Why Consumer Finance Valuations Keep Surprising People Who Aren't Paying Attention
Most people who look at consumer finance stocks are reading the same three metrics: net charge-off rates, cost of funds, and revenue growth. That's not wrong. It's just incomplete and it gets you nowhere when you're trying to figure out whether a name is actually undervalued or just looks cheap for a reason. The real work happens in the spaces between those numbers. The modern consumer finance landscape looks nothing like it did five years ago. Servicing platforms have matured, alternative data streams are actually useful now instead of being gimmicks, and regulation has shifted in ways that penalize companies that didn't plan ahead. When I started covering this space, you could spend a week building a model from scratch using SEC filings and Fitch reports. These days you need to pull from payment trend APIs, card-issuing network data, and borrower behavior signals that aren't in any public document. Here's the practical part. The best consumer finance research I've seen recently starts with the balance sheet but doesn't stop there. You need to understand the funding stack in detail — variable rate warehousing, securitization tranches, and wholesale credit facilities. A company might look like it has a 4.5% cost of funds on paper, but if 60% of its debt is floating rate in a higher-for-longer environment and they haven't locked in hedges, that number is about to move against them faster than anyone realizes. I learned this the hard way with one particular lender in 2023 where the annual report showed stable margins while their quarterly supplemental data quietly disclosed a $200 million variable facility repricing mid-year. The stock dropped 18% in three weeks when the market caught up.
What actually separates good work from generic coverage is how you handle the charge-off curve. Everyone looks at trailing 90-day delinquencies. That's lagging. The forward-looking signal is the roll rate from current to 30 days past due, especially in the first 60 days of a new vintage. If you're modeling a originator's profitability and you're using last year's NCO rate as your baseline without adjusting for the shift in borrower quality, you're going to be wrong. The average consumer finance firm's vintage now shows a 3-5% difference in early-stage roll rates compared to pre-2020 behavior, and most models don't capture that adjustment properly. I use a specific workaround when the data gets messy. Instead of relying solely on the company's reported NCOs, I cross-reference their loss emergence patterns against Federal Reserve consumer credit delinquency data at the state level. Certain markets — particularly Arizona and Texas — show delinquency deterioration two quarters before it hits aggregate national numbers. If a lender has heavy concentration in those states and you're seeing the regional signal turn while the company's reported numbers look fine, that's your early warning. This approach typically cuts the gap between when something goes wrong and when the market prices it in from about six weeks down to roughly two weeks. Another area people consistently get wrong is the servicing right valuation. On the surface it looks like a straightforward DCF, but the actual driver is the prepayment speed assumption and how it interacts with the amortization schedule of the underlying loans. When interest rates drop, prepayments accelerate and servicing fees get eaten faster than most models account for. I've seen analysts value servicing rights using flat prepayment assumptions when the 30-year mortgage constant was moving by 50 basis points in a single quarter. That kind of error can swing the servicing asset valuation by 30-40%.
The common pitfall I keep seeing is treating all consumer finance companies as the same sub-sector. They're not. A buy-now-pay-later platform has a completely different risk profile and capital structure than a traditional subprime auto lender, which is different from a personal loan originator using warehouse facilities. The BNPL companies trade on take rate and gross originations because their balance sheet is offloaded through investor syndications. The auto lenders are all about the spread between vehicle loan yields and their securitization costs. The personal lenders sit somewhere in between but lean heavier on wholesale funding. If you apply the same multiple to all three, you're not doing research, you're guessing. Alternative data has gotten genuinely useful in the last couple years. Transaction-level data from providers like Yodlee and Plaid can show you a company's borrower repayment behavior before it appears in quarterly reports. I tracked one regional consumer finance lender whose early repayment rates improved noticeably in their September data even though their Q3 earnings release didn't mention it. The improvement showed up in the transaction stream three weeks before the company filed its 10-Q. That kind of signal doesn't change everything, but it nudges your assumptions in the right direction when you're trying to beat consensus by a margin that matters. Regulatory risk is another area where the surface-level reading is insufficient. The CFPB's recent enforcement actions and proposed rule changes around debt collection and credit reporting aren't abstract — they directly affect operational costs and could force model recalibration for companies that built their underwriting on older data sources. A company might be trading at 8x earnings because everyone assumes regulatory headwinds are already priced in, but if the new rules require a fundamental shift in how they assess borrower ability to repay, the compliance cost could hit operating margins harder than the market expects. I'd recommend tracking the CFPB rulemaking docket directly instead of waiting for secondary analysis. The gap between proposed rules and final rules often contains the details that matter most.
Get the Full Details
One thing I want to flag about current market conditions: the spread environment is compressed in a way that rewards companies with deposit franchises. A lender that can fund growth at 3-4% deposits versus 6-7% wholesale borrowing isn't just slightly better — it's the difference between healthy origination economics and negative spread compression. During my coverage of this space, I've found that companies with strong retail deposit bases tend to outperform during rate transition periods, but that advantage disappears quickly if deposit costs spike without corresponding yield movement on the asset side. The downside of this approach is that it requires access to data sources most individual researchers don't have. The transaction-level data and payment stream analytics cost money. The regulatory tracking takes time. And even with all of it, consumer finance research has a high error rate because borrower behavior is increasingly hard to predict when macro conditions shift rapidly. You can do everything right and still get surprised by a weather event that affects agricultural communities, or a geopolitical shock that moves commodity prices and then consumer spending patterns in a way no model captures. When I run into situations where the data is too thin to make a confident call, I fall back on comparing a company's valuation to its peer group's historical spread relative to broader market indices. If a consumer finance name is trading at a significant discount to its five-year average spread versus the S&P 500 and the discount coincides with an idiosyncratic issue rather than a sector-wide problem, that's usually worth investigating. If the discount is sector-wide, it's probably justified.
