Credit Acceptance Corporation is an auto subprime lending company. I have followed this company for a while, and I initially disliked it. However, I’ve realized it has been a misunderstood business that has performed well and will likely continue to do so.
On August 31, 2020, the Massachusetts AG announced a lawsuit seeking $120 million in damages. The company’s stock price declined 20%, erasing roughly $1.5 billion in market capitalization. The shares have continued to sell off with general market weakness since then. I’ve entered into position on this sell-off because I believe it provides a good entry point to own a great business.
The following write-up will describe how CAAC makes money, profitability levers, management, the auto subprime market, and outstanding legal issues.
CACC makes money through two time-tested, straightforward offerings
CACC offers car dealers two programs: 1) the portfolio program and 2) the purchase program.
Under the portfolio program, the dealer keeps the customer’s down payment and receives a cash advance from CACC. CACC collects payments from the consumer until it recovers its full advance. Afterwards, CACC will forward remaining payments to the dealer net of a servicing fee.
The flagship portfolio program creates an incentive structure that improves loan performance. Since the dealer bears risk for the non-advanced part of the loan, he has an incentive to participate in the customer’s success. This includes selling quality vehicles and taking steps to ensure the customer understands and is able to pay off the loan.
From a risk management standpoint, the portfolio program insulates CACC. If a customer defaults on his first payment, CACC loses the cash advance net of repossession and collections recovery. Since the cash advance is less than half of the loan amount, the total loss CACC would suffer in this case would be small. If the consumer fully repays the loan, the dealer enjoys the corresponding upside.
Under the purchase program, CACC pays the dealer to receive full assignment. Typically, CACC pays a larger amount to acquire the loan outright than it does through dealer advances in the portfolio program. The higher initial outlay and full responsibility for the loan make this program risker for CACC.
Since the portfolio program has a 50-loan minimum, small dealers tend to use the purchase program. CACC also offers the purchase program to generate business with larger dealerships who do not wish to commit to the portfolio program. This program has become a greater part of CACC’s loan book as competition has increased.
CACC excels in forecasting accuracy and capital discipline, making it a great underwriter
Underwriting loans is a commodity business, with price competition eroding economic profit. There is no barrier to entry; execution is the only thing that leads to success.
The two keys to successful underwriting are: forecasting accuracy and capital discipline. CACC’s history shows that it is the best executor in this space under these criteria.
Each year, the company projects how much it will collect and provides a comparison against actual results.
CACC has demonstrated consistent ability to accurately forecast loan performance. Since 2003, the worst spread between actual and projected collections was -2.6%. Over the same period, the company generated a cumulative 60 bps in positive collections over initial projections.
The company estimates that a 100 bps negative spread in collections adversely affects its total return on capital by 60 bps. Since the company’s spread over cost of capital has averaged above 6% in the last 5 years, it would need see a negative variance of 10% before it books an economic loss (the negative variance would technically result in accounting losses a few hundred bps before -10% because of administrative expenses). To put that in perspective, the company’s worst year was in 2001 where it faced a -3.1% variance in collections. The current variance in forecast for 1H 2020 is just -0.4%.
Management is always comparing collections against forecasts and adjusting its model. Results for 18 years show they are good at it:
In this industry, lenders make worse loans as competition increases. Once the market hardens, usually as a result of an economic downturn, the ability to write new loans deceases since outstanding loans underperform. Competitors exit the space or dial back their operations. Those with strong balance sheets and market-beating performance continue writing new loans at substantially more attractive rates because of a less competitive environment.
When competition increases, the company does not aggressively expand by undercutting price and taking on increased risk. It does not unsustainably increase its leverage either. Instead, CACC adjusts its dealer advance rate while maintaining an adequate spread between the advance rate and the forecasted collection. In competitive environments, maintaining this stable spread generates consistent operating results at the cost of volume growth. Management has repeatedly stated that it will not grow at all costs. When management cannot generate enough volume to deploy all its excess capital at the desired spread, it buys back shares.
CACC’s performance through the GFC demonstrates the above operating philosophy in practice. Because CACC did not overcommit during the pre-GFC boom, it did not suffer outsized losses when the downturn hit. As a result, it was able to keep writing loans to generate major underwriting profits and gain market share in the proceeding years, even as the economy stalled.
The subprime auto market is large, fragmented, and provides CACC a long runway for growth
There isn’t a clear source stating the size of the subprime used car market. Data on sizing gets even scarcer as you move further down the credit ladder. As of 2020, I estimate that CACC has about 9% market share based on management comments in Q4 2012. This number matches the one I calculated using a top-down analysis involving loan volume, total used cars sold, typical rate of financing, and auto subprime originations as a percentage of total car loans. Under this framework, Santander Consumer has 27% market share and Ally Financial 18%. Regional financing companies and captive financing arms of dealerships tend to make up the remaining 46% fragmented market.
CACC’s market share depends largely on its dealer partners. It currently has 13,000. There are an estimated 50,000 viable car dealers across the country. Assuming the company’s TAM is 25,000 dealers, the company has a long runway to grow earnings both through onboarding new dealers and taking loan volume share from fragmented players.
The subprime auto space does not pose significant structural risks and, even if it does, CACC is insulated because it is the most conservative underwriter
There are major misconceptions with how this space works. People see interest rates, collection rates, increased subprime lending, etc. and immediately scream “bubble.” They also say that’s it’s predatory, with too high rates, unsavory collections practices, etc.
Those two ideas are mutually exclusive.
A “bubble” forms when large amounts of loans are underwritten at significantly below-market clearing prices. Saying the subprime dealers are predatory with high rates and aggressive collections practices would mean that they are pricing significantly above the market’s hurdle rate. This implies there is no bubble to burst.
Regardless, a downturn in the subprime market hurts the worst players. These are the ones with lax underwriting standards, high leverage, and a low margin of safety. As shown above, CACC sacrifices growth to maintain a significant spread in every loan it accepts. It also has the least leverage. It is therefore the most insulated player.
To the extent that “predatory” practices are further regulated, the need to accurately forecast and properly comply increases. Therefore, further regulation makes execution an even more important requisite for success. As discussed above and next, CACC’s management has the best execution in this space.
Management’s historical execution and compensation structure show it to be well-aligned and competent
Ownership & Incentives
First, insiders own 25% of this business. The founder owns no less than $700m worth of shares. The CEO owns $125m worth of shares.
Second, the company pays Brett Roberts, the CEO, $1m in cash compensation each year. RSUs vesting over a 15-year period and tied to improvement in economic profit are the rest of his compensation. Currently, Roberts has $133 million in total RSUs, $41 million of which are subject to forfeiture. The remaining amount is to be paid in equal installments over many years, starting in 2022. Even after adjusting for historic stock sales, >90% of Roberts’s family wealth is tied to the per-share value of CACC common stock many years down the line.
As a positive kicker, the CFO, Kenneth Booth, has purchased $1.5m worth of shares over the last year.
The entire incentive structure forces management to take a long-term view of the business.
Operational Competence & History
The Chairman, Donald Foss, essentially created the subprime auto lending space decades ago. Before going public, his accounting system was based on intuition. The company underwrote loans through its portfolio program and found that it generally earned more money than it spent. When the company went public, Foss learned that his accounting system was unequipped to meet disclosure requirements and provide a proper understanding of the business to both himself and outsiders. The old system broke down in the late 1990s, resulting in economic losses.
Brett Roberts joined CACC in the accounting department while in his 20s. After taking over as CEO in 2002 at the age of 31, Roberts formalized the financial systems and forecasting methods within the company. He has delivered 3,000% TSR over his tenure (nearly 7,000% before the lawsuit sell off a few weeks ago).
Management’s non-GAAP financials accurately reflect the reality of the business
The company’s financials are relatively complicated. Management provides numerous non-GAAP metrics and does not account for loans like a direct lender. At best this creates confusion, at worst this creates pessimism about the statements’ reliability. Sell-side analysis who have covered this company for years still ask questions on earnings calls showing a basic lack of understanding of the business, key levers, and the financials.
Under GAAP, any positive changes in the company’s forecasted cash flows are recognized as revenues over the life of the loan. Negative changes are recognized as upfront expenses.
CACC holds thousands of loans. GAAP requires that any time a single loan has an unfavorable forecast change, CACC must expense that whole value in the current period. If that same loan suddenly becomes performing again, CACC cannot claw back that provision, and instead must count against it over the life of the loan through higher realized revenues.
This creates an income statement that shows consistent incremental provisions for loan losses even though individual loans fluctuate between underperforming, performing, and outperforming, largely cancelling each other out over the lifespan of a pool of loans.
To provide a better economic picture, CACC provides adjusted earnings that recognize both positive and negative variances from the company’s initial forecast over time. This method has proven accurate and trustworthy. Adjusted GAAP financials have shown only $50m lower cumulative net income compared to standard GAAP financials over a trailing 18-year period.
The new CECL (Current Expected Credit Losses Accounting) adds another layer of complexity, contributing to further misunderstanding of the business. Before, the company would carry loans at 70% of their face value on its books (it’s initial forecasted collections) and then add or remove loss provisions based on changes in its forecast from that point. Now, the company must carry loans at their full-face value and recognize a provision for losses immediately to account for the 30% it never expected to collect. On the balance sheet, you see a higher loan amount and a higher provision for losses. CECL just generates larger credit provisions at origination that are offset by higher GAAP revenues over the life of the loan. The net carrying value of loans, economic profit, and timing of cash flows remain exactly the same.
Once you peel back the layers, you find that the accounting makes sense, is conservative, and accurately reflects the ongoing economics of the business. To state that it does not after these methods have successfully existed for nearly 20 years ignores reality. This is especially true given that the company returned over $1.79 billion via share repurchases on a cumulative $4.18 billion of GAAP net income ($3.82b Adjusted Net Income) over the last 15 years, excluding an additional $300m repurchased during the March sell off. It has done that while maintaining significantly less balance sheet leverage than both Santander Consumer and Ally Financial.
The valuation is pretty simple. Assume a historically low 10x P/E multiple from here for a business generating 25%+ ROE, consistently repurchasing shares, operating in a fragmented market with a long runway, and fielding the best/most-aligned managers in the space.
The market is overpricing the impact of legal and regulatory risk on CACC
Subprime Lending Woes & Lawsuits
The MA Attorney General alleges unfair lending behavior, securities fraud, and predatory collection practices. The AG is seeking up to $120 million. On top of this, several other states are investigating the company. It’s easy to look at a pending lawsuit and multiple investigations and determine that the business is untouchable. But let’s analyze the lawsuit and its context.
- CACC lends to people it knows will default
- CACC engaged in unfair collection practices, primarily through excessive collection calls
- CACC used an improper repossession notice form from 2013 to 2018
- CACC uses dealers who mark-up cars to skirt usury limits
- CACC uses dealers who sell vehicle service contracts to skirt usury limits
- CACC used deceptive practices in its ABS offering disclosures
Point (3) arises from recent court rulings in MA. In late 2018, an appeals court in MA determined that the notice form that nearly all auto financing companies were using did not comply with a state statute. In a separate case, an MA court decided in early 2020 that the ruling applied retroactively to incorrect form usage before 2018. Although CACC and presumably all financing companies stopped using the old forms as soon as the 2018 case was decided, the 2020 ruling gave cause of action to the AG for prior violations. PNC Bank, the subject of the adverse ruling in 2020 for loans it made many years ago, has filed an appeal with the MA Supreme Court that will need to resolve before the AG can move forward against CACC on this point.
Point (6) is a non-starter. The premise of alleging securities fraud is based on form (I.E. standard/boilerplate) SEC filings for ABS offerings that state the following:
“The Originator and the Seller do not expect that the characteristics of the Loans and related Contracts purchased during the Revolving Period will be materially different from those transferred on the Closing Date, and each Dealer, Dealer Loan, Dealer Agreement, Purchased Loan, Purchase Agreement and Contract must satisfy the eligibility criteria specified in the Transaction Documents.”
The AG argues that the loans included in the ABS offerings ended up having projected collections 4% worse than what was listed in the offering documents. She alleges this is “material” because CACC management, in its discussion with common stockholders, often says deviations in forecasts >1% are material. The AG is implying that non-defaulting ABS instruments sold to sophisticated parties over many years, containing interrelated terms, and involving offsetting cash collateral considerations are fraudulent based on a form clause and a contrived interpretation of material. To allege fraud, you must show intent as well. The AG offers no evidence or statement showing that CACC executives and underwriters intended to consistently place lower performing loans into the pools. To add onto the kitchen-sink nature of this allegation, the AG does not list even one complaining party. Is there any investor in MA that even bought into those specific ABS pools?
So, we are left with Points (1), (2), (4), and (5). This is the usual gamut of disliked subprime lending practices.
Let’s try to back out CACC’s exposure by using Santander’s record $550m settlement for similar infractions as an example. Santander’s payout rested on multiple compliance failures. First, Santander did not verify the income of borrowers. Second, Santander routinely provided loan modifications without adequate disclosures to consumers. Third, Santander routinely turned a “blind eye” to dealers who engaged in predatory activity such as openly conditioning loans on Vehicle Service Contracts (VSCs).
CACC’s exposure relative to Santander is mitigated in two key ways: 1) its controls are much stronger (see management competence above) and 2) it originates about 1/3 the loans as does Santander.
Dealers offer VSCs and increase nominal vehicle values to economically justify offering these loans in the face of a ceiling on interest rates. The FDIC commissioned a study in 2015 that found VSCs and vehicle value mark-ups to be a natural consequence of setting the usury limit below the market clearing price. Whether you agree with them or not, subprime companies have been engaging these practices for decades. State and federal governments, regardless of political affiliation, have made clear that they will extract fines, increase regulation, and demand greater compliance. However, they have never contemplated removing the two fundamental pillars supporting the $120 billion auto subprime market: making loans to people with a high chance of defaulting and allowing ground-level dealers to clear the market through creative offerings.
Here is what Santander’s settlement resulted in when it was similarly accused of Points (1), (4), and (5):
“Under the settlement announced today, Santander is prohibited from originating any loan if the consumer’s residual income – the consumer’s net monthly income minus the consumer’s total debt obligations – is zero or negative. Santander is also required to monitor auto dealers for possible embellishment or falsification of loan applicants’ income information.“
This result shows that CACC’s exposure is limited largely to payouts and increased oversight, not an injunction on its primary business activities. Full stop. To argue otherwise ignores precedent and the actual contents of the lawsuit against CACC.
So, how do we calculate the dollar exposure here?
Good-faith, customer disclosures, and internal controls have been the determinants of consequences. The AG alleges CACC’s dealers used vehicle markups and conditioned loans on service contracts to inflate loan values. She offers an email in which a CACC legal compliance officer tells a field employee to notify a dealership that it cannot make VSCs a requisite for loans. Contrarily to the AG’s assertion, this evidence shows that CACC does not actively encourage the use of VSC and took good-faith steps to ensure dealers did not make them requirements.
The AG was trying to use the email to imply that dealerships whose loans come with VSCs 100% of the time are illegally requiring them. However, she provides a table showing dealership inclusion of vehicle service contracts ranging from 100% to 0% of loans originated, with many 100% dealerships originating <5 loans. Even the dealerships showing 100% VSCs have inconsistent patterns, with many offering VSCs in some years and not in others. This stands in stark contrast to Santander which apparently did nothing when it internally suspected some dealers were requiring VSCs.
Furthermore, nowhere in the complaint is CACC’s vetting of borrowers questioned. Unlike Santander that was verifying less than 10% of borrowers’ incomes, CACC seems to be compliant.
CACC is also insulated from regulatory risk because the portfolio program, representing the ultra-majority of loans made in the last 10 years, do not count as loans to consumers on CACC’s books. They count as loans to dealers, which removes a significant amount of regulatory scrutiny the government can apply to CACC for those loans. (The government must actually go after the individual dealers who underwrote those loans)
Overall, the things CACC is being accused of are significantly less egregious than what Santander admitted to doing. When coupled with the fact that CACC originates about 1/3 the number of subprime loans, the company’s nationwide exposure looks significantly smaller than $550m.
How this will likely playout
This isn’t CACC’s first encounter with lawsuits. Over the next few years, I would expect the AGs of various states to launch lawsuits. Those lawsuits will be bundled via Multi-District Litigation in the federal courts. Eventually, the various AGs will reach a settlement involving loan forgiveness, payouts, and new controls. Although CACC is 1/3 of the size of Santander, let’s assume a full $550m payout. So, 3-5 years from now, CACC will be required to pay $550 million, likely over a period of a few more years after that. That’s not a lot compared to the several billion in market cap the company lost after the MA AG announced the lawsuit. It’s also not a lot compared to the billions in underwriting profits the company will generate before the lawsuit settles, all while buying back shares at a depressed valuation.
Regulatory hardening benefits CACC
Subprime continues to face an avalanche of regulation and scrutiny. There are only two regulations that would fundamentally impair the business: 1) you cannot lend when you expect a high probability of default and 2) individual dealers cannot mark-up vehicles or provide service contracts at all.
Both have far-reaching consequences that go beyond subprime, making them non-starters. Historically, every AG in every state has had ample opportunity to break one or both of the above pillars, but they have not. They know they cannot, or thousands of dealers would go out of business and tens of thousands of people (many of whom do pay their loans and improve their credit) will no longer be served. Instead, as demonstrated with Santander, the AGs have focused on increased compliance and regulation. This will only serve to consolidate the space by reducing competition from smaller players who cannot meet those requirements.