased DAC.
While insurance commissioners have targeted FASC repeatedly for its compliance failures, the plethora of service plan brands and corporate entities connected to them has masked the ultimate decision-maker not only from consumers, but from regulators as well. Credit Acceptance chooses the warranties for its dealer-partners. Credit Acceptance receives the bulk of the profits from the sale of each warranty. It even handles the day-to-day operational aspects of maintaining the warranty relationships with dealers. In response to one lawsuit discovery question, Armstrong Auto Sales, Inc. answered as follows: "Wynn's did not provide any training to Travis Armstrong or Armstrong Auto Sales, Inc. Any training would have been performed by Credit Acceptance Corporation ('CAC'). Wynn's provides a Dealer Kit and a rate sheet for dealerships to consult when marketing service contracts."64 During a deposition in the same case, the dealership owner also stated that Credit Acceptance is the first party to receive warranty paperwork—not the warranty provider. "This is what we would send in to Credit Acceptance, and then they send it to Wynn's to start their extended service plan."65
Credit Acceptance gets an additional bonus: if and when a car with a VSC agreement is repossessed, it is entitled to the balance of the amount remaining in the trust fund for that service contract—nearly $700.66 Even if the car buyer paid cash up-front, the refund does does not necessarily go back to the buyer.67,68 This is outlined in a confusingly worded clause in the RIC, buried among other terms and conditions. The clause gives Credit Acceptance an additional incentive to repossess vehicles, which court records show, it often does.
This is consistent with previous research on the VSC industry. In 2011, the Better Business Bureau of Eastern Missouri and Southern Illinois released a report entitled "Vehicle Service Contract Industry: How Consumers Lost Millions of Dollars."69 According to that report, based on a survey of 660 complainants, VSC agreements paid for claims only 6.55% of the time, leaving 93.45% of claims denied. On this basis, the Better Business Bureau, which received no response to its report from Missouri's Department of Insurance, concluded, "Lax oversight of the VSC industry has created a national scam that has been evolving for at least 25 years, costing consumers millions of dollars."
While it could be argued that the regulatory landscape for VSCs has fluctuated over time, FASC's track record is too abysmal to ignore. The implications for consumers are real, as complaints filed with state departments of insurance could have a real impact on both FASC's and Credit Acceptance's business practices were it clear how and where VSC products were regulated. Nor do recently-implemented (and optional) electronic contracts solve the key problem that VSC claims are rarely honored due to terms that are poorly understood, if they are visible at all. Given the large number of Credit Acceptance deals that depend upon VSCs, increased regulatory oversight, new legislation, and/or a class-action lawsuit able to recoup losses for affected consumers could each pose serious threats to Credit Acceptance's business model.
Credit Acceptance Corporation's business practices expose it to considerable regulatory risk depending upon how government policies shift over the coming years.
The Garnishment Machine
Many Credit Acceptance investors believe that the company's long history gives it experience in the subprime lending market and that the company uses its experience to underwrite better loans. The reality is that the company prioritizes learning as much as possible to maximize collections over information that might inform an objective lender about the customer's creditworthiness. The entirety of its underwriting process for borrowers without a FICO score consists of checking that wages reported on paystubs are accurate—not that they are sufficient to pay any particular amount.70 This is in keeping with industry standards (where verification may not happen at all), and is perhaps why subprime loans are sometimes referred to as "liar loans." According to Bloomberg, "Jeff Brown, Ally Financial Inc.'s chief executive, said verifying income isn't the norm. Ally, he said, checked incomes on 65 percent of its subprime car loans. GM Financial's AmeriCredit unit checked roughly the same percentage."71
Court records reveal that Credit Acceptance's true business model is based upon an ugly foundation of wage and income tax refund garnishment. Between 1995 and 2006, it applied to garnish either wages or income tax refund deposits in 81.94% of the 17,480 Credit Acceptance cases for which data was available in Michigan's 36th District Court, which includes Detroit. Similarly, in the 18,284 Credit Acceptance cases for which data is available from 2007 through 2017, the company requested garnishment 63.43% of the time. Generally, three out of four Credit Acceptance borrowers who find themselves in court will end up with a garnishment order against them due to their car loan. Such orders can last for years, and sometimes, decades, following people from job to job. Yet garnishment only appears to result in a complete collection of funds about 30% of the time on average, depending on how many years a collector is willing to pursue funds.
Credit Acceptance has never disclosed to its investors the extent to which its business model completely depends upon judges doling out garnishment orders in case after case after case after case, often without carefully reviewing the underlying loan documents.72 This is problematic, because any time a publicly-traded company depends on a government process to make money, it exposes itself to regulatory risk should that process suddenly change. Amendments to local court rules, state law, federal law concerning garnishment, and federal law concerning bankruptcy73 could all have a substantial impact on Credit Acceptance's ability to collect from defaulted borrowers. That is why internally, Credit Acceptance's computer systems track which states are more permissive of wage garnishments—another material fact not disclosed to investors.
The company's business model has severely impacted impoverished individuals all over the United States, but especially in Detroit, where between 2008 and 2016, the company's share of the Michigan 36th District Court's docket increased by nearly a factor of ten, from 1.39% to 11.37% of all civil lawsuits filed. In 2017, not yet over at the time of this report's writing, Credit Acceptance's share of the 36th District Court's docket is at an all time high of 11.97%74, surpassing even the company's share during high rates of litigation in the late 1990s. Still more incredible is the sheer number of lawsuits the company is involved in: at least 153,842 non-federal legal actions that we could find nationwide since 1995.75
It is highly unusual for one company to represent such an enormous percentage of any major city's court system activity. Perhaps even more alarming, the 36th District Court charges a recurring $15.00 garnishment fee (payable by check in person at the courthouse), which may make the court dependent on a major filer paying tens or hundreds of thousands of dollars in fees annually. As a result, Credit Acceptance Corporation has helped bring at least $2,349,450 in fee revenue into the coffers of the 36th District Court, out of $24,254,400 charged total during the time period we examined between 1995 and 2017. By 2017, for the first time ever, more than 30% of all of the periodic wage garnishment docket entries in the court's computer system were associated with Credit Acceptance cases filed that year.
The overall court filing statistics also reveal a stark change in Missouri. While Credit Acceptance was involved in an average of 924 Missouri lawsuits per year from 1995 through 2010—almost always as the plaintiff—by 2012, it was only involved in ten lawsuits filed that year, and of those, it was a defendant in seven. From 2013 on, Credit Acceptance virtually disappeared from the Missouri court system. It is possible that the company was reacting to the $1.2 million judgment in the case of Carrie Peel, a car buyer from Independence, Missouri. Or it is possible that other changes in state law convinced management to pull out of the state. Either way, investors were never told about the change.
Perhaps even more astounding is the fact that a 1995 Credit Acceptance garnishment judgment on which the company finally finished collecting in 2017, twenty-two years later, is by no means an outlier. A scatter plot of the company's Detroit dockets containing "JUDGMENT SATISFIED" entries clearly shows that in essence, the company will never stop collecting so long as it can find a customer, and long after it has told Wall Street that a loan has been written off. Plenty of cases from the 1990s were satisfied after 20 years, and many are likely still being collected on today thanks to constantly renewed garnishment judgments and the efforts of persistent lawyers.
That Credit Acceptance may be having trouble collecting of late is hardly out of keeping with industry trends. A Q2 2017 Equifax analysis of subprime auto debt stated, "Performance of recent deep subprime vintages is awful."76 While Credit Acceptance says it doesn't change the terms of loans to mitigate poor collections, it does enter into separate debt collection agreements with completely different terms.
According to a lawyer who has represented Credit Acceptance borrowers in New York, borrowers are frequently approached about settling the balance of their debts once their repossessed vehicle is sold. Borrowers are told that the next step is for the company to take them to court and seek wage garnishment. As a result, borrowers often agree to payment plans, but are disappointed to see that even after a car is sold, their loan balance is little changed. That's because vehicles often sell at auction for amounts far below the elevated prices that borrowers have paid.
One former employee said he was aware of debt restructuring companies calling Credit Acceptance and asking for deals to be restructured so that individuals could keep their cars and continue paying. He believed terms were changed, including interest rate reductions, though this contradicts the company's public stance.
It goes without saying that Credit Acceptance is abusing the legal system, and it backs off quickly when borrowers hire counsel. Even if each lawsuit were perfectly crafted and justified, there would still be reason to question the company's litigation volume. At such a massive scale, it is inevitable that there will be problems with the final work product in a large number of cases.
"Reasonable Efforts:" Going To State Courts On Bad Math
Credit Acceptance sued Shelley Williams in the Circuit Court of Mobile County, Alabama on May 12, 2014.77 The company's lawyers had already filed over 120 lawsuits in Mobile County that year by the time they got to hers, but Williams's case ended up being different. Unlike most of the other car buyer-borrower-defendants who ended up with judgments against them, Ms. Williams ultimately hired a lawyer, fought back, and won.
By the time Williams's lawyer, Judson E. Crump, entered her case, she had already lost. The court entered a default judgment against her on July 22, 2014 in the amount of $19,371.69 plus court costs—approximately double her 2003 Volkswagen Passat $10,990 cash price at purchase. Crump filed a Motion for Relief from Judgment and Stay of Garnishment on her behalf on October 3, 2014. On November 21, 2014, Judge Rick Stout granted the motion and dismissed the case in its entirety—a remarkable turn of events.
Crump's legal magic trick was checking Credit Acceptance's math. Credit Acceptance's 2014 complaint against Williams knowingly asserted that Williams continued to owe interest on the full amount of her balance starting from January 1, 2012, even though she had made payments as recently as 2013. In addition, Credit Acceptance sought compounding interest of $6,639.34 starting from January 1, 2012 on an amount of $11,071.61 that it was only reasonably able to calculate for the first time starting on August 15, 2013, after the car was repossessed. As Crump noted in his motion, "The 'Affidavit of Account' upon which this Court relied in entering the default judgment in this matter, was obviously false. Rule 60(b)(3) of the Alabama Rules of Civil Procedure authorizes this Court to relieve a party from a judgment entered against it for 'fraud, misrepresentation, or misconduct of an adverse party.'" In Alabama alone, Credit Acceptance has filed upward of 4,000 lawsuits, many using the same Birmingham law firm of Zarzaur & Schwartz, P.C.
Alabama isn't the only locale where math errors are apparent. In the case of Cassie Cannon,78 filed in San Francisco Superior Court, errors were everywhere. To start, the car dealer, Valencia Auto, printed her Installment Sale Contract on a mis-aligned dot-matrix printer. Consequently, each line of text, including each number, was printed in the space below the designated line, making the document hopelessly confusing to read. For example, instead of a $100.00 "Document Preparation Fee," it appeared that Ms. Cannon was being charged a $100.00 "Smog Fee Paid To State," which in contrast to the former fee would be statutory and non-negotiable.
Furthermore, Credit Acceptance filed a Declaration Regarding Interest Computation that similarly computed interest from September 27, 2008, "the date of default." Yet the story told by the Credit Acceptance Customer Payment History Report, filed separately as an attachment to another Declaration,79 isn't nearly as clear. To start with, the Balance column is computed incorrectly for some days where multiple transactions appear and is therefore unreliable. In addition, it appears that on September 27, 2008, a Western Union payment bounced, resulting in a subsequent Not Sufficient Funds (NSF) fee and corresponding a transaction reversal. Subsequent payments after the date of default appear on the Customer Payment History, also made via Western Union, but they are not accompanied by NSF fee entries, implying that they cleared. Even so, there are more inexplicable transaction reversals with no accompanying reference number or memo field that nullify each subsequent payment. It is not clear if these reversals were automated or manually entered, but at the end of the day it didn't matter. On June 19, 2013, the judge ruled in Credit Acceptance's favor to the tune of $13,510.68.80
Whether Credit Acceptance's conduct in these cases was fraud, misrepresentation or merely negligent misconduct, the company's enormous litigation volume obviously increases the likelihood that its cases might suffer from defective mathematical calculations. These types of errors caught the attention of Human Rights Watch, which produced a 2016 report entitled, "Rubber Stamp Justice: US Courts, Debt Buying Corporations, and the Poor." The report specifically singles out Credit Acceptance:81
"Several years ago the state district court in Southfield, Michigan began asking its clerks to scrutinize garnishment requests submitted by debt buyers to make sure they were free from errors. 'My court administrator was clearly troubled by these cases,' Judge William Richards told Human Rights Watch. 'She saw some problems and really took it on herself to try and engender some reforms. Our court was more aggressive than most at screening requests for garnishment. At one point we were doing 9,000 garnishments per year and one clerk was screening all of them.'"
"In 2005, the court returned numerous garnishment requests loaded with apparent mistakes to the attorney who had filed them on behalf of debt buyer Credit Acceptance Corporation. 'He filed 60 or 70 garnishment requests in a single day,' Judge Richards recalled. 'There were thousands of dollars' worth of errors.' Some of the garnishment requests appeared to relate to judgments that were void or already satisfied while others appeared to include excessive interest. The court's clerk asked Credit Acceptance's attorney to correct errors and provide additional supporting documentation. Rather than accede to these requests, the attorney sued the court on behalf of his client. He argued that the court's clerks had no right to request additional documentation in support of his garnishment requests."
In the end, Credit Acceptance won its case before the Michigan Supreme Court. The result according to Judge Williams (as told to Human Rights Watch) is that, "with his court administrator and clerks barred from taking the work on, he said, there is no practical way to apply meaningful scrutiny to the garnishment requests."
Harassing Debt Collection Calls That Never Stop
The typical legal strategy of the few plaintiffs' attorneys who have made it a significant part of their career to go after Credit Acceptance has been to focus on the Telephone Consumer Protection Act of 1991 (TCPA), 47 U.S.C. § 227(b)(3). While there are clearly broader issues with Credit Acceptance's business practices, the fact that these suits are so narrowly tailored to violations of this particular law has more to do with the potential monetary rewards per violation (e.g. per robocall) than it does with actual justice for the individuals filing. After all, if someone is receiving non-stop robocalls from Credit Acceptance's collections department, it generally means that a lot had to go wrong with their car and/or loan for them to get to that point.
Employees describe the Credit Acceptance collections department as a "typical collections job," with "[n]ine hour days with an hour lunch break. Solely spent on a dialer system" with "the best catch...at 5am."82 One noted, "very few of those gaming the system were terminated even when proof of such gaming was provided..."
Given the preponderance of lawsuits and CFPB complaints surrounding Credit Acceptance's telephonic debt collection practices, the company has considerable exposure in this area.
In addition to telephone harassment, Credit Acceptance has an enormous problem with out-of-control contractors hired to repossess cars. That's according to Credit Acceptance, which sued one of them in federal court in 2016 when the contractor started repossessing cars and keeping them.83
Credit Acceptance customers have reported repo men who have damaged vehicles, garage doors, air conditioning units, municipal poles,84 and in one notable instance, a customer alleged that repo men working for Credit Acceptance committed blatant assault that could be fairly described as a hate crime.85
According to its Q3 2015 earnings call transcript, the company repossesses vehicles attached to about 35% of its loans.86 Episodes such as these could lead to significant damage to the company's reputation.
The Credit Acceptance business model has caught the attention of a number of state Attorneys General over the years. So far, since 2014, six different states have begun to investigate the company,87,88,89 as well as one agency of the New York City government90 and two federal agencies.91
The recent case of the New York City DCA92 is particularly interesting as it highlights several problematic business practices discussed in this report. Specifically, the DCA expressed concern over Credit Acceptance's repeated price hikes in the instance of one consumer loan, and the fact that loan documents were not provided as they should have been. On November 29, 2017, the case settled for $300,000 across all parties involved, with the DCA referring to Credit Acceptance and other similar companies as "predatory" lenders in its press release.93
In addition, subprime auto lender J.D. Byrider recently found itself named in a lawsuit filed by Massachusetts Attorney General on September 26, 2017 in Suffolk County Superior Court, in which the government alleged that the company originated credit-damaging loans to consumers that were "set up to fail."94
The Consumer Financial Protection Bureau
When the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010 was passed, the National Automobile Dealers Association lobbied for an exemption to the Consumer Financial Protection Bureau (CFPB)'s regulatory authority. The result is a half-measure that requires car dealerships to actually service vehicles in order to qualify for exemption, per the terms of 12 U.S.C. § 5519.95 To the extent that Credit Acceptance dealer-partners do not service cars themselves, dealers may therefore fall under the jurisdiction of the CFPB.
The CFPB started taking action against "Buy-Here-Pay-Here" auto lenders in 2014.96 Effective August 31, 2015, the CFPB implemented a new rule codified at 12 C.F.R. § 1001 and 12 C.F.R. § 1090 entitled, "Defining Larger Participants of the Automobile Financing Market and Defining Certain Automobile Leasing Activity as a Financial Product or Service," which allowed the Bureau to begin regulating major auto lenders. Accordingly, its jurisdiction likely now extends to Credit Acceptance.
As of October 2017, the CFPB had received 655 complaints specifically concerning Credit Acceptance Corporation. Many of the complaints relate to the company's debt collection calls. Other complaint sub-categories include, "Used obscene/profane/abusive language," "Debt was paid," "Debt is not mine," "Fraudulent loan," "Account status incorrect," "Information is not mine," "Seized/Attempted to seize property," "Billing problem," and "Debt was discharged in bankruptcy."
Given the Trump Administration's extremely skeptical stance on the CFPB and the recent resignation of Richard Cordray as the Bureau's Director, its regulatory power in the immediate future may be purely theoretical. Already, a rumored lawsuit against Santander Consumer USA, Inc., the largest subprime auto lender in the United States, may have been delayed due to the controversial installment of Mick Mulvaney as CFPB Director, which is pending review by the United States District Court for the District of Columbia.97,98 Mulvaney is alleged to have ties to Santander lobbyists.99
Securities Law Meets Patent Law
Credit Acceptance likes to boast about its patented CAPS system. In fact, right on the home page, its Investor Relations website states, "Our Guaranteed Credit Approval program provides automotive dealers with the opportunity to deliver credit approvals to consumers within 30 seconds through the Internet using our patented Credit Approval Processing System (CAPS®.)"100 This statement almost certainly violates SEC Rule 10b-5, 17 C.F.R. § 240.10b-5.
This is because this statement is true only in the most technical sense. While CAPS was once patented as U.S. Patent 6,950,807 B2: System and Method for Providing Financing, that patent is now completely invalid. After a bruising battle before the newly-formed Patent Trial and Appeal Board (PTAB), plaintiff Westlake Services LLC first convinced the USPTO to invalidate claims 1-9, 13, and 34-42101, and then the remaining claims 10-12 and 14-33102.
Credit Acceptance appealed once to the United States Court of Appeals for the Federal Circuit from the district court action underlying the PTAB proceedings, but voluntarily abandoned the appeal in the middle. It also appealed to the Federal Circuit from the second of the two PTAB actions. On June 9, 2017, the Federal Circuit handed Credit Acceptance a defeat, cementing its patent's doom.103
Still, the fight is not over. Westlake Services LLC, also an auto lending business, is continuing to pursue Credit Acceptance in court, alleging that Credit Acceptance filed "sham litigation" while attempting to monopolize the market with an invalid patent.104 The case is ongoing.
In short, competitors are now completely free to copy the Credit Acceptance business model (which they could nearly do before by making minor changes to their business methods to avoid infringement). The questions that investors should be asking are why Credit Acceptance feels so comfortable brazenly lying to investors, and whether copying the company's business model is actually a good idea in the first place.
The stock market has become extraordinarily complacent about corporate wrongdoing. Our research has shown that Credit Acceptance Corporation is reliant on the judicial system to enable its business, yet its model appears to necessitate routine legal violations in order to be sustainable. Furthermore, its public narrative does not match its actual operations. With its vaunted patent-ineligible CAPS technology, Credit Acceptance creates an enivronment in which its used car dealer-partners can and sometimes do commit fraud, charge usurious rates of interest, and sell largely useless VSC products that quickly become the pretext for designed-to-fail loans.105 The company then abuses court systems nationwide (that already favor wealthy corporate litigants) to wring wage garnishment orders out of judges based on faulty interest rate calculations and mathematical errors—which it has sued over to prevent the courts from examining—repossesses the cars it helped sell on fraudulent terms, pockets the balance of the remaining dollars in trust for the aforementioned VSCs, destroys the credit of its borrowers, and then begins the cycle anew with their repossessed and in some cases dangerous vehicles.
The fact is that many Credit Acceptance customers are worse off than they were before coming into contact with the company. The lengths to which Credit Acceptance has gone to exploit its customers has in the past undeniably resulted in short-term gains for investors. In the long term, however, Credit Acceptance is exposed to legislative, regulatory and judicial uncertainty that combined could interrupt or fundamentally alter its business model. That the company refused to answer any of the substantive questions106 we posed to it for this report, despite being publicly traded, does not suggest that its management is acting in good faith.
More broadly, the Credit Acceptance story reveals the importance of transparency in our economy. Car buyers have repeatedly fallen prey to multiple finance charges hidden in standardized Truth In Lending Act paperwork. The company's disclosures in SEC filings have historically fallen far short of the level of detail necessary to understand its business. Limitations in court database systems have even contributed to the curtailed perspective that has prevented policymakers from fully understanding the scope of damage wrought by the company. To avoid future stories like this one, policies must be put in place to ensure that public is better informed. When companies "change lives," it should be for the better.
Appendix A: Re-Examination of the Offshore Re-Insurance Industry
Perhaps the most surprising page on the Southwest Re website is a PDF document entitled "REINSURANCE COMPANY FORMATION, Checklist and Instructions."107 The Southwest Re logo appears in the top right corner of the first four pages, but the remainder of the document is actually a compilation of forms to be sent to several different entities, including a company called Caribbean Management Services Limited, as well as the Financial Services Commission of the Turks and Caicos Islands (TCI).
A reinsurance company is simply an insurance company that insures other insurance companies. Just as a car owner pays an auto insurance company premiums to cede the risk of getting in a car accident, an insurance company pays a reinsurance company premiums to cede the risk of having to pay out claims on insurance policies.
These entities are referred to as PORCs: Producer-Owned Reinsurance Companies. The PORCs that Southwest Re specializes in creating are located offshore for several reasons. First, the Turks and Caicos Islands—the most popular locale for offshore insurance companies in the world according to data from the National Association of Insurance Commissioners (NAIC)—offer lower capital requirements to form a reinsurance company than any locale within the United States. Second, locating a company in the TCI allows the owner to legally claim no tax or a lower tax rate than would otherwise be possible. Since 2003, when the IRS last seriously examined the issue of PORCs, there has been remarkably little scrutiny from U.S. regulators, despite PORCs being identified under the heading of "Abusive Tax Avoidance Transactions" in a November 4, 2004 IRS Exempt Organizations division "Fiscal Year 2004 Accomplishments" briefing.110 Third, corporate filings from offshore jurisdictions such as TCI, Bermuda, St. Kitts and Nevis, and the British Virgin Islands, are virtually impossible to request. Fourth, regulators are extremely lax in offshore jurisdictions. And fifth, possibly through its own offshore entity (which may be Caribbean Management Services Limited), Southwest Re has a thriving business as a third-party administrator of the companies that it facilitates the creation of. Unsurprisingly, Credit Acceptance also has its own TCI subsidiary, CAC (TCI) Ltd.111
According to 2016 Ohio Department of Insurance documents,112 James B. Smith's insurance company, Dealers Assurance Company (DAC), has reinsurance agreements with 259 offshore firms, 240 of which are located in the TCI with names such as Drive Away Confident Reinsurance Co. Ltd. and Sweet Gum Reinsurance Company, Ltd. Each one of these reinsurance companies likely belongs to a different car dealership or dealership owner located in the United States, and since so many are selling Southwest Re products, they are likely to be Credit Acceptance dealer-partners.
What this means on a practical level is that in many cases, with no specific disclosure to the car buyer or any government agency, the ultimate insurers of FASC/Southwest Re VSC products selected and priced by Credit Acceptance and sold by car dealerships are the car dealerships themselves. That is, unless the dealership has not elected to create its own reinsurance company, in which case Credit Acceptance steps in with its own reinsurance company domiciled in Washington, D.C., VSC Re Company.113
As of December 31, 2013, VSC Re was re-insuring over $40 million of VSC premiums insured by Dealers Assurance Company, making Credit Acceptance DAC's largest client by far. Credit Acceptance's SEC disclosures explain only that VSC Re Company is a subsidiary of Credit Acceptance, and that "VSC Re currently reinsures vehicle service contracts that are underwritten by one of our third party insurers."114 That "one" third party is presumably Southwest Re.
Whether the ultimate insurer of the VSC agreement is the car dealership or Credit Acceptance, James B. Smith wins either way. According to filings in a lawsuit between Southwest Reinsure, Inc. and TCI-based reinsurance company Saffa Reinsurance Co., Ltd., Southwest Re provides a sweeping array of "corporate administration," "tax return preparation and filing," "trust account establishment" and "financial reporting" services to its PORC clients, effectively running them from top to bottom on behalf of the dealerships.115 This is consistent with the PDF on the Southwest Re website, which asks car dealers to provide the type and amount of information that a corporate agent would need to incorporate a company offshore.
In the end, the pile of no-tax or low-tax offshore money becomes working capital for the car dealership, supplemented by the reinsurance premiums that the PORC receives in exchange for assuming the risk of the VSC products paying out more claims than expected. So long as Credit Acceptance car buyers can't get their claims for car service paid, however, that risk doesn't seem very large at all. Some PORCs are even themselves reinsured.
The involvement of hundreds of offshore reinsurance companies in the Credit Acceptance business model, as well as Credit Acceptance's own massive reinsurance company, VSC Re, not to mention its TCI subsidiary, gives its entire enterprise the distinct flavor of fraud. Not only is the VSC a largely useless product practically designed to fail, but it also is the basis for multiple levels of reinsurance madness that in a different era was an IRS "abusive tax avoidance" enforcement priority.
At Southwest Re, James B. Smith is well aware that his business model is, to say the least, aggressive. In a revealing March, 2014 presentation at the "Agent Summit" at Caesars Palace in Las Vegas, NV, Southwest Re CFO Bill Bigley made a number of remarkable statements. His slides describe the PORC industry as having been born from obscure lawsuits and IRS regulatory rulings. As his slides state,
"[A] lawyer came to the rescue. His name is Kirk Borchardt and on behalf of one of his clients, he requested a revenue ruling that would allow an 'administrator' to file a tax return as an insurance company. The [IRS] subsequently released TAM9601001, which allowed service contract administrators to file a tax return as an 'insurance company'... Kirk subsequently became the CEO of Dealers Assurance Company (DAC), our affiliated carrier."
More recently, Bigley described that,
"In 2013...we took a controversial position that our 953(d) electing companies did not have to provide this information. Our position was very sound because once our clients' [sic] made the election for their [reinsurance company] to be taxed as a U.S. corporation, then the company was no longer a 'foreign' corporation. We took this position even though the IRS original instructions were contradictory to this interpretation and to the regulations themselves, which some people criticized us for. Fortunately, our position was later vindicated..."116
The statements from the 2014 presentation are hardly Southwest Re's most controversial move. From 2011 until late 2016, both Southwest Reinsure, Inc. and James Bradford Smith personally were named in a separate lawsuit filed by Security Life Insurance Company of America.117 Southwest Re had agreed to set up an offshore TCI reinsurance scheme for Security Life, and the companies had signed a number of agreements when things started to go wrong. In April 2006, Smith suddenly moved Security Life's $1.28 million trust account from INA Trust FSB to Ohio-based Fifth Third Bank. Security Life stopped receiving statements.
After inquiring about the situation, Security Life was told that the new account was simply the successor to the old one. Paragraph 86 of the Amended Complaint filed April 4, 2012 states, "The [Southwest Re] employee who was responsible for sending monthly statements for the INA Trust Account, and who then told Security Life she would send monthly statements for the Fifth Third Trust Account, understood that the funds in those trust accounts were at all times 'pledged' to Security Life and that Security Life was the beneficiary of those accounts."
It wasn't. The lawsuit continues, "Security Life became aware that the assets...were being held at Fifth Third under a 2006 Trust Agreement, to which Security Life was not a party." The next two paragraphs elucidate further:
"The 2006 Trust Agreement is between Ideal as Grantor, Fifth Third as Trustee, and an SRI-affiliated company called First Automotive Risk Retention Group, Inc. ('First Automotive'), as Beneficiary. Security Life is not mentioned in the 2006 Trust Agreement and had no knowledge of that agreement, or the removal of Security Life as the Beneficiary of the trust assets, until March 2011. Smith signed the 2006 Trust Agreement as President of both Ideal and First Automotive. On information and belief, either Smith or Southwest Re, Inc., the holding company of which he is the sole shareholder, is the sole shareholder of First Automotive."
Smith allegedly appropriated a life insurance company's trust fund for use by FASC118 , a key provider of Credit Acceptance VSC products. From there, the situation devolved quickly. "On or about March 30, 2011, Security Life...once again requested that Fifth Third not permit any distributions or withdrawals from the Fifth Third Trust Account until Security Life was able to resolve its issues with Ideal/SRI and enter into a mutually acceptable plan for the handling of the assets held in trust."
Nonetheless, "Late in the day on April 1, 2011, Fifth Third notified Security Life, through its counsel, that a distribution request had been made, apparently by [Southwest Re] on behalf of Ideal and/or on behalf of First Automotive, and that all assets held in the Fifth Third Trust Account had been distributed. The withdrawal requests, once signed by Smith, were submitted to Fifth Third by an employee of [Southwest Re]." In other words, Smith was personally responsible.
Security Life sued for a declaratory judgment under the theory of alter ego liability, arguing that Smith's many companies were all really one and the same: fronts for James Bradford Smith. What Security Life may not have realized was the extent to which all of those companies existed to serve the interests of Credit Acceptance Corporation.
It remains to be seen whether IRS scrutiny of the "controversial" positions Southwest Re has assumed might increase in the future. If so, Southwest Re's entire business model may no longer be viable. With Credit Acceptance so dependent upon Southwest Re VSC products for profitability, and Southwest Re so tied up in the undisclosed and rather questionable affairs of its founder, James B. Smith, both companies could be exposed to considerable liability.
Appendix B: Court Data Methodology
With the exception of Alabama's Alacourt site, most of the court websites we used for this report allowed us to download a full result set as needed. In order to derive statistics regarding Credit Acceptance Corporation's presence in the Michigan 36th District Court, we developed three software programs in the PHP programming language to obtain, convert, and analyze court records in bulk.
Up through November 27, 2017, the first program scraped court cases from the 36th District Court's public Court Case Inquiry website at http://jis.36thdistrictcourt.org/ROAWEBINQ/. We determined that this website serves as a basic front-end for an IBM mainframe system that stores the actual case records. Due to limitations of the website's design, searches for any given entity result in a maximum of 1,000 results, with no ability to limit searches to a specific timeframe or re-order results. Since Credit Acceptance Corporation clearly had more than 1,000 cases filed in the court, we decided to download every possible case from the court since 1995 to find them all.
Case numbers in the 36th District are of the form XXYYYYYY, where XX represents the two-digit year (for example, 99 for 1999 or 17 for 2017) and YYYYYY represents a six-digit serial number starting at 100,000 for each year. We simply requested every single case number for each year from 1995 on until we reached cases filed on or around the last business day in December and began to observe invalid case number errors. Once we had assembled all of the files for each case available, we converted the HTML files to plain text in order to reduce file sizes and ensure accurate string parsing.
As indicated in the report, for earlier case years, fewer cases were available electronically, and for every year we deleted case files that appeared to contain error messages of various sorts and attempted to re-download those files at later times, depending upon server availability. (The court's server appeared to automatically reboot late on Sunday nights, rendering case files we attempted to obtain during those time periods temporarily unavailable.)
With all of the available text files available on our local area network, we then proceeded to analyze the files using several regular expressions based on standard strings we observed in the court dockets. The analysis software was designed to compile statistics for the court as a whole and a subset of cases pertaining to Credit Acceptance Corporation simultaneously.
Our regular expression to find Credit Acceptance Corporation cases searched for the broader case-insensitive string "CREDIT ACCEPTANCE," which potentially could have resulted in false positives (such as a hypothetical company called "General Credit Acceptance"). To mitigate this possibility, we manually reviewed the names of all companies found to match the expression, and did not observe non-Credit Acceptance Corporation matches.
In addition, while calculating the length of time between initial case filing and a "JUDGMENT SATISFIED" docket entry if applicable, we noted many incorrect dates, such as mistyped dates (e.g. 2050 instead of 2005) and system default dates (e.g. January 2, 1989). We manually corrected these errors by searching for time intervals that were negative or excessively large.
We compiled output from our custom software in Microsoft Excel, which we then used to analyze the data. The source code for our analysis software is available at https://www.plainsite.org/realitycheck/caccanalysis.txt.
Appendix C: Unanswered Questions
We never received any response to the e-mail below or to follow-up phone calls placed to the Credit Acceptance Corporation Investor Relations department.
I do research for investors and am working with a team that's writing a report on Credit Acceptance. We have a number of questions for management, which are listed below.
We plan to publish the report on Monday, at which point, it will be available to the public.
8 Zacks Investment Research has since rated CACC as a "Hold."
13 Dealer holdback is discussed on page 4.
17 "The Don Foss Story," Credit Acceptance Corporation Video
20 The Truth In Lending Act, 15 U.S.C. §§ 1601-1667f, begins with a statement of its purpose: "to assure a meaningful disclosure of credit terms so that the consumer will be able to compare more readily the various credit terms available to him and avoid the uninformed use of credit, and to protect the consumer against inaccurate and unfair credit billing and credit card practices."
23 Corveta Houston v. Dealz Auto Trade, Inc. et al, Cuyahoga County, Ohio Court of Common Pleas, Case No. CV 14 827795, Document 1
27 Under the Magnuson-Moss Warranty Act, 15 U.S.C § 2301, "service contracts," defined in section 8, are considered separate and apart from various kinds of warranties. Under the law, a "service contract" is "a contract in writing to perform, over a fixed period of time or for a specified duration, services relating to the maintenance or repair (or both) of a consumer product." VSCs are therefore different than automotive warranties, which are provided by a car's original manufacturer.
29 Credit Acceptance Corporation CAPS Funding Guidelines 2015
33 Corveta Houston v. Dealz Auto Trade, Inc. et al, Cuyahoga County, Ohio Court of Common Pleas, Case No. CV 14 827795, Document 1, Page 50
34 Credit Acceptance Corporation v. Nelson J. DeCuire et al, Superior Court of California, County of San Francisco, Case No. CGC 16 550042, Document 1
48 In 2015, Credit Acceptance began offering dealers the option of using electronic contracts in conjunction with CAPS 2.0, which when implemented changed procedures involving document availability. As before, documents may now be easier or harder to obtain depending upon individual dealer compliance with Credit Acceptance rules.
55 We called back and asked a receptionist.
63 Texas Department of Licensing and Regulation Press Release, January 12, 2004. Archived via the Internet Archive.
72 We found passing references to "wage garnishment" with no detail in two SEC disclosures about vehicle repossessions from 2000 and 2009. See https://www.sec.gov/Archives/edgar/data/885550/000095012401001945/k61157e10-k405.txt and https://www.sec.gov/Archives/edgar/vprr/1001/10012667.pdf, respectively.
73 According to the federal PACER database, Credit Acceptance Corporation is linked to over 37,200 bankruptcy cases.
74 If only available dockets are analyzed, Credit Acceptance comprises 12.01% of the 2017 docket. The comparable figure in 1996 was 15.35%, but only 49.76% of dockets are available on-line from that year, as opposed to 99.6% in 2017.
75 This figure includes 4,000 Alabama lawsuits that we know of, not represented in the corresponding graphs because Alabama's Alacourt system makes comprensive searches of the state's courts prohibitively expensive for researchers.
77 Credit Acceptance Corporation v. Williams et al, Circuit Court of Mobile County, Alabama, Case No. 02-CV-2014-901364.00
86 Less than one year later, in its Q2 2016 earnings call, in reference to "average age...of vehicles" and "repo rates," Credit Acceptance CEO Brett Roberts stated, "Neither of those are numbers that we'd disclose at this point."
89 The Capitol Forum, January 6, 2016, "Auto Financing: Update on Credit Acceptance Investigation; Fair Lending, Deceptive Add-ons, Aggressive Debt Collection and Repossession Are Likely Areas of Focus in State Inquiries"
95 Auto Dealer Law, August 9, 2011, "What Most Dealers Need to Know About the Consumer Finance Protection Bureau." http://www.autodealerlaw.com/2011/08/what-most-dealers-need-to-know-about-the-consumer-finance-protection-bureau/
103 There is a small chance that the Oil States Energy Services, LLC Supreme Court decision could reverse the Federal Circuit's ruling if the Supreme Court finds the entire PTAB to be unconstitutional and retroactively nullifies its decisions.
105 These loans are then coupled with a massively complex offshore tax evasion scheme in which Credit Acceptance is both a knowing participant and an enabler, described in Appendix A.
106 See Appendix C.
109 Although a privacy guard service is set up to protect reinsure.com's WHOIS information against disclosure, visiting http://www.reinsure.com immediately redirects to the Southwest Re website at https://www.southwestre.com.
113 According to previously undisclosed documents from the Washington D.C. Department of Insurance, Securities and Banking (DISB), VSC Re Company was incorporated within DISB itself as a "captive insurer" under Chapter 39A of the D.C. Code on October 28, 2008.
118 Technically, the funds were transferred to First Automotive's Risk Retention Group, a special type of insurance vehicle and yet another James B. Smith entity. It is safe to assume a connection between the Risk Retention Group and FASC.
References on PlainSite
By Bassem Banafa, Aaron Greenspan and Christine Richard
Credit Acceptance Corporation (NASDAQ: CACC), which services auto loans to subprime borrowers, faces a growing solvency crisis of its own. Despite a long history of disputes with its auditors, the SEC and the IRS over its unusual revenue recognition methodologies, Credit Acceptance has managed to convince investors that its business is stunningly profitable and predictable. In reality, Portfolio Program loan pool performance—the company's bedrock—has cratered, with over 60% of loan pools underperforming since 2011. Our analysis indicates that the poor performance of loans originated in 2015 has already eaten through the portion of losses allocated to future dealer payments that shields the company from setting aside loss provisions.
We believe that the company's poorly performing loan portfolio is becoming harder to hide. Our detailed examination of company financials has yielded the following:
Investors have failed to see the warning signs thanks to securitization wizardry, lack of loan portfolio transparency, and management's continual rebuffing of crucial questions by analysts. Fundamentally, we believe that the company's model amounts to modern-day alchemy: an ill-fated attempt to turn lead—last-ditch auto loans—into gold. It doesn't work, but that doesn't stop some from trying.
In Part I of this report, we examined how Credit Acceptance's business model exploits the vulnerable position of its customers. We also introduced some key terms that are crucial to understanding how Credit Acceptance works: Portfolio Program, Purchase Program, Dealer Holdback, Pool, Vehicle Service Contracts (VSCs) and Retail Installment Contracts (RICs). Here in Part II, we examine the company's financial history, structure, and its relationships with Wall Street.
Understanding Credit Acceptance's finances requires an overview of its track record, after which we briefly discuss how corporate financing and securitization works. The mind-numbing complexity of the company's business (which seemingly negates the risk of lending to subprime borrowers), the sheer amount of widely-dispersed information necessary to piece together the puzzle, and especially the company's refusal to disclose vital metrics, are some of the many reasons why Credit Acceptance has been able to stonewall (and in some cases fool) regulators, ratings agencies, investors, and the media. As several individuals we interviewed told us, the company has a long and solid track record in a difficult space, and its lenders always seem to get paid. The seven-point-three-billion-dollar question is whether there may come a day in the near future when lenders don't.
We encourage the public to examine the data we compiled for this report, available at https://www.plainsite.org/realitycheck/cacc/data.zip.
A Checkered Past
Credit Acceptance, its ratings agencies, investors, and analysts often preface any discussion of the company with a reference to its long operating history. Credit Acceptance's executives hold themselves out as grizzled but triumphant veterans of the subprime auto lending industry. Yet while they may be industry veterans, their story is anything but one of unequivocal success.
Credit Acceptance was initially incorporated in 1972 as Auto Finance Corporation, a captive finance and collections company—a type of company that exists for the express purpose of helping another company, similar to (and sometimes overlapping with) a subsidiary—that serviced founder Don Foss's dealerships exclusively. It would be 17 years before Foss began financing outside dealerships. In 1986, shortly before Foss concocted the idea of paying advances to outside dealers in exchange for loan assignments, he appointed Arthur Andersen as Credit Acceptance's outside auditor.
In 1991, Foss hired Richard Beckman, an Arthur Andersen partner, to prepare the company to go public and serve as its CFO and Treasurer. On June 5, 1992, Credit Acceptance went public in a deal managed by William Blair & Company, LLC, an investment bank that itself invested in Credit Acceptance stock. In 1994, William Blair placed one of its own executives, Thomas Fitzsimmons, on Credit Acceptance's Board of Directors.
As a publicly-traded company, Credit Acceptance began an immediate, rapid, and "irrational"1 expansion strategy that included setting up operations in Canada and the U.K. Between 19922 and 1996, Credit Acceptance's dealer network expanded from 748 dealerships to over 5,300, including over one thousand located internationally in the United Kingdom, Ireland, and Canada.
At the end of 1996, Credit Acceptance reported $1 billion in loan receivables, on which the company had advanced $502 million to dealers. They also reported that 34.1%, or $351 million, of their loan receivables were in "non-accrual" status, meaning the company had not collected any material payments from those borrowers in at least five months. The company's allowance for credit losses, reflecting expected losses on over $1 billion in subprime loans, was booked at just $13.1 million.3. This absurdly low provision was possible because Credit Acceptance booked the full principal and interest due on a loan as an asset while booking future performance-dependent payments to the dealer—known as dealer holdback—as a liability. If a loan underperformed, the company would first offset losses against that loan's future dealer holdback, which could be anywhere from 30% to 80% of a given loan's principal and interest balance. With dealer holdback acting as a buffer, a loan would need to incur substantial losses before Credit Acceptance reported reductions to its bottom line.
The advances on these loans were largely funded by unsecured debt and equity. Credit Acceptance's management team members were also the company's controlling shareholders, so there were limited outside shareholder interests or secured parties to curb reckless growth. This led to a situation where Credit Acceptance was not accountable for the loans it was financing as it expanded its dealer network by 620%.
On January 23, 1997, Mercury Finance Co., a subprime auto lender, reported "accounting irregularities" that had inflated its earnings by $90 million. Mercury's stock price plummeted and the company was eventually forced into bankruptcy.4. Mercury's executives were later indicted and convicted for accounting fraud.5. The same week, another auto lender, American Auto Funding Corp., declared bankruptcy,6 and on January 30th, yet another auto lender, Jayhawk Acceptance Corp., reported a $15 million loss, causing its share price to tumble along with Mercury's.7
This two week period in January of 1997 touched off intense scrutiny of subprime lenders, who then fell victim to their own lending practices. One by one they declared bankruptcy or were acquired as the industry consolidated. In February of 1997, after Crain's Detroit Business observed Richard Beckman, the company's then-President and CEO, attempting to sell 98,000 shares of CACC stock amidst the downturn, Beckman publicly stated that Credit Acceptance would not have similar problems.
"Beckman said Credit Acceptance has salted away the kind of reserves Jayhawk hasn't 'for at least the past four or five years' and thus won't need to take special charges. 'All we can do is keep our business and continue to do what we've always done, which is make money using our business model. We've been doing that for 25 years,' Beckman said. You've also got to remember that Mercury is apparently a fraud (situation), and as somebody once said, fraud is not contagious."8
In July of 1997, Credit Acceptance filed a $300 million shelf registration for public issuance of senior notes. The notes9 were rated Baa3 by Moody's and BBB- by Standard & Poors ("S&P"). By October, Credit Acceptance had charged off $467 million in loans, stating that "a new loan servicing [computer] system" had determined that they were uncollectible.10,11
The company's loan losses were so large that they exceeded the underlying dealer holdback buffer in many cases. As a result, the company had to book an $85 million increase to its allowance for credit losses in 1997, wiping out substantially all of its net income. Moody's dropped the credit rating on Credit Acceptance's Senior Notes from Baa3 to Ba2, while S&P lowered its rating from BBB to BB-. The company's unsecured lenders responded by increasing its interest rate by 0.50%.
On October 23, 1997, Beckman explained in a Credit Acceptance press release that the adjustments being made were, in fact, a positive development:12
"The new level of information available to us will allow us to better identify dealers that have not provided us with profitable business, allowing us to focus our resources to enhance our services to our remaining dealers. We believe that this presents opportunities for the Company to grow while mitigating risk that is inherent in our business. At the same time it allows us to manage our cash more efficiently without as much dependence on borrowings to support contract originations."
Don Foss also chimed in regarding this process enhancement, touting the ability to identify which dealer originated an individual loan, as many loan servicers had done on paper for decades.
"...the Company has recently deployed proprietary management information and reporting software which allows for the detailed analysis of the Company's portfolio on a dealer-by-dealer basis, replacing technology which, while sophisticated in the consumer finance industry, did not allow for the level of analysis that is now possible."
An analyst quoted in Crain's Detroit suggested that Credit Acceptance had deeper problems than accounting for costs: "I think they...compromised some basic tenets they had been using before in terms of the quality of the loans and the quality of the dealers they were lending to."13
Years later, management would concur with this analyst, stating, "In the mid 1990's, we had an unsustainable business model... We faced irrational price competition and responded with irrational pricing of our own."14
The downward spiral continued through 1998, starting with a series of shareholder lawsuits in January.15,16. (The lawsuits were settled in 2001 with no acknowledgement of wrongdoing.. Credit Acceptance continued its charge-offs, writing off an additional $450 million in loans. The company also reduced its dealer network from 5,385 dealers at the end of 1996 to 1,729 at the end of 1998.17 There was a corresponding decrease in new loans, from $965 million in 1996 to $580 million in 1998.
On March 23, 1998, Credit Acceptance filed its 1997 year-end financial statements, reporting continued fallout related to its post-IPO excesses. On March 31, Beckman resigned as President of Credit Acceptance. Two weeks later, the company's auditor, Arthur Andersen LLP, resigned as well.
"On April 16, 1998, Arthur Andersen LLP informed the Company that the client-auditor relationship between Arthur Andersen LLP and the Company had ceased. Although the Company had not communicated such fact to Arthur Andersen LLP, this notification followed the Company's determination to seek competitive bids from independent accounting firms, including Arthur Andersen LLP, with respect to the engagement of independent accountants to audit the Company's financial statements for the year ending December 31, 1998."18
On June 24, 1998, Moody's downgraded Credit Acceptance's senior notes again, from Ba2 to Ba3, pushing Credit Acceptance to deregister the notes and seek non-public financing.
The following year kicked off with an Internal Revenue Service audit of Credit Acceptance's tax filings for the years 1993 through 1996, later expanded to include all years through 1999.19 Credit Acceptance continued to "discover" that substantially all of the loans it had originated between 1995 and 1997 were unprofitable, taking a $47 million charge directly against their net income. The company continued to shed dealer-partners, reducing its network from 1,729 in 1998 to 1,191 at the end of 1999. Credit Acceptance announced that it had seen the error of its ways and was merely cleaning up old, problem loans—the kind it would not make going forward.20
In 2000, Credit Acceptance rolled out CAPS, which it claimed would "significantly streamline and enhance the credit approval process" by handling applications over the internet.21 (We discussed CAPS in detail in Part I of this report.)
By 2001, the IRS audit, which had taken three years, had expanded again to include the entire period between 1993 and 2000. Credit Acceptance was required to amend its tax returns for most of that period, in addition to paying a settlement for back taxes owed. That year, the company made a series of appointments to its management team including:
Credit Acceptance reported "Forecasted Collections" for the first time in its 2001 annual report via the shareholder letter, outside of its audited financial statements. This metric was established to replace nearly every credit quality metric ordinarily provided by lenders. According to Credit Acceptance, the metric (somewhat unbelievably) encapsulates all loan evaluation criteria including actual collections. Forecasted Collections immediately began declining and Credit Acceptance blamed the decline on "a difficult collection system conversion" that "negatively impacted collection results."22
In 2002, company President McCluskey announced new imperatives: exclusive territory and guaranteed credit approval through the internet. He also announced that the company would only add 200 high-quality dealers per year through 2011.23
"'The growth of our business is not based on the sheer volume of dealers we approve to join our dealer-partner network,' explained Mr. McCluskey. 'We're much more interested in selecting a very limited number of the best dealers in each market and partnering with them to build their businesses through our Patent-Pending guaranteed credit approval program.'"
Credit Acceptance began using its newfound ability to forecast collections to adjust its accounting. Rather than charge off loans after nine months of delinquency, the company began charging them off when it no longer "forecasted" incoming loan payments, including "post-repossession collections," which could take decades.24 Six months after implementing this policy, on June 27, 2003, Credit Acceptance issued its first asset-backed securitization (ABS). With a new accounting methodology, a new growth strategy, and new management, Credit Acceptance pursued its next hurdle to changing its own life: financing.
Problems with the SEC and Auditors
Credit Acceptance's Q2 2003 financials foreshadowed potential problems with loans originated that quarter:
"[T]he risk of an unintended adverse change in the profitability of loan originations is increased during periods of high growth. The growth rate experienced in the second quarter of 2003 is higher than the Company's expected long-term growth rate."25
On October 30, 2003, the company's warning proved prescient when it announced that its Q3 2003 earnings release would be delayed as a result of "a desk review of its periodic reports by the Securities and Exchange Commission." After the SEC's review, Credit Acceptance reversed its "forecasted collections" charge-off policy.26 The company made other substantial changes to its financial statement reporting and issued the rest of its 2003 financial statements without any additional delays.
Parallel to Credit Acceptance's troubles with the SEC, Credit Acceptance's new independent auditor, Deloitte, began having regulatory troubles of its own. The Public Company Accounting Oversight Board (PCAOB), an outgrowth of the Enron and WorldCom accounting scandals, launched its inaugural inspections of U.S. public accounting firms by targeting the Big Four, including Deloitte.27 Naturally, Deloitte began cleaning up its audits and reviews, and gave Credit Acceptance a second look.
On May 11, 2004, referring to Deloitte, Credit Acceptance announced that it would be "unable to file its quarterly report on Form 10-Q within the prescribed time period as additional time is needed to complete a review of two accounting issues with its independent auditors."28
With the PCAOB looking over its shoulder, Deloitte ascertained that Credit Acceptance's forecasting methodology included the assumption that the company would retain 100% of dealer holdback payments. Additionally, Deloitte learned that Credit Acceptance, which elected to pool loans together (a common practice), was then grouping loans with varying terms and allowing new loans to be added to those pools—less common practices that were violations of U.S. Generally Accepted Accounting Principles (GAAP), a set of accounting standards endorsed by the SEC that publicly-traded companies are encouraged to use in order to promote financial understanding and transparency. Credit Acceptance ceased these practices and took the opportunity to announce the changes as an enhancement to its business.
"During the first quarter of 2004, the Company enhanced its methodology for applying SFAS No. 91 such that finance charge income and the amount of the provision for earned but unpaid income at the time a loan is transferred to non-accrual status can be calculated for each individual loan. Prior to the first quarter of 2004, the Company...ma[d]e various assumptions and estimates which impacted the timing of income recognition and the classification of finance charge revenue and the provision for earned but unpaid revenue... The Company believes that these enhancements improve the precision of the Company's calculation of finance charge revenue and the provision for earned but unpaid revenue."29,30
On August 25, 2004, despite the financial reporting delays, Credit Acceptance closed another $100 million term securitization—a securitization instrument with a defined term, such as 24 months, or in the case of warehouse lending, a facility with a defined term—with Wachovia, also insured. This time, Credit Acceptance's ABS received a AAA rating from S&P. It would be the company's last securitization for two years.
The next day, the PCAOB issued its first inspection report of Deloitte, finding a number of "significant issues" that had resulted in material restatements by Deloitte clients. The PCAOB specifically pointed out that "consultation beyond the local office level was warranted" for significant or judgmental accounting issues that really needed the attention of the "National Office."31,32
Also in August, Credit Acceptance began taking questions directly from investors through its website to circumvent its auditors. The website was used by company management to publish non-GAAP financial measures and forward-looking statements outside of company financial statements and without Deloitte's input.
In December of 2004, American Institute of Certified Public Accountants (AICPA) Statement of Position (SOP) 03-3, later called Accounting Standards Codification (ASC) 310-30, went into effect overriding ASC 310-20. Both standards involve different approaches to timing for recognition of loan income and losses. Generally, an ASC is a codification of U.S. GAAP maintained by the Financial Accounting Standards Board (FASB), with input from various accounting-related bodies such as the American Institute of Certified Public Accountants (AICPA). Previously, a standard set by the AICPA was referred to as an SOP.
On March 10, 2005, Credit Acceptance announced that it had not completed its FY 2004 financials and that it was restating its FY 2003 and Q1 through Q3 2004 filings because it had "...incorrectly accounted for income taxes in prior periods related primarily to its foreign subsidiaries." In reality, Credit Acceptance's audit team at Deloitte had referred the question of Credit Acceptance's accounting, particularly with regard to ASC 310-30, to Deloitte's National Office. The National Office responded on April 8, 2005: Credit Acceptance could not book loans as an originator—it had to record them as a lender to dealers.33 The change would have a considerable negative impact on its balance sheet.
Credit Acceptance balked and contacted the SEC to make its case and resolve the issue. As the dispute unfolded, the company took two measures to increase the number of dealers enrolled in their program: a pilot program that later developed into the Purchase Program, and an enrollment option that allowed dealers to defer the up-front $9,850 fee. The fee was again made non-refundable.
Two months later, on June 24, 2005, the SEC concurred with Deloitte, and on the same day, Credit Acceptance fired Deloitte. By July 20, Credit Acceptance was several months past NASDAQ's deadline for issuing its FY 2004 financials. The company declined to pursue additional extensions of time and was delisted.34
The company abandoned McCluskey's more qualitative strategy in favor of "just looking to expand the number" of dealers. Credit Acceptance would later attribute its ability to grow originations despite a difficult "competitive environment" during this period to dealer growth. The ever-mercurial and unpredictable competitive environment would become a favorite scapegoat of company executives.
On January 30, 2006, Credit Acceptance issued its FY 2004, Q1 2005, Q2 2005, and Q3 2005 financial statements and on March 10, 2006, Credit Acceptance issued its FY 2005 financial statements. In these new financial statements, Credit Acceptance restated its previously filed financial statements for years 2000 through 2003. The company also replaced Deloitte with a smaller firm, Grant Thornton, that had been absorbing former Big Four clients.35
Over the next year, Credit Acceptance convinced the SEC Office of the Chief Accountant and Grant Thornton to sign off on its three pet issues: "Forecasted Collections" recognition of delinquent loans, the assumption that the company could retain 100% of dealer holdback payments, and the use of loan pools with varying terms and constant new loans. This non-standard trifecta was summarized by the company as an approach "similar to" ASC 310-30.36 Credit Acceptance would later attempt to clarify this in correspondence with the SEC:
"While our loan accounting policy is analogous to certain concepts within ASC 310-30, we do not believe ASC 310-30 is the authoritative guidance on how to account for our loans. Our policy for aggregation is consistent with the approach that was developed and agreed upon among us, our external auditors and the Office of the Chief Accountant Staff in 2005 when we changed our loan accounting methodology, as reflected initially in the restated financial statements the Company filed with the SEC in early 2006."
This stance put the company in a position where it was effectively arguing to be exempt from GAAP accounting rules—a practice occasionally allowed by the SEC for good reason, but only with proper disclosure to investors.
On April 26, 2006, Credit Acceptance was relisted on the NASDAQ and closed at $25 a share, a 15-month high. Foss used the high stock price as an opportunity to have the company repurchase $30 million of his stock. In June, company Director and former Enron executive Daniel Leff resigned. Two months later, as the dealer network approached 2,200, McCluskey did as well. By the end of 2006, Foss extracted over $100 million through share repurchases, forcing Credit Acceptance to escalate its securitizations and originations.
In February of 2007, the IRS began another audit of Credit Acceptance, this time for its filings for 2004, 2005 and 2006. (Later, the IRS audited the company's 2007 and 2008 returns.)
In March of 2007, Scott Vassalluzzo, a long-time shareholder,37 filled a vacancy on the Credit Acceptance Board of Directors due to a resignation. Simultaneously, the company expanded its Purchase Program, acquiring loans from dealers it had never had a relationship with before, and ending 2007 with nearly 20% of originations from purchased loans. As economic storm clouds gathered, these purchased loans were used to provide collateral for Credit Acceptance's last insured securitizations: Credit Acceptance Auto Loan Trust (CAALT) 2007-1 and 2007-2. The company's ability to continue issuing securitizations to finance its business was hamstrung by the lockup of credit markets as the mortgage crisis developed.38
The 2008 Financial Crisis
In 2008, JP Morgan Chase dropped out of Credit Acceptance's sole warehouse lending facility, Warehouse Facility II, and Wells Fargo, the remaining investor, cut Credit Acceptance's borrowing limit by $100 million and increased its interest rate by 35 basis points, or 0.35%. Then, the company established another warehouse lending facility with Fifth Third Bank, Warehouse Facility III, with a borrowing limit of $50 million. By the end of the year, Credit Acceptance had used all of it. In addition, Credit Acceptance's last insured securitization. CAALT 2007-2, was downgraded from AAA to A- by S&P, and to A3 by Moody's. The company continued amending its forecasting methodology and stated that these changes were the basis for an 81.2% increase in its provision for credit losses.
On April 18, 2008, Credit Acceptance issued its first uninsured Term ABS, CAALT 2008-1, a single class of notes with a 12- month revolving period which received an A rating from S&P. Without an insurance policy, the cost to provide sufficient credit enhancement for an AAA rating, through either subordination or overcollateralization, was prohibitively expensive.
Through 2009, Credit Acceptance was unable to fund sustainable growth as the effects of the 2008 financial crisis permeated the marketplace, and it reduced its forecast for collections.39 In response to the financial crisis, the Federal Reserve Bank of New York launched the Term Asset-Backed Securities Loan Facility ("TALF"), which provided low-interest financing for the purchase of "qualifying ABS.. One of the qualifications was a AAA rating by at least two of the major ratings agencies, referred to as Nationally Recognized Statistical Rating Organizations, or NRSROs. As early as March 25, 2009 Credit Acceptance was pursuing a AAA rating on its subordinated and overcollateralized securities in order to qualify.40 On December 3, 2009, Credit Acceptance closed CAALT 2009-1 with an 18 month revolving period, which received an AAA rating from S&P and DBRS, Inc.
As part of a 2010 senior notes issuance, Credit Acceptance agreed to hold quarterly earnings conference calls for the first time. Historically, the company had refused to hold such calls, stating that filings and information on the investor relations portion of their website were sufficient to communicate with investors. According to one newspaper report, "'They're not going out to meet Wall Street. They don't do conference calls. If you want to meet them, you gotta go there, or send them an e-mail,' said Randy Heck, a general partner at the Connecticut-based Goodnow Investment Group L.L.C., an institutional investor that manages about $600 million and has owned Credit Acceptance stock 'for seven or eight years.'"41 Finally, after 18 years on the NASDAQ, the policy changed:
"As our long-term shareholders know, this is our first quarterly earnings call. We have not held quarterly calls historically... However, as a result of our recent senior note offering, we agreed to hold quarterly calls. As we gain some experience with this new process, we hope to find a way to make it a productive use of everyone's time."
Much of the analysis in this report starts in 2011, roughly when the company began operating as we know it today.
In 2012, the Monthly Reporter of the Detroit Chapter of the Institute of Internal Auditors reported that Hal Hebel joined Credit Acceptance as the Director of Internal Audit. From 1997 through 2012, he had worked in the Southfield, Michigan42 office of Credit Acceptance's newest auditor Grant Thornton, where he "specialized in audits of publically held financial institutions and specialty finance companies." His hiring continued a Credit Acceptance tradition of bringing in management from financial partners, starting with William Blair & Company's Thomas Fitzsimmons joining the Credit Acceptance Board, and followed by Treasurer Doug Busk joining in November 1996 from Comerica Bank.
By 2014, founder Don Foss continued a pattern of extracting funds from the company through share repurchases, and escalated his involvement in CARite (one of Credit Acceptance's top accounts), poaching Credit Acceptance's Chief Administrative Officer, John Neary, to be CARite's CEO. In 2015, as he was being inducted into a used car "hall of fame," Foss suggested that he wanted to replace Credit Acceptance with CARite:
"If I could do it all over again, starting today, how would I want to do it. The answer was that I'd want to have one brand. I'd want to sell better cars, newer cars, lower miles with a better customer experience... A few years ago, I started another venture called CARite. It's a network of branded used-car dealerships with a captive subprime leasing company..."43
When the Board of Directors caught wind of Foss's intention to take CARite public, it was forced to issue a disclosure regarding "actual or perceived conflicts of interest."44 Viewed by his own company as a nuisance and a liability, Foss was forced into an early retirement and sold 660,000 CACC shares for $128 million.45
How The Company Works
Credit Acceptance issued the most current versions of its Portfolio Program and Purchase Program contracts for used car dealers in 2007. These contracts, called Dealer Service Agreements, outline the relationship between the company and each dealer, how loans are to be assigned between them, and how funds flow between borrowers, Credit Acceptance, and each dealer.
Each Dealer Service Agreement also designates a Dealer Lot Number (DLN), a three-character identifier (e.g. "AB3") used to group loans throughout their lifecycle: application, forecasting, advance/purchase, collections, and securitization. Accordingly, pools of loans and dealer ratings in Credit Acceptance's internal systems correspond to a specific DLN. Although a dealer may have more than one DLN, each DLN is considered a unique dealer for accounting purposes. Subsequent references to a "dealer" in this report will therefore indicate a unique dealer as identified by DLN.
As mentioned previously, a DLN may have one or more associated loan pools based on the associated dealer's program participation. According to Credit Acceptance's Dealer Service Agreement there are three types of loans, described below.
Portfolio Program: Dealer Loans
Dealer Loans are loans to used car dealers secured by Retail Installment Contracts (RICs), servicing rights46, and cross-collateralization47 rights. Dealer Loans are placed into sequential dealer pools as a dealer "caps" or closes each pool of 100 loans. When a dealer closes its first pool of 100 loans, the dealer becomes eligible for "Portfolio Profit"48 and "Portfolio Profit Express"49 payments later on. These payments constitute "dealer holdback," and are subject to Credit Acceptance's cross-collateralization rights, which allow the company to retain (i.e. to avoid paying) up to 100% of any dealer holdback under certain circumstances, such as not closing a pool.50
Dealer holdback payments are calculated on a pool-by-pool basis. For dealer rating and accounting purposes, dealer pools are grouped by DLN to facilitate dealer-by-dealer calculations. This is important because a dealer may stop receiving dealer holdback payments on all pools if any pool's loans underperform. The company's response to a comment letter by SEC staff states, "The amount of these payments, if any, are dependent upon collection performance of all of the dealer's underlying consumer loans, not just those in a given pool."51
Purchase Program: Purchased Loans52
Purchased Loans are payments to dealers for outright purchases of RICs. Purchased Loans are placed into a single pool that is never capped. Purchased Loan pools are uniquely identified by their originating DLN for dealer rating purposes. For accounting purposes, every Purchased Loan is aggregated by the month and year of vehicle purchase, regardless of DLN. In effect, every month and year group (e.g. April 2017) is one giant pool. Credit Acceptance does not make further dealer holdback payments on Purchased Loans once they are placed into a Purchased Loan pool. Collection shortages are borne by the company.
Collection Only Loans
Collection Only Loans are a minor part of the business and are placed into a single pool that is never capped. The company's interest in the loan is limited to servicing; after servicing fees are deducted, loan proceeds are paid. Credit Acceptance does not report these loans separately on its financial statements.
Unless otherwise noted, subsequent references in this report to Dealer Loan pools or Purchased Loan pools will refer to pools as they are grouped for accounting purposes: by DLN for the former and by vehicle purchase month and year pairing for the latter.
Credit Acceptance's underwriting process has three key components:
Like many large companies, Credit Acceptance has different ways of financing its short-term and long-term capital needs. To facilitate its lending operations, the company must borrow heavily from capital markets.
Credit Acceptance finances its lending in the short-term through a secured line of credit and warehouse lending facilities.53. The secured line of credit is provided by a syndicate of financial institutions. Many of these financial institutions have relationships with Credit Acceptance in addition to their participation in the syndicate. The same is true for Credit Acceptance's warehouse lenders.
A secured line of credit does not typically involve the examination of the quality of the specific loans that make up the collateral. In contrast, warehouse lenders receive a list of loans and make lending decisions based on the quality of those loans (the collateral). Secured line of credit lenders tend to focus more on the credit quality of the borrowing party itself. On the other hand, one benefit of using a warehouse lending facility is that borrowing capacity is paid for in advance, so there is a "use it or lose it" aspect to the arrangement.
Credit Acceptance has structured its warehouse lending facilities as follows:
Credit Acceptance finances its lending operation in the long-term through asset-backed securities. Prior to the 2008 financial crisis, Credit Acceptance issued "AAA" rated ABSes that were credit-enhanced with bond insurance. Once the bond insurance marketplace collapsed during the financial crisis, this was no longer an option, and the company's initial rating for its highest tranche fell to "A."
It is crucial to note that ABS ratings only reflect the risk that a particular ABS might not pay back amounts due to bondholders, often based upon a rating agency's formula represented in a spreadsheet. ABS ratings do not necessarily reflect the creditworthiness of the sponsoring company, its business practices, or its management. Credit Acceptance is itself rated BB even while its top-tier ABS tranches are rated AAA.
Table 1: Credit Acceptance ABS Issuances
From 2008 on, Credit Acceptance began issuing ABSes with other credit enhancements including subordination, overcollateralization, excess spread, and a reserve account. Attorney and subprime auto specialist Joseph Cioffi explains how these credit enhancements work (and sometimes don't) on his Credit Chronometer blog:55
"Subprime auto ABS issuers use the same types of credit enhancements that were common in subprime RMBS, including [overcollateralization], excess spread and subordination. [Overcollateralization] exists where additional collateral (pooled loans) is placed in the deal so that the aggregate principal of the collateral exceeds the aggregate principal of the certificates issued to ABS investors. The purpose of [overcollateralization] is for inflowing principal and interest payments on the [overcollateralization] to allow payments to continue to investors even when borrowers default or are late on payments. Excess spread is the difference between the interest on the collateral and the coupon paid on the certificates held by ABS investors. The idea is that funds provided by the excess spread will be available to help pay investors if borrowers default. Subordination reflects where losses are first allocated to lower tranches before reaching senior levels."
After 2011, Credit Acceptance's use of securitization grew at a rapid clip, from approximately $330 million outstanding to over $3 billion outstanding. ABSes also grew as a percent of Credit Acceptance's total outstanding debt, from 33% to nearly 80%. Credit Acceptance issued 15 ABSes between 2011 and 2017.
With two exceptions56, Credit Acceptance has structured its ABSes accordingly:
Table 2: Credit Acceptance ABS Ratings and Initial Purchasers
The securitization process is complex, but the benefits ultimately outweigh the costs: Credit Acceptance is able to raise hundreds of millions of dollars at a time from Wall Street to fund its lending operations. With each new ABS, the company typically involves at least two third-party ratings agencies to evaluate underlying asset risk. As illustrated by Table 2, at various points in time, Credit Acceptance ABSes have been rated by S&P, DBRS, Kroll and Moody's. Each ratings agency has its own proprietary mix of quantitative and qualitative measures that it employs to assign a grade to each tranche of a security. These methodologies have varying levels of transparency.59
Measuring Credit Acceptance's Solvency
One way to evaluate the solvency of a going concern is to ask the question, "Were the entity liquidated tomorrow, would the sale of its assets generate enough cash to cover its liabilities?. In the case of Credit Acceptance, we believe the answer is "no," for reasons that should be clear by the end of this report. In the absence of a comprehensive and accurate liquidation analysis, which would require data Credit Acceptance does not make available, we offer an analysis that examines known data points.
Empirical Data Method
Cash Position: Almost No Unrestricted Cash On Hand
For a multi-billion dollar lender with over a thousand employees, Credit Acceptance Corporation has almost no unrestricted cash on hand. It reported $4.9 million in cash as of September 30, 2017, down from $14.6 million at the end of 2016.60
According to data from FinViz.com,61 as of January 27, 2018, the price-to-cash ratio for CACC is an astronomical 1,459.42, meaning that the company's cash on hand represents 0.068% of its $7.284 billion market capitalization. It has the highest price-to-cash ratio by far of any American publicly traded company in the financial sector offering credit services.62 The average price-to-cash ratio for the 42 credit services companies in the financial sector is 62.42; if Credit Acceptance is excluded, that average drops to 25.66. The median price-to-cash ratio for the credit services industry is just 6.32. Put another way, on average, other companies in the same industry have an amount equivalent to about 3.8% of their market capitalization in unrestricted cash—more than fifteen times as much as Credit Acceptance has.
It is of course important to view any single stock metric in context. Credit Acceptance has recently benefited from a high stock price in historical terms, which has inflated the numerator (price) in the price-to-cash ratio. Even with a stock price half of its peak value, the company would still boast the highest price-to-cash ratio in the credit services industry (where many companies have similar price-to-earnings ratios). Yet with $4.9 million cash on hand and 1,609 "team members" in 2016,63 it is not even clear how the company is making payroll without using borrowed cash, let alone financing other core operations. This has not been explicitly disclosed to investors.
That the company is operating on so little unrestricted cash is consistent with the steady drumbeat of news about the state of the subprime auto industry in general, for as fewer borrowers pay their bills, cash is harder to come by. As of December 21, 2017, Bloomberg News reported, "Subprime Auto Defaults Are Soaring, and PE Firms Have No Way Out."64 A month later, a headline from the same publication read, "Worthless Auto Trade-Ins Signal Riskier Loans."65
Were the company to be held broadly liable for Truth In Lending violations of the kind discussed in Part I of this report beyond those already identified by the New York City Department of Consumer Affairs (DCA), it is not clear that Credit Acceptance would even have the cash on hand to pay the penalties. So far, the company has emerged relatively unscathed, but only due to the narrow scope of investigations. The DCA, which focused on specific used car dealers, forced Credit Acceptance to make restitution on 12 individual loans in amounts ranging from $3,500.00 to over $11,000 per loan, and threatened a future civil penalty of $500.00 for future violations of the Consent Order's terms.66 Future Consent Orders may not be so forgiving.
Dependency Upon Litigation Efforts: Steadily Increasing
For decades, Credit Acceptance has filed lawsuits in state and local courts to effect delinquency collections via garnishment orders, as described in Part I of this report. Further analysis of the demand amounts specified in those lawsuits' dockets, all of which are from Michigan's 36th District Court (which covers Detroit), shows that the amounts being fought over have grown almost linearly since 1995 and have tripled to an average of about $9,000 per lawsuit over a twenty-year period, far exceeding inflation. This steady increase in demand amounts has coincided with a massive increase in the number of lawsuits filed in the District since 2008.
If the Credit Acceptance business model were succeeding, the amount of money it sought to recover from lawsuits would be going consistently down or perhaps staying level, but not rising. The company has led shareholders to believe that its Forecasted Collection Rate—the lynchpin of its business in many ways—is generally accurate with immaterial variances. Yet the Forecasted Collection Rate has fallen consistently since 2009, from 79.4% to 65.1% in 2016,67 while average litigation demand amounts have increased by 50% (alongside frequency) during the same period, giving the appearance that the company's Detroit litigation level correlates inversely with collections success. We have already shown in Part I that a surprising number of borrowers take more than twenty years to fully discharge their garnishment orders. Therefore, if litigation is inversely correlated with collections success and it is a notoriously slow process, one might also ask why the company continues to litigate more and more.
Vehicle Repossessions: 35% of Loans Already Defaulting, Maybe More
Over the past decade, Credit Acceptance has offered few disclosures regarding repossessions. In 2015, Treasurer Douglas Busk confirmed that approximately 35% of all vehicles financed by Credit Acceptance are repossessed.68 In other words, at that time, Credit Acceptance had an approximately 35% default rate69 on its loans, roughly consistent with its reported 67.8% Forecasted Collection Rate that year.70 In 2016, Credit Acceptance stated that it books loans with 120-month lives regardless of actual loan term,71 allowing the company up to twice a loan's contractual term to perform repossessions and deficiency collections, without write-offs.
S&P Global Ratings has noted that peak losses to CAALT ABSes occur around month 26.72 On average, the company repossesses vehicles within 15 days of a borrower's second missed monthly payment. This is consistent with our analysis showing that repossession sales within the first three years of loan origination rarely recover the principal and interest balance on a RIC and repossession sales within the first year generally result in a deficiency balance and a loss to Credit Acceptance. Viewed another way, as loan periods have grown over time, in turn lowering average monthly payments, the company's ability to cover its advances to dealers has also dropped.73
Table 3: PlainSite Analysis of Performance Program RIC Per Loan Average Figures
In Table 3, we compiled data that shows when loans are paying off. The "Payback Period" is the time necessary to repay Credit Acceptance's dealer advance from monthly borrower payments.74 The "Profit Period" is the time required in subsequent months to fully cover the spread (the total of the loan's principal and interest, its forecasted loss, and the advance).
Used car depreciation rates aren't helping Credit Acceptance. Since many repossessed cars are sold at auction, repossessions bring in more cash for the company when used car prices are high overall (or when depreciation is low), and less when used car prices are low (or when depreciation is high). According to CEO Brett Roberts, repossession sales between 2009 and 2011 brought in proceeds much higher than normal,75 possibly due to the federal Car Allowance Rebate System (CARS) program, known as "cash for clunkers," signed into law in July 2009.
By 2016, used car prices began to decline sharply. In the company's Q3 2016 earnings call,76 CFO Ken Booth stated that used vehicle values were dropping, and that repossession proceeds were a small portion of forecasted cash flows.
"What we've seen so far this year is a steeper decline in terms of the depreciation on the vehicles in our portfolio...probably the steepest declines this year going back to, I think 2008 was the last time we saw depreciation this steep..."
In the company's Q4 2016 earnings call, Roberts also stated that repossession proceeds, as a percent of Forecasted Collections, had been reduced by approximately "300 basis points," or 3.00%, since their peak in 2011. Should used car prices fall even further, the company could suffer an additional hit to revenues from auction sales.
Dealer Holdback: Potentially Already Zero for 2015
Since dealer holdback is paid to dealers based on actual collections long after consumer loans are originated, it acts as Credit Acceptance's buffer against issuing provisions for losses. Once bad loans have eaten through dealer holdback, in which case actual collections must be worse than predicted according to the Forecasted Collection Rate, the company is on the hook for losses.
For example, at the outset of the loan origination process, Credit Acceptance might advance $3,000 on a three-year loan expected to generate $10,000 in total cash flow. The company therefore calculates a yield, or an internal rate of return, based on $3,000 of "expected cash flow.. The remaining $7,000 (ignoring Credit Acceptance's 20% servicing fee for the sake of simplicity) is considered projected dealer holdback payments. In Year 2, based on payment history, Credit Acceptance determines that total cash flow on this loan will only be $3,100. Projected dealer holdback payments are now only $100, but expected cash flow remains the same at $3,000. The company is not required to book a provision for loss because the two components of net asset value, expected cash flow and yield, have not changed. In other words, yield is not adjusted downward and expected cash flow is not impacted unless loan performance is so bad that the dealer holdback payments buffer is wiped out.
It appears that Credit Acceptance's unusual prescience in its ability to forecast losses is not so much a factor of its methodology as it is the wide margin of error it has allowed itself. Using data from SEC filings that the company began reporting in 2011 (see Appendix D, Table 15), we attempted to isolate the impact of dealer holdback on profitability. According to our analysis, for the cohort of loans outstanding at the end of 2016, Credit Acceptance forecasted losses of nearly $6 billion. Over $5 billion was taken out of future dealer holdback to maintain Credit Acceptance's spread, leaving a $1 billion buffer remaining.
Table 4: PlainSite Analysis of Dealer Holdback First Loss
In Table 4, the "Percentage Variance from Forecasted Collection Rate at Origination to Cause First Loss" is calculated as forecasted dealer holdback payments for the quarter multiplied by the percentage of forecast not realized (as indicated by the inverse of the figures in the latest quarterly data, presently page 38 of SEC Form 10-Q for Q3 2017), all divided by the aggregate RIC principal and interest volume for the quarter (which is the disclosed quarterly advance total divided by the disclosed advance percentage). "Disclosed Actual Variance" is indicated on page 20 of SEC Form 10-Q for Q3 2017. "Cash Flows at Risk of Loss" is calculated as the total RIC principal and interest volume for the quarter multiplied by the Forecasted Collection Rate for that quarter to result in a dollar amount of Forecasted Collections, multiplied once again by the percentage of forecast not realized.
The end result: these calculations indicate that dealer holdback has already been wiped out for loans originated in 2015, and may soon be wiped out for loans originated in 2014. This sobering reality demonstrates why getting the mysterious and proprietary Forecasted Collection Rate right is so crucial to Credit Acceptance's success. In the words of CEO Roberts, "I think the thing to focus on there is the performance relative to our initial expectation. So we don't really care what the absolute collection rate is as long as it is what we thought it was going to be when we wrote the loan."77
Since 2011, at least 60% of Dealer Loan pools, consisting of all loans originated by a dealer, have underperformed.78 How is this possible when Credit Acceptance has also reported increases to its Dealer Loan forecasts every year since 2011. The answer lies in the same misleading tactic that Credit Acceptance has used since their IPO: burying losses in dealer holdback.
Forecasted Collections Variance: More Material Than Reported
Credit Acceptance notes that longer loan terms are consistent with lower collection rates,79 which aligns with the fact that most successful collections take place in the first few months after a loan originates. Borrowers tend to be reluctant to make payments once vehicles start to exhibit problems requiring service—a frequent issue fully described in Part I. (In 2017, the average initial loan term increased yet again from 53 months to 55 months.80)
In 2016, CFO Ken Booth reported that Credit Acceptance does not extend loan terms for borrowers having trouble making payments, stating "[we do not] extend or rewrite contracts, except where we're required to by law, bankruptcy, FCRA, things like that. So we don't, we just don't do that.... Yet sources we spoke with indicated that Credit Acceptance sometimes does modify loan terms by effectively cancelling old loans and originating new ones with different terms as part of collections lawsuit settlement negotiations. To the extent they actually occur, such modifications naturally affect Forecasted Collections.
Paradoxically, CEO Brett Roberts has discounted the importance of Forecasted Collections variance in earnings calls:
"...The other thing to keep in mind is when you look at the forecasted collection percentages, you've got to ask yourself what's a material number. And I would argue that all the [variance] numbers on the page with the exception of maybe 2009 when we had a 750 basis point positive variance, they were all immaterial..."81
"...What constitutes a big number, what constitutes a small number. We estimate that a 100 basis point variance in the collection rate impacts our after tax return by about 40 basis points. So if you're looking at a 10 basis point change or a 20 basis point change, it's really not a significant change..."82
Credit Acceptance's exclusive use of percentages misleadingly suggests that these seemingly low error rates (and by extension, collections after 60 months) are immaterial. In fact, variance from the Forecasted Collection Rate matters a great deal. We were able to convert these Forecasted Collections percentages into dollars as of year-end 2016. According to our analysis, a 1% reduction in Credit Acceptance's total Forecasted Collections represents a cash loss of $161.45 million. At 3%, the cash loss increases to $484.34 million, and at a 7% variance the cash loss is approximately $1 billion, sufficient to wipe out dealer holdback in every Dealer Loan pool.
Magic: Always Beating Forecasted Collections by the Same Amount
The deep subprime auto lending market is an unpredictable business, so it is surprising that Credit Acceptance consistently reports to DBRS that for its various securitizations it has exceeded its Cumulative Forecasted Collection Rate for each ABS by numbers that tend to hover closely around 112% in some cases and 120% in others. The Cumulative Forecasted Collection Rate is calculated by adding together all actual monthly collections for the ABS up through the present month and comparing it to all monthly forecasts up through the present month, with the effect of smoothing out potential dips. Yet other disclosures demonstrate, on average, that Credit Acceptance's risk management procedures have a failure rate—the inverse of the Forecasted Collection percentage—that has increased from 20.8% to 35.0% from 2009 to 2017.83
The consistent numbers may stem from the fact that during the revolving period, Credit Acceptance forecasts that it will collect the same amount each month84 so long as the required ABS advance rate is maintained and the total Net Book Value of the collateral contributed is at least the initial amount. In addition, the company presumably wants to avoid the third rail of each ABS by a wide margin: a "Revolving Trigger Level" of 90%, and a "Revolving & Amrotization [sic] Trigger Level" of 65% ("amortization" is misspelled in hundreds of CAALT DBRS reports), both of which execute only after applying for three consecutive months. Early amortization would speed repayment of the bonds and bring cash flow to Credit Acceptance to a halt.
DBRS and other ratings agencies obtain figures from Monthly Servicer's Certificate reports85 provided by Credit Acceptance, which are neither audited nor verified. The figures from these reports are simply plugged into spreadsheets that show consistently exceeded forecasts. Some of the Monthly Servicer's Certificate figures can be found on monthly, publicly-available Performance Analytics Reports issued by DBRS.
Table 5: Select Credit Acceptance Corporation Collections Figures Reported by DBRS
These amazingly consistent figures—possibly reported as cumulative percentages to mask any underlying problems—appear to be the result of carefully-planned financial engineering designed to disguise the fact that Credit Acceptance has experienced, and continues to experience, loan losses far in excess of any disclosures. Sharp drops from the elevated baseline, such as in June 2017 for CAALT 2016-3, can potentially be explained by the creation of new ABS vehicles (CAALT 2017-2 started in June 2017), since an ABS's first month of operations requires two months worth of collections revenue.
Credit Acceptance's most significant and ongoing SEC disclosures regarding loan performance consist of Forecasted Collections broken down by origination year, going as far back as the year the company went public.86 These figures tell a different story, with the overall Forecasted Collection Rate decreasing since 2009.
The ABS Treadmill: Signs of Extremely Poor Loan Quality
In 2013, Credit Acceptance suddenly brought in a third ratings agency, Kroll, to rate CAALT 2013-1, a potential sign of underlying conflict. (See Table 2.. One industry expert we spoke with surmised that Credit Acceptance brings in new ratings agencies as a matter of course. Yet the next ABS, CAALT 2013-2, was not rated by Kroll.
By January 22, 2014, going strong with 6,394 active dealers, Credit Acceptance issued $300 million in senior notes and used the proceeds, along with other borrowings, to redeem its previous outstanding senior notes.87 Later, the company closed its CAALT 2014-1 Term ABS, rated as per usual by S&P and DBRS. Credit Acceptance then closed CAALT 2014-2. For this ABS, DBRS dropped out entirely. Instead, the company once again brought in Kroll for the second and (thus far) final time.
After CAALT 2014-2, DBRS continued to rate every subsequent Credit Acceptance securitization through 2017 with no substantial changes to its reporting, suggesting something unusual about the underlying assets in that ABS.
Table 6: Credit Acceptance 2014 ABS Comparison (Dollars in Millions)
Table 7: CAALT Class88 Sizes (Dollars in Millions)
As illustrated by Table 6, relative to the immediately prior CAALT ABS, the much higher percentage of defaulted Dealer Loans in CAALT 2014-2, as well as the fact that it contained nearly twice as many purchased loans as CAALT 2014-1 (but nearly the same number of dealer advances) are indicators of substantial changes in the underlying ABS collateral quality. To be clear, an "Already Defaulted RIC" refers to a loan that was already in default when it was pledged to the ABS, on day one.
The sudden change is perhaps best explained by the fact that over 30% of CAALT 2014-2's Purchase Program loans were originated prior to 2011 and had a weighted average original term of 51 months, or 4.25 years. On average, this means that those loans were originated just prior to the 2008 financial crisis, and may have performed especially poorly accordingly.
Meanwhile, the once-miniscule "C" class of each CAALT ABS began to grow starting in about 2013, shown in Table 7. Historically, the company did not even attempt to get the C class rated, resulting in "NR" (no rating) indications on reports. With CAALT 2015-1, the C class, now with over $30 million of very low-quality loans, finally received an "A" rating. By the end of 2017 with CAALT 2017-3, the C class was over $53 million while the A class remained about the same size as it had been since 2012. With CAALT 2018-1, the C class grew even further.
In 2015, for the first time, Credit Acceptance attempted a third securitization within one year. After struggling to close CAALT 2015-3, the company gave up and went to its underwriters for a massive secured loan, larger than any ABS issued yet, at a variable interest rate that is now the company's most costly source of financing. The attempt at CAALT 2015-3 was instead converted into CAALT 2016-1, previously discussed in footnote 56.
In late 2015, Credit Acceptance began to roll out "e-contracting,"89 potentially worrying lenders due to loan quality concerns. The following year, lenders instituted a three-month seasoning requirement—a requirement that loans be older than a certain period of time as a proxy for loan quality—on any ABS collateral. In addition, Credit Acceptance's long-time contact at Wachovia/Wells Fargo, Chad Kobos, left his position after handling every securitization issued by Credit Acceptance since 2003. Mr. Kobos did not respond to a request for comment.
By the end of 2016, the interest rates on Credit Acceptance's warehouse facilities were twice their 2014 rates, so the company chose to rely more heavily on fixed rate ABS issuances. In 2017, Credit Acceptance was able to issue three securitizations in the same year for the first time.
As discussed previously, Credit Acceptance's ability to forecast collections has largely been a function of its ability to mitigate losses through dealer holdback retention, the massive size of Purchased Loan pools, and the ten-year term attributed to every loan pool. Absent these factors, we believe the company would be unable to forecast collections with any degree of accuracy, and with over 30% of its portfolio directly exposed to additional losses, we expect the company to escalate its apparent earnings management and attempt to issue many more securitizations to stem the bleeding.
Dealer Retention Problems
In 2006, Credit Acceptance reported that the "estimated life of the dealer-partner relationship was 20 months."90 With the company in business for over 40 years and a limited number of interested used car dealers in the country, Credit Acceptance has already churned through a shocking number of potentially interested dealers. All other financial concerns aside, its dealer-partner attrition percentages alone would be cause for alarm at most businesses.
Furthermore, the company appears to involve recycling dealers who at some point previously abandoned the program. From 2003 to 2016, approximately 50 former dealers would return every year. In 2015, this figure leapt to over one hundred. (See Table 8.)
In Credit Acceptance's Q4 2015 earnings call, CEO Roberts admitted that many dealers will never see dealer holdback payments.
Brett Roberts: "...if we miss our collection forecast by $1 million, $800,000 of that goes to dealer holdback in the impact to us is only $200,000..."
Analyst: "Hey, guys. Sorry if I missed this. Have you ever talked about what percentage of dealers you are working with today are eligible for holdback payments?"
Brett Roberts: "It would be all the portfolio dealers."
Doug Busk: "Yes."
Brett Roberts: "Now –"
Analyst: (Multiple speakers) "where they have to—you have to compute at least 100 loans?"
Brett Roberts: "There is. They have to, on the portfolio program, they have to complete 100 loans before they're eligible for dealer holdback. You know the problem with answering your question is for those dealers sort of around 40, 50, 70 you can't say definitively that it won't be eligible for dealer hold back at some point."
In 2005, Credit Acceptance disclosed that, historically, only 50% to 60% of dealers become eligible for Dealer Holdback (by closing 100 loans).91 Furthermore, as Credit Acceptance has added more dealers to its business, each dealer has become less "productive" on average with lower per-dealer loan volumes, making it that much less likely that each dealer will close 100 loans, form a pool, and receive the elusive dealer holdback payments. Therefore, the more the company grows and saturates the market, the lower the probability that its dealers will have an ongoing incentive—dealer holdback—to stick with the program. A high-growth strategy could ultimately backfire if enough dealers tire of the program and quit for good.
Table 8: Dealer Churn
By June 30, 2008, only 130 of the 2,943 dealers who deferred their enrollment fee had closed 100 loans. By December 31, 2008, per-dealer volume was 37.2 loans per dealer, a little over half the per-dealer volume in 2005 that Credit Acceptance suggested would indicate 50% of dealers close a pool. At the time of the Q4 2015 earnings call, per-dealer volume had dropped even further to 31.5 loans per dealer—nowhere close to 100. Coupled with Credit Acceptance's increased reliance on Purchased Loans as compared to 2005, these results indicate that the number of dealers that have capped a pool, and that are generating sufficient volume to allow for cross-collateralization rights payments to occur, has plummeted.
We have demonstrated that Credit Acceptance is unusually strapped for cash; that its litigation—a slow, costly way to scrounge up cash from delinquent borrowers—is through the roof; that its baseline loan failure rate is 35%, but probably higher (since not all delinquent loans necessarily involve repossessions, which involve a two-month delay); that negative variance from Forecasted Collections may have already eliminated dealer holdback from 2015 loans, with 2014 next to bat; that the company has tried to distract from this issue, but that it remains quite significant; that Credit Acceptance ABSes involve both questionable numeric figures in DBRS reports and questionable behavior involving rating agency participation; and that the company has a serious dealer retention and incentivization problem. The sum total of all of these indicators is that the company is struggling to survive. With almost no cash on hand, it may already be insolvent, while facing unrelenting trends that threaten to expose the downside of its carefully staged business model.
The Struggle To Stay Afloat
Round-Tripping, with a Swing and a Sweep
The use of circular cash flow, or "round-tripping," to sell an asset to another company and later on repurchase it at approximately the same price is a hallmark of accounting fraud. The goal is often to increase market capitalization, but according to the authors of one article on the practice, it can be used for other purposes: "Companies using asset-based financing are particularly prone to fraud involving the inflation of working capital assets because this increases the company's ability to borrow cash."92
Credit Acceptance is cash poor because most of its cash is restricted or allocated to its ABS vehicles. Simultaneously, it appears that the company has begun using one lending facility to pay off another on a regular basis, in a loop of deteriorating loan quality and cash losses. We know the company engages in this practice because it admits as much: past disclosures include, "[W]e will need to...[a]ccess an additional $300-350 million in additional capital to allow us to refinance existing amortizing debt and fund new loan growth,"93 and "we used the net proceeds from the 2021 notes, together with borrowings under our revolving credit facilities, to redeem in full the $350.0 million outstanding principal amount of our 9.125% senior notes due 2017."94
Management's goal is to continuously meet ABS minimum threshold payments both for the sake of appearance—after all, bondholders would be extremely unhappy to learn that payments were not being made in full and word travels fast on Wall Street—and to avoid the first contractual trigger. For CAALT 2017-2, the early amortization trigger (which winds down the ABS and pays bondholders early) occurs when cumulative actual collections as a percentage of base forecasted collections are less than 90% for three consecutive months.
Payments to the ABS, which are constantly being wired to Wells Fargo, ordinarily come from borrower payments. An obvious problem arises when borrower revenues fall short, at which point the next-best option is to borrow money on a short-term basis to continue making the necessary payments. Since Credit Acceptance uses the same bank and numbered account at Comerica Bank for its revenue collections and ABS wires,95,96 and Comerica is also the deal agent and lead bank for the company's secured line of credit,97 it can easily borrow against that secured line of credit and wire funds to Wells Fargo for the ABS if necessary. Based on the amounts borrowed between 2015 and 2017, we believe that Credit Acceptance began utilizing a sub-facility within its secured line of credit called a "swing line" intended for extremely short-term borrowing.98 A swing line advance can be for one to thirty days and is offered as a routine commercial product by Comerica Bank. The swing line maximum limit on June 24, 2014 was $25 million, eventually raised to $30 million. An additional feature offered by Comerica called "business sweep to loan" that facilitates automatic loan repayments completes the picture.99 Swing line advances can be facilitated through a sweep agreement, which we believe may be in use at Credit Acceptance based on its low cash reserves.
Starting in Q1 2015, for seven of ten quarters, the average amount borrowed was roughly equal to the swing line credit limit. Prior to that period, the average amount borrowed always exceeded the swing line credit limit. See Table 12.
On the ABS side, Wells Fargo does not check the source of the aggregate funds that are wired to it for each ABS, and does not require Credit Acceptance to provide a detailed breakdown of the funding source on a loan-by-loan basis—it simply checks to make sure that the money is there. Based on our analysis, since approximately 2015, the process has worked as follows:
In 2012, Credit Acceptance's use of its warehouse facilities began to accelerate. In Q2, warehouse borrowing peaked at $430 million with average quarterly turnover—the amount borrowed in a quarter divided by average balance for the period—of 2.33. In Q4, borrowing amounted to $315 million, but with average quarterly turnover of 6.20. This rate of turnover was surpassed in Q3 2015, after which it accelerated throughout 2016 and 2017, with average quarterly turnover of 19.49 in Q1 2017. (See Appendix D, Tables 12-14.)
During the first quarter of 2015, Credit Acceptance borrowed over a billion dollars on its secured line of credit with a $310 million limit (which it recently felt the need to increase).101 The company was able to do this by borrowing and paying back an average of $100 million approximately nine times throughout the quarter. At times, it appears that the company has also borrowed from its VSC Re subsidiary.102
By Q2 2017, the company had borrowed nearly $2.4 billion from lenders (about $250 million borrowed and paid back every week and a half), but with an accretable yield balance of only $1.5 billion on June 30,103 raising the question of why so much borrowing was necessary. In recent quarters, the company also made sure to keep balances low by each quarter end before ramping them back up again, perhaps so as to make numbers more palatable for reporting purposes. This pattern of heavy borrowing is unsustainable and makes the company vulnerable to external shocks that could originate from credit markets or a newly volatile stock market.
While it is certainly true that non-bank lenders routinely borrow money from capital markets to finance their lending operations, Credit Acceptance's situation is markedly different because it is attempting to balance the books of a multi-billion dollar corporation on the back of a $30 million swing line. Its disclosed loan pool performance is poor and getting worse, yet its ABSes beat forecasts by about the same percentage month after month. To keep ABS holders satisfied in this delicate dance, we believe that the company is filling the gap with huge amounts of short-term borrowing.
Credit Acceptance is able to pull of this feat so long as the music doesn't stop, to use the parlance of Citi's Charles O. Prince from 2007 (whose quip is portrayed in the film Margin Call). Aside from a booming stock market that added over $1.2 billion to Credit Acceptance's market capitalization in December 2017 and January 2018 alone, the company has also benefitted from generous interest rates on its secured line of credit thanks to the syndicate led by Comerica Bank.104 This beneficial arrangement may exist because billionaire founder Don Foss is also a Comerica customer.
It should also be noted that Wells Fargo has played an instrumental role in enabling the Credit Acceptance business model. Even despite numerous public pronouncements dating back to at least as early as 2010 that Wells Fargo would soon be exiting or had already exited the subprime auto lending market, the bank has persisted in offering warehouse lending and ABS services to Credit Acceptance, allowing the company to use its ABSes to repay outstanding indebtedness and repurchase shares rather than finance new loans.
"Over-reliance on securitization as a funding source" is a lending red flag according to the Office of the Comptroller of the Currency, as is "significant growth or pressure for growth."105
Dealer Collections Purchases
In addition to a dizzying amount of short-term borrowing, Credit Acceptance has introduced confusing jargon regarding its collateral assets. Four terms in particular are crucial. Eligible Dealer Loans are Portfolio Program advances, secured by RICs, made to dealers. Eligible Dealer Contracts are underlying RICs assigned to Credit Acceptance Corporation to secure a pool of Eligible Dealer Loans. Eligible Purchased Loans are RICs that Credit Acceptance purchases outright from dealers under the Purchase Program. These are acquired immediately after origination by the dealer. Finally, Eligible Purchased Contracts are Eligible Dealer Contracts that once secured a Eligible Dealer Loan pool, purchased by Credit Acceptance long after origination.
In 2012, Credit Acceptance's institutional lenders began including a description of a practice called "Dealer Collections Purchases" (hereinafter, "DCPs") in securitization deal documents. DCPs are instances where Credit Acceptance "extinguishes," or erases, a pool of Eligible Dealer Loans by purchasing all of the pool's underlying RICs outright. As shown in Table 9, the practice was initially referenced in securitization deal documents (usually Section 3.2 or 3.3) to identify the resulting Eligible Purchased Contracts as assets belonging to the securitization trust and not Credit Acceptance Corporation. An example of the contractual verbiage appears as follows:
"...the related Dealer Loans...shall be deemed to be extinguished...and the loans thereunder shall be deemed Purchased Loans. For the avoidance of doubt, all Collections on such Purchased Loan Contracts shall be included in Available Funds."
Credit Acceptance has a program already in place for loan purchases: the Purchase Program. In its Q4 2015 earnings call, CEO Brett Roberts admitted that the Purchase Program is much riskier for the company than the Portfolio Program.106 In his words, "You have to be a little bit more careful about the dealer that you do business with. You have to be a little bit more conservative about your collection forecast. You have to have some different risk management procedures in place.... The company has also repeatedly used the boilerplate disclosure, "For Purchased Loans, the decline in forecasted collections is absorbed entirely by us. For Dealer Loans, the decline in the forecasted collections is substantially offset by a decline in forecasted payments of Dealer Holdback."107
On December 27, 2012, Credit Acceptance extended a $325 million revolving warehouse facility, Warehouse II, with Wells Fargo. The company announced that the interest rate on the facility had been reduced from the commercial paper rate plus 2.75% to the commercial paper rate plus 2.0% and that there were no other material changes. The Warehouse II documents were later changed to include DCP verbiage.
Table 9: Warehouse Lending Facility Dealer Collections Purchases Clauses
In 2013, Warehouse IV, with BMO, amended its facility documents to include the DCP reference and in 2014, Warehouse III, with Fifth Third Bank, was replaced with Warehouse V, which also included the DCP reference in its facility documents. The company's securitizations through the end of 2014 (CAALT 2013-1, 2013-2, 2014-1, and 2014-2) also included discussions of DCPs. By the end of 2014, all of Credit Acceptance's sources of asset-backed financing included a discussion of the DCP practice.
In 2015, Credit Acceptance's lenders began including an admonition that DCPs may not be used solely to extinguish Eligible Dealer Loan pools of poor credit quality, by adding the phrase "and not based on poor credit quality of the Dealer Loan Contracts" within parentheses in the deal documents. In 2016, the company's lenders amended the section heading to read "Dealer Collections Purchase; Replacement of Dealer Loan with Related Purchased Loans.. Otherwise, DCPs and their impact on Credit Acceptance's finances are not directly disclosed in any SEC filings except via opaque references in attachments of securitization agreements, nor is the practice discussed at all in ratings documents issued by DBRS or S&P.
Credit Acceptance has only disclosed that purchased loans are acquired in at least two ways: through actual purchases at origination and through forfeiture of dealer holdback (disclosed as "Transfers" on previous Credit Acceptance Corporation financial statements). DCPs are not the same as Transfers, which are listed as a component of Loans Receivable recently averaging about $1.1 million per quarter and defined by a Form 10-Q footnote that states, "Under our Portfolio Program, certain events may result in Dealers forfeiting their rights to Dealer Holdback. We transfer the Dealer's outstanding Dealer Loan balance to Purchased Loans in the period this forfeiture occurs.. In other words, if a used car dealer goes out of business or voluntarily elects to stop using Credit Acceptance's program, the company will sometimes buy the dealer's loans outright and move them to the Purchase Program. Or, if a dealer fails to complete a pool of 100 loans, "they would forfeit their right to the dealer holdback and those loans would be transferred to purchased loans," as explained by Senior Vice President and Treasurer Doug Busk on the company's Q1 2017 earnings call.108
In 2017, DBRS, which has rated nearly all of Credit Acceptance's securitizations since 2009, began publishing Performance Analytics Reports that contained sudden shifts. Given the lack of disclosure, these shifts could represent:
Actual DCP volume for each month is listed on each CAALT Monthly Servicer Certificate, but these are not routinely available to the public. The magnitude of the shifts reported by DBRS makes it highly unlikely that these shifts are explained by Transfers. In April 2017, DBRS suddenly reported CAALT 2016-2 Eligible Purchased Contracts of $216 million, approximately 200 times greater than the Transfers Credit Acceptance disclosed in the same time period.111. In September 2017, pledged collateral contained $275.2 million of Eligible Purchased Contracts, meaning that Credit Acceptance had by that point "extinguished" $275.2 million worth of advances on RICs that had previously been grouped by DLN in small, volatile pools.
Even though DCPs are clearly different than Transfers, Purchase Program loans acquired through Transfers are used to adjust collections forecasts, which suggests that Purchase Program loans acquired through DCPs can also be used to adjust collections forecasts. As the company has reiterated in multiple SEC filings, "The forecasted collection rates and advance rates presented for each Consumer Loan assignment year change over time due to the impact of transfers between Dealer and Purchased Loans."112
Since they are grouped by month and year, Purchase Program pools are very large and therefore far less likely to be affected by performance of individual loans. In contrast, Portfolio Program pools grouped by DLN are small and severely impacted by the performance of individual loans. Credit Acceptance has admitted as much.113
"...GAAP treatment was originally developed considering large pools of loans, such as those under our Purchased Loan portfolio that are aggregated by month of purchase. For our Dealer Loan portfolio...loans are aggregated into pools by individual dealer. As a result, we have numerous pools (in excess of 20,000 at December 31, 2016) that contain a relatively small number of loans. While our forecasts for loans assigned in any period have been stable and accurate in aggregate historically, our forecasts at the individual dealer level are highly volatile due to the small number of loans in each pool."
In 2015, the company booked a provision expense of $41.5 million due to such volatility.114. By booking RICs under the Purchase Program into very large, less volatile pools of loans grouped by month and year of vehicle purchase, Credit Acceptance could avoid booking provisions for losses against the company's net income. Effectively, the company could quietly use the Purchase Program as a "bad bank."
Once in the bad bank, the RICs, now rebranded "Eligible Purchased Contracts," are maintained as collateral in addition to any replacement Eligible Purchased Loans and Eligible Dealer Loans already in the monthly pools.115. Cash flow generated by Eligible Purchased Contracts, replacement Eligible Purchased Loans, replacement Eligible Dealer Loans, pre-existing Eligible Purchased Loans, and pre-existing Eligible Dealer Loans is made available to bond holders to the extent necessary to meet any dealer requirements. Excess cash flow is made available to Credit Acceptance Corporation.
Alternatively, DCPs could be advantageous to Credit Acceptance because the company can only securitize assets it actually owns. In the case of the Portfolio Program loans, this means that it can only securitize the portion involving the company's advance to the dealer; the full RIC is a contract between the dealer and the borrower, even though it is facilitated by Credit Acceptance. In contrast, Purchase Program loans are completely owned by Credit Acceptance, and can be fully securitized. By moving a loan from one program to the other, the company stands to gain the ability to report additional assets as collateral.
Whichever explanation may be correct, the maneuver achieves two short-term goals: stabilizing "actual collections" versus forecasted collections and increasing Net Income through a reduction in provisions for credit losses. However, this practice also places Credit Acceptance at risk of significant losses should these Eligible Purchased Contracts—purchased loans with no dealer holdback buffer—underperform.
There is some evidence that the first goal has indeed been achieved. In the March 2017 collection period for CAALT 2016-2, DBRS reported that Credit Acceptance had achieved exactly 100.00% of cumulative collections as a percentage of the company's forecast: the company had collected every single dollar of the $311,381,933 it forecasted well in advance, and not a penny less. In the deep subprime market, this is frankly impossible, even on the best of days. Furthermore, DBRS reports that cumulative collections as a percentage of the company's forecast for each ABS always have remarkably consistent figures that hover within fractions of a percent near 112% or 120% depending on the securitization, as previously discussed. This suggests that either Credit Acceptance could earn more profit across the entire financial sector with its "proprietary credit scoring system," or the reported figures are being manipulated.
Only one industry observer appears to have noticed similar activity, though its description may refer to Transfers. In a September 2017 report entitled "The Subprime Auto Bubble," a fund called 3-Sigma Value wrote:
"CACC moves the loans from the partner program to the purchase program, in part to maintain the credit performance in the partner program required for securitizations. Since 2013, transfers from the portfolio program to the purchased program have exceeded write-offs and the process has accelerated. In effect, the purchased portfolio serves as a 'bad bank' as CACC segregates non-performing loans from performing ones."116
Table 10: Comparison of DBRS Reports and Credit Acceptance Financial Statements Filed with the SEC
As of September 2017, Credit Acceptance reported $4.8 billion in loans receivable: $3.5 billion in Dealer Loans and $1.3 billion in Purchased Loans. Of the $4.8 billion in loans receivable, the company reported approximately $3.2 billion pledged as collateral to currently outstanding ABSes.117 See Table 10.
DBRS lists an additional $1.46 billion in collateral as being pledged to these securitizations. This additional collateral, undisclosed in Credit Acceptance's financial statements, is made up of Eligible Purchased Contracts. To the extent that there is an innocuous explanation for the Eligible Purchased Contracts figures appearing on both the company's own Monthly Servicer's Reports and DBRS Performance Analytics Reports, but not in its financial statements, it is not possible for an average investor to discern that explanation from the company's public disclosures, which itself a problem. In a decade of financial statements and reporting, Credit Acceptance has not once told investors that it engages in DCPs as described in its securitization deal documents. There does not appear to be any disclosure regarding Credit Acceptance's apparent purchase of $1.46 billion in loans from existing Portfolio Program pools, or regarding this additional $1.46 billion in collateral being pledged to existing securitizations. Though unlikely, this could conceivably mean that Credit Acceptance has fewer assets than it has been telling investors it has. With just $4.9 million in cash, a hole this large would be difficult to plug.118
Personal Wealth Extraction
On February 1, 2010, Credit Acceptance issued $250 million in senior notes, purportedly to reduce the company's heavy reliance on asset-backed financing, to "grow originations," and "reduce our reliance on the short-term credit and bank markets."119 The funds from these senior notes were used to finance the repurchase of 4 million shares of Credit Acceptance stock at a cost of $200 million. 3.7 million of the shares sold were beneficially owned by Don Foss, the company's founder and Chairman of the Board. The company stated that the share repurchase was financed by borrowings under its line of credit and warehouse facilities.120 Yet on page 37 of the same Form 10-K, Credit Acceptance also stated that the senior notes were used to pay off borrowings under their line of credit and warehouse facilities:
"...on February 1, 2010, we issued $250.0 million of Senior Notes. The net proceeds from these financings were used to repay outstanding indebtedness under our revolving credit facility and our $325.0 million secured warehouse facility."
The senior notes included several restrictions on repurchasing stock, the most significant of which was a limitation based on Net Income from Q4 2009 and the quarter preceding the repurchase.
"...The Company shall not...directly or indirectly... purchase, repurchase, redeem, defease or make any other acquisition or retirement for value of any Capital Stock of the Company held by any Person..."
Credit Acceptance's repurchase took place in Q3 2010. Net Income during Q4 2009, Q1 2010, and Q2 2010 was $18.5 million, $32.0 million, and $49.0 million, respectively, for a total of $99.6 million. The company's Q3 2010 share repurchase exceeded this restriction by nearly $100 million. Additionally, repurchase amounts were classified as restricted payments. The senior note documents count all restricted payments made since the senior notes were issued towards the Net Income calculation.
On March 3, 2011, Credit Acceptance issued an additional $100 million in senior notes. The proceeds, along with additional borrowings, were used to finance Credit Acceptance's repurchase of 1.9 million shares for $125.0 million. 1.8 million of those shares were beneficially owned by Don Foss.121
Credit Acceptance disclosed that the repurchase was made using proceeds from the March 2011 senior note issuance. However, it appears that Credit Acceptance violated its senior note restrictions in conducting this repurchase. The repurchase occurred in Q1 2011. Credit Acceptance's Net Income between Q4 2009 and Q4 2010 was $246.1 million, but Credit Acceptance had already repurchased $200 million in shares in 2010. Credit Acceptance's new share repurchase exceeded available Net Income by approximately $78 million.
In 2012, for the first time in five years, Credit Acceptance issued two term ABSes, CAALT 2012-1 and CAALT 2012-2, for a combined total of $453.3 million. The company used this financing to repurchase 1.74 million shares of common stock worth $152.5 million from unknown sellers.
This set of repurchases, which also appears to violate senior note restrictions, also represented a shift in financing for Credit Acceptance. The 2010 and 2011 share repurchases were completed wholly through a tender offer and the shares were ultimately purchased from Don Foss. The 2012 repurchases were split: 1 million shares purchased pursuant to a tender offer, including 557,023 from board members or their affiliates; and 727,000 shares purchased on the open market from unknown parties. The following year, Credit Acceptance also issued two Term ABSes, CAALT 2013-1 and CAALT 2013-2, for a combined total of $338.1 million.
Credit Acceptance executives were well aware that they were not allowed to purchase the company's stock with the money they had borrowed. Instead of purchasing the shares outright, they paid for them using their secured line of credit and their warehouse lending vehicles, which actually executed the trades—and then repaid their lenders the following day for the transactions that they had been forbidden to execute themselves. In this manner they attempted to comply with the letter of the agreements, while clearly violating their spirit.
In April of 2013, Credit Acceptance filed registration documents with the SEC for listed shareholders selling 1.5 million shares, with an option for the underwriters to purchase another 225,000 shares from those shareholders. Credit Acceptance repurchased 1.2 million shares for $135 million dollars. This time, substantially all of the repurchases were made on the open market.
Why No One Noticed
Throughout the course of investigating Credit Acceptance, we spoke with individuals involved in several aspects of their business. Some of the most surprising statements came from employees and former employees of ratings agencies.
Ratings agencies were famously highlighted in the aftermath of the 2008 financial crisis as having played a large role in allowing mortgage-backed securities to effectively implode the economy. In an infamous scene in The Big Short, an S&P ratings analyst appears with literal blinders on.122 S&P Global Ratings ultimately paid a $1.375 billion penalty for defrauding investors in the lead-up to the financial crisis,123 while Moody's paid a penalty of $864 million for its role.124 Here again, we found that the fundamental model around ABS ratings, where corporate customers pay for analysis to support a misleadingly narrow rating they request up-front, is fundamentally broken and not at all changed from a decade ago.
We spoke with a ratings analyst familiar with the Credit Acceptance model. This individual expressed admiration for Credit Acceptance, saying that "it's a company and a business model we've gotten very comfortable with," and telling us, "I think their track record does speak volumes for them, particularly in that segment of the auto space; to be around that long and be that successful, you're clearly doing a lot of things right.. The analyst described some of the challenges of rating the company's ABSes:
"It's not really like any of the other, what you might call 'typical' auto loan ABS securitizations. In the course of doing it we twist ourselves into a pretzel, but we're trying to look at it through our typical auto ABS lens. So, in other words, we're trying to stuff it into our auto loan criteria box, but at the same time recognizing that, well, there are some funky things going on, so we need to add a few other bells and whistles to it..."
The analyst also emphasized that the ratings agency felt comfortable with already defaulted loans being pledged to securitizations because of cross-collateralization, despite the fact that we were told, "nobody [else] ever does that.. The analyst also told us, much to our surprise, that the models used to rate the company's ABSes only focus on performing contracts, and that the model ignores defaulted loans for cash flow purposes (but not necessarily as related to other metrics).
When asked about the steadily increasing number of outstanding ABSes the company relies on for long-term financing, the analyst stated, "It doesn't seem to be anything they can't handle.. Regarding Dealer Collections Purchases, the analyst was familiar with the practice, but downplayed its significance, stating, "I don't think that's a big percentage of anything that's ever happened."
Many have attributed Credit Acceptance's success in effect to its uniqueness, pointing out that over the years other companies have tried and failed to replicate the model. These supporters seem unaware that those failures might be attributable to the fact that the model does not work. To the contrary, ratings agencies appear to view their analysis as leaning conservative because it focuses on the bonds, and not the company itself. As the analyst we spoke with put it, "The bonds are a subset—eighty percent—of the advances, which are themselves a subset of the forecasted collections. You could have reasonably catastrophic losses, and still pay off the bonds. So I think there's a lot of conservatism built into the structure, as there should be.. In contrast, when we spoke with others from rating agencies with knowledge of Credit Acceptance, we were told that we had indeed identified several errors in the company's securitization reports.125
The blame does not fall on the major ratings agencies alone. In 2015, the SEC charged ratings agency DBRS with misrepresenting its ratings methodology for residential mortgage-backed securities, among other "certain complex financial instruments.. According to the SEC press release,126 "Without admitting or denying the findings in the SEC's order, DBRS agreed to disgorgement of $2.742 million in rating surveillance fees it collected from 2009 to 2011 plus prejudgment interest of $147,482 and a penalty of $2.925 million. DBRS also agreed to be censured and retain an independent consultant to assess and improve its internal controls among other things."
The Wall Street analysts who prepare reports for institutional clients also appear to have missed the forest for the trees. We examined over 300 analyst reports issued on CACC common stock going back to 2012, and found them to be formulaic, uninformed, overly optimistic, and lacking for meaningful insight. Many of the reports were automatically generated by number-crunching algorithms that simply reported various metrics, while others drafted by actual analysts sliced and diced the company's SEC disclosures in a few predictable ways, without offering anything new and while overlooking crucial details. Many of the reports parroted what little Credit Acceptance discussed on its notorious short and uninformative earnings calls, such as the "competitive environment.. To their credit, several of these analysts have recommended that investors sell CACC in recent months.
Auditors fare slightly better in this story—more than once, the company's auditors departed under a cloud, either because they resigned or were fired. We would welcome feedback from Grant Thornton, especially on its conflict of interest policy.
Regulators have been asleep at the wheel for years, and now under the Trump administration, they are being dismantled—especially the CFPB, which previously had jurisdiction over Credit Acceptance. After the financial crisis, the SEC was tasked with implementing section 942 of the Dodd-Frank Wall Street Reform and Consumer Protection Act, which concerned ABS transparency. The SEC implemented this portion of Dodd-Frank by issuing a final rule in 2011 suspending reporting and transparency obligations for certain (read "many") securitization issuers. A handful of comment letters from the American Securitization Forum, Chris Barnard, Cleary, Gottlieb, Steen, & Hamilton LLP, the CRE Finance Council, the Investment Company Institute, MetLife, Inc., and the Mortgage Bankers Association were all that was needed to turn "public interest" legislation against the public once more.127
The issue of the need for transparent asset-level data arose again in 2014, when the SEC issued a Final Action.128 Despite all of the SEC's work, the Credit Acceptance case makes clear that barely any ABS data, let alone asset-level data, is transparent to the public at all. Public companies continue to profit from private securitizations, at great risk to investors.
While various Attorneys General have looked into some of Credit Acceptance's business practices, its complex finances have remained largely unexamined. Almost every investigation has ended with a slap on the wrist as more lives have been "changed" for worse by the company.
The accounting practices used by Credit Acceptance Corporation are "not especially well understood"129 and may be signs of insolvency, ongoing accounting fraud, or both. As recently as January 30, 2018, Brett Roberts answered a basic analyst question about a $27 million adjustment with, "I don't know, you stumped us on that one."130 To the extent that there are reasonable explanations for apparent circular cash flows, rapid borrowing, Dealer Collections Purchases, share repurchases, and share dumping by insiders, the company has completely failed to disclose them to shareholders, and to date has still refused to answer any of our questions posed in Part I, Appendix C, or our single request on January 29, 2018 to explain its DCP discrepancies.
With more transparent reporting, it would be far easier to discern the company's financial health, which is potentially why the company refuses to provide it. We would welcome feedback from the company on our estimates.
Instead, like many companies, Credit Acceptance only provides investors with the positive outlook. In its 2013 annual shareholder letter,131 the company states a number of "Key Success Factors" that purport to summarize "the key elements of our success today."
The company claims, "We have developed the ability to offer guaranteed credit approval while maintaining an appropriate return on capital.. Yet Credit Acceptance refuses to disclose key metrics about its return on capital, because if it did, we believe those metrics would reveal a company in crisis.
The company claims, "We understand the daily execution required to successfully service a portfolio of automobile loans to customers in our target market.. The company has filed hundreds of thousands of lawsuits to keep up with its cash flow needs.
The company claims, "We have learned how to develop relationships with dealers that are profitable.. Dealer churn is significant, suggesting that the program is not working for many dealers at all.
The company claims, "We have developed a much more complete program for helping dealers serve this segment of the market.. This may be a reference to CAPS, the company's patent-ineligible database that allows dealers to stealthily increase vehicle prices—a practice that recently resulted in a settlement with New York City.
The company claims, "We have strengthened our focus on our core business.. In fact, the focus of Credit Acceptance's business has shifted from its Portfolio Program to its Purchase Program in large part thanks to Dealers Collections Purchases, a practice invisible to investors and only discussed in hundreds of pages of securitization documents that has left a $1.46 billion question mark on the company's balance sheet.
The company claims that, "We have developed the ability to execute our loan origination process consistently over time.. In fact, it nearly went out of business twice in the mid-1990s and after the 2008 financial crisis, has been delisted from the NASDAQ once, and has had major disagreements with nearly all of its auditors.
Finally, the claim that, "We devote a large portion of our time to something we call organizational health" is simply untrue. We believe Credit Acceptance is deeply unhealthy. Perhaps once again it can stage a miraculous recovery, but unless management can convincingly answer questions about its accounting and business practices, its days may be numbered.
In Part I of this report, we followed the case of Samantha Rajapakse, a car buyer who eventually sued Credit Acceptance twice in federal court. Ms. Rajapakse provided us with a copy of a December 20, 2017 letter that Sandy Pollack, a Regulatory Compliance Manager at Credit Acceptance, sent to the Attorney General of Michigan, accusing her of vexatious litigation and "harassing behavior" by filing legal complaints and creating YouTube videos concerning her cases that encouraged others to do the same. The apparent intent of the company's letter was to block further attempts by Ms. Rajapakse to exercise her legal rights without the assistance of counsel.
The Attorney General replied to Ms. Rajapakse on January 2, 2018 by informing her that "there is no action we can take on your behalf" since her case had already gone to court and that "we regret that we are unable to further assist you in the resolution of this matter.. That letter did not address whether the Attorney General of Michigan could take broader action, as many other Attorneys General have.
Credit Acceptance followed up with Ms. Rajapakse on January 19, 2018 with a written notice demanding $5,649.63 on her past due balance. On February 3, 2018, the company repossessed her truck. Her second lawsuit is ongoing.132
Also in Part I, we reported that Credit Acceptance Corporation comprised 12.01% of the Michigan 36th District Court's entire docket. Due to an error in our analysis software that repeatedly counted unexpected mentions of the phrase "CREDIT ACCEPTANCE" occurring in the same file beyond the parties to each lawsuit, we over-reported this figure by 0.31%. The correct figure is 11.70%. Although this has change no material impact on our conclusions, we regret the error.
Appendix D: Data Tables
Table 11: FinViz.com Credit Services Industry Company Comparison (1/27/2018)
Table 12: Secured Line of Credit Turnover (Millions of Dollars)
Table 13: Warehouse Lending Facilities Turnover (Millions of Dollars)
Table 14: Aggregate Turnover (Millions of Dollars)
Table 15: Underperforming Dealer Loan Pools (Dollars in Millions)
2 Loan and write-off figures from 1992 to 2005 refer to the entire principal and interest amount. After 2006, the company began reporting loan balances as Net Book Value of the associated dealer advance (in most cases less than 50% of the principal and interest balance of the underlying loan).
9 The senior note issuance was managed by William Blair & Co. In August of 1997, Thomas Fitzsimmons, the William Blair & Co. executive who took Credit Acceptance public in 1992, resigned from his position at William Blair. He publicly stated that his resignation had nothing to do with any trouble at Credit Acceptance that may have impacted William Blair's beneficial holdings of over 4 million shares. "Speaking from San Diego, where he is building a home, the 53-year-old denies any connection to the stock slide and his exit as a Blair principal. He plans to start a merchant bank..."
11 Credit Acceptance also adjusted its revenue recognition policy by lowering its "non-accrual" threshold from 120 days of non-payment to 90 days, and the charge off requirement from 12 months to 9 months.
17 In its Q3 1997 SEC Form 10-Q, Credit Acceptance claimed that its dealer network had dropped by over 200% as a result of an "internal review."
19 After Congress passed federal tax reform in 1986, the IRS began to enforce a new (and widely disliked) provision that required used car dealers to record all of the potential revenue from a loan up-front. This change had a significant impact on Credit Acceptance dealers. Credit Acceptance also sued the United States Department of the Treasury over the tax treatment of repossession expenses. See 236 Mich App 478 (1999). Such antics resulted in years of audits. In 2007, an IRS Technical Advice Memorandum (TAM) unfavorable to dealers again inflamed tensions with the IRS.
27 To highlight how much these PCAOB audits strained relations between Deloitte and its clients, two years before the inspections began in H1 2001, Deloitte lost 34 publicly traded clients through resignation or dismissal. In H1 2002, it lost 35. In H1 2003, the PCAOB's inaugural inspection year, Deloitte lost 74. The trend continued through H1 2004 with 62 client losses, and H1 2005 with 64 client losses.
30 One of the "assumptions and estimates" in question referred to Credit Acceptance pooling loans together and then using the entire pool's weighted average loan term to calculate its revenues and losses. Adjusting the calculation to use actual loan terms had a material, negative impact on the company's net income.
32 Thus, by 2004, Credit Acceptance had already been stopped by its auditors and the SEC from implementing three types of creative accounting: disregarding actual delinquencies in favor of across-the-board cash flow predictions extended out for up to ten years; ignoring dealer holdback payments when forecasting future collections and loan spread; and disregarding individual loan terms and adjusting forecasted cash flows based on actively changing pools of heterogeneous loans. Credit Acceptance would continue pursuing these three objectives despite the temporary regulatory setback, going so far as to make its case directly to investors through its website.
33 Credit Acceptance argued that it was an originator of loans from consumers because it performs substantially all parts of the loan origination process. A distillation of this argument can be found in the company's 2006 SEC Form 10-K, in a paragraph discussing how CAPS interfaces with other corporate systems. "Consumer Loan contracts are written on a contract form provided by the Company...the Company, not the dealer-partner, is listed as lien holder on the vehicle title. The customer's payment obligation is directly to the Company..." Deloitte argued that a repurchase clause within dealer servicing agreements gives the dealer an interest in the loan and that this interest shifts Credit Acceptance's transaction away from origination at the consumer level to secured lending and third-party servicing at the dealer level.
34 Instead, the company made its shares available on the OTCBB, which did not require audited financial statements.
35 In the first half of 2005, Grant Thornton picked up 47 former Big Four clients, 12 of which came from Deloitte.
37 Mr. Vassalluzzo remains on the Credit Acceptance Board of Directors and in his individual capacity is the largest direct holder of CACC stock aside from founder Don Foss and CEO Brett Roberts. His investment firm, Prescott General Partners LLC, is the top institutional holder of CACC stock, with over 2 million shares. He joined Thomas N. Tryforos, another current Board member with Prescott ties. There are only four Directors total; one is the CEO.
38 Simultaneously, Credit Acceptance formed the captive VSC Re Company to re-insure VSCs peddled by its dealers against claims by car buyers. Credit Acceptance used its formation as an excuse to use more flexible accrual accounting.
42 Credit Acceptance is a specialty finance company with its headquarters located at 25505 West Twelve Mile Road in Southfield, Michigan.
46 Credit Acceptance charges a 20% servicing fee. However, Credit Acceptance also retains 100% of collections until its loan to the dealer is repaid. After the loan is repaid, Credit Acceptance retains only the 20% servicing fee unless it exercises its cross-collateralization rights.
47 In a broad sense, cross-collateralization refers to the ability to offset losses from one loan with gains from another. Cross-collateralization between separate pools is eliminated when pools are securitized in different ABSes or warehouse lending facilities.
48 Portfolio Profit represents a dealer's 80% share of collections after Credit Acceptance's loan has been repaid.
49 Portfolio Profit Express payments are up-front payments of approximately 25% of future Portfolio Profit payments. Dealers may elect to defer their initial enrollment fees in exchange for 50% of their initial Portfolio Profit Express payment, reducing the up-front payment to approximately 12.5% of future Portfolio Profit payments. During the first year this option became available, 1,277 dealers elected to defer their initial enrollment fees. Of those 1,277 dealers, only 13 became eligible for Portfolio Profit payments by the end of the year.
50 Within each dealer pool, performing RICs securing Dealer Loans are identified as "R Lot Receivables.. Delinquent RICs securing dealer loans are identified as "C Lot Receivables:" either RICs with balances that are 90 days overdue or RICs where the vehicle has been repossessed, sold at auction, and the proceeds received by Credit Acceptance. If a dealer has multiple Dealer Loan pools, and if any of those Dealer Loan pools has a Dealer Loan balance that exceeds that Dealer Loan pool's R Lot Receivables balance, the company can use collections from any or all the dealer's other Dealer Loan pools to pay down the deficient Dealer Loan balance until it is lower than its associated R Lot Receivables balance.
52 This program was initially piloted as a "No Dealer Holdback" program in 2005.
53 Credit Acceptance also has two senior note issuances outstanding, but these have been used to finance share repurchases rather than their lending operations.
54 In practice, the reality is more complex. These advances are often paid by more borrowing in the form of long-term securitization.
56 CAALT 2008-1 was a single tranche issuance. On February 26, 2016, Credit Acceptance announced that it had entered into a deal for CAALT 2016-1, "a $385.0 million asset-backed non-recourse secured financing" with Wells Fargo and Bank of Montreal (BMO). This security featured a highly unusual structure: a hybrid with all the costs of a warehouse facility, but with all the downside of a Term ABS.
57 Prior to 2008, Credit Acceptance's securitizations and trusts were identified as "Credit Acceptance Auto Dealer Loan Trust" or "CAADLT.. Wells Fargo, the indenture trustee, backup servicer, and initial purchaser, has underwritten these deals since 2003 and continues to list servicing documents under "Credit Acceptance Auto Dealer Loan Trust" on their trust document portal.
58 BMO Capital Markets/GKST, Inc. specializes in municipal bonds. Possibly due to company politics, BMO participated in CAALT transactions using this entity because it was Chicago-based, while BMO Capital Markets Corp. is in New York.
59 Many credit reports from Moody's are not public and can cost anywhere from $200 to $750 per report. Recent S&P RatingsDirect reports are publicly available, but accessing archived reports, to the extent that they are even available, can require a subscription costing tens of thousands of dollars per year. Shockingly, Kroll told us that it deletes its old reports once an ABS has begun its amortization period. We were therefore unable to obtain certain Kroll pre-sale reports.
62 The next-highest values are Manhattan Bridge Capital, Inc. (LOAN) at 489.48 and Regional Management Corporation (RM) at 64.52, leaving Credit Acceptance in a league of its own.
69 This does not necessarily indicate that 35% of vehicles are repossessed and sold. Credit Acceptance allows third-party contractors, who work on a contingency basis, to attempt collections on the past-due balance. If the collections are successful, the vehicle may be returned to the borrower. Otherwise, the vehicle is sold at auction.
72 S&P Global Ratings CAALT 2017-3 Presale Report, October 12, 2017, Chart 5. See also S&P Global Ratings CAALT 2018-1 Presale Report, February 7, 2018, Chart 5.
73 Between 50 to 60% of loans are sold with a Vehicle Service Contract which, by virtue of being a VSC, is uncollateralized. Additionally, VSCs are entirely refundable when a vehicle is repossessed, causing an instant loss of any remaining premiums.
74 Overall, the Payback Period tripled between 1992 and 2017, with an island of stability from 1996 through 2006, due to decreasing quality of the dealer base.
78 See Appendix D, Table 15, based upon the "Loan Pool Performance Less than Initial Estimates" tables in Credit Acceptance SEC filings.
84 This policy appears to have begun in 2011, prior to which each month's forecast was calculated separately.
85 One such Certificate for CAALT 2017-2 matched the numbers DBRS provided for that same period. However, while the Certificate cited the 65% trigger level on page 6, similar descriptions on page 5, lines 137-139, read "75%." DBRS appears to use the 65% figure in its Performance Analytics Reports.
86 This data can typically be found in a table under the "Consumer Loan Performance" or "Consumer Loan Metrics" heading in the company's SEC Form 10-Ks, around 20-25 pages in.
89 E-contracts, discussed briefly in Part I, involve the use of computer-based borrower signatures on loan documents. Some borrowers have complained that they never actually signed documents with their "electronic signatures" on them. E-contracts were rolled out as an opt-out enhancement to CAPS.
97 Collateral is not always specifically identified for the secured line of credit. The company is effectively borrowing against the totality of its assets that are not delineated as collateral for warehouse lending facilities and securitizations. Such assets include the company's (limited) intellectual property, such as trademarks, which as of 2004 began to be assigned to Comerica Bank. See the Trademark Assignments tab.
100 Advances can remain unfunded for up to 45 days before Credit Acceptance's funding guidelines require any action.
102 We analyzed restricted cash and restricted securities line items in SEC quarterly reports filed by Credit Acceptance to determine the quantity of restricted cash associated with VSC Re specifically. (VSC Re is the only disclosed source of restricted securities." From this analysis, it was apparent that VSC Re's assets would drop when the company appeared to be having cash flow problems. See also Report on Examination VSC Re Company, December 31, 2013.
109 We inquired with S&P Global Ratings about DCPs. S&P confirmed that the Eligible Purchased Contracts numbers reported by DBRS also appeared in the Monthly Servicer's Certificates provided to S&P by Credit Acceptance, making a typographical error by DBRS less likely.
110 We contacted DBRS to discuss these Performance Analytics Reports. DBRS was unable to shed any light on them.
111 Nor are DCPs the same as the rarely discussed Collection Only Loans, which are placed into one pool that is never capped. The aforementioned securitization agreements define DCPs with the statement, "the Dealer agrees to sell...one or more Pools of Dealer Loan Contracts" (emphasis added). Collection Only Loans are not pooled.
113 In 2011, Credit Acceptance also began reporting the dollar value of Dealer Loan pools and Purchased Loan pools with cash flows below forecasts. (Credit Acceptance has not disclosed whether this means that cross-collateralized payments have been taken and have not been sufficient to offset the losses." For Dealer Loan pools, the percentage of underperforming pools has never dropped below 60% between 2011 and Q3 2017. See Appendix D, Table 15. For Purchased Loan pools, the percentage has varied wildly, from 22% in 2011 to 4% in 2014. In 2016, it was 29% and in the nine months ended September 30, 2017, it jumped up to 38%. It is unclear why the Portfolio Program, with its smaller, volatile pools, appears to perform so consistently, while the Purchase Program with its larger pools is in reality so volatile.
115 New collateral must be acquired as Eligible Purchased Contracts and by the terms of the securitization can only be used towards the overcollateralization balance and not towards minimum collateral balance.
117 According to DBRS, that $3.2 billion is comprised of $1.8 billion in Dealer Loans, $0.7 billion in Purchased Loans and $0.8 billion in Unidentified Collateral. DBRS did not produce Performance Analytics Reports for CAALT 2014-2 and CAALT 2016-1.
118 In its Q4 2017 earnings conference call on January 30, 2018, the company addressed the issue of Transfers:
David Scharf, JMP Securities: "Can you expand a little more explaining the language in the press release regarding the reclassification of some dealer loans into purchased loans. Number one, is this something that you've historically done in the past but maybe not to as great a degree. And for our purposes, are there any implications about how the yields, going forward, on those loans are going to be treated?"
Brett Roberts: "I don't think I would worry about that too much. The answer is we do reclassify loans from dealer loans to purchased loans if the dealers don't meet their performance obligations; we're entitled to do that under the terms of the agreement. It's typically small dollars. What happened this quarter is we had a catch-up. We hadn't been reclassifying those loans in a timely basis, and so we caught that up this quarter. The reason it's in the release is it impacts some of the numbers in the collection and spread tables, so we changed we way we do those tables. From this point forward, the tables will reflect the way the loan was initially recorded, and that will just allow us to keep those numbers consistent over time. So we highlighted it in case you had a record somewhere of what was in prior releases so you wouldn't be wondering why they didn't tie out."
123 United States Department of Justice Press Release, February 3, 2015, "Justice Department and State Partners Secure $1.375 Billion Settlement with S&P for Defrauding Investors in the Lead Up to the Financial Crisis."
125 One agency, clearly worried about reputational risk, would not comment without communications personnel present, who never materialized.
References on PlainSite
By Aaron Greenspan and Christine Richard
Web content coming soon. Please download the PDF using the link above.
Have a tip?
E-Mail us at tips at plainsite.org