Show me a lender bragging about record originations, and I'll show you a lender who still has some explaining to do.

How many of those loans will make their first payment?

How many will reach 30 days past due?

How much did the lender spend acquiring those borrowers?

What will the company have left after credit losses, funding, servicing, collections, and operating expenses?

Funny how those questions can take the celebration right out of a record funding month.

I've spent more than two decades in the business of lending money to people other lenders frequently turn away. Payday loans. Installment loans. Car title loans. Storefronts. Online lending.

One lesson keeps repeating itself.

Three major publicly traded lenders recently reported earnings. They operate different businesses, serve different borrowers, and use different accounting methods.

But their results reveal three very different operating bets.

And every owner, CEO, credit officer, and multi-store manager ought to recognize the choices.

ONEMAIN: GROW THE PORTFOLIO. MANAGE THE CREDIT COST.

OneMain Holdings reported third-quarter 2026 earnings on October 30.

Consumer loan originations reached approximately $4.5 billion, up from $3.9 billion a year earlier.

Managed receivables reached $28.1 billion, an increase of 9%.

That's meaningful growth.

But the provision for finance receivable losses increased from $488 million to $615 million.

That's an increase of approximately 26%.

Before somebody jumps to the conclusion that OneMain's underwriting deteriorated, there's an important accounting distinction.

A provision for credit losses is not the same thing as loans actually charged off during the quarter.

OneMain reported that $146 million of its provision reflected an increase in the allowance for finance receivable losses, driven by receivables growth.

Its reported net charge-offs were approximately $469 million for the quarter, versus $427 million a year earlier.

Those are different numbers measuring different things.

Now look at delinquency.

Delinquency improved year over year. The charge-off ratio was higher than a year earlier, although it improved from the preceding quarter.

Meanwhile, net income declined from $199 million to $185 million.

Is OneMain making a bad bet?

Not necessarily.

Growing receivables requires capital. It can require additional reserves. And the lifetime economics of newly originated loans won't be fully visible in the quarter they're funded.

The question is whether the additional loans will produce enough lifetime revenue to cover their expected losses and costs.

What a storefront operator should learn

Suppose your store manager increases monthly funding by 20%.

Wonderful.

Now ask whether the additional loans came from better conversion, stronger demand, higher loan amounts, or weaker underwriting.

Those are four different explanations.

And they can produce four very different financial outcomes.

OneMain's bet: Expand lending while managing the credit and capital costs associated with that growth.

The lesson: Growth is not free. Somebody has to fund it, service it, reserve for its risks, and collect it.

WORLD ACCEPTANCE: PROTECT CREDIT QUALITY, EVEN WHEN NEW-CUSTOMER GROWTH SUFFERS.

World Acceptance reported fiscal 2027 second-quarter results on October 23.

Its numbers tell a different story.

New-customer loan volume fell 18.1% from the prior-year quarter.

Former-customer loan volume declined 9.9%.

Refinancing volume increased 4.4%.

And the total customer base declined 5.4% over the preceding twelve months.

World Acceptance explained that it had tightened new-customer underwriting because of its portfolio composition and economic uncertainty.

It has since loosened underwriting and expects to increase new-customer lending carefully.

That is a familiar operating problem.

Tighten approvals and you may reduce the number of bad loans.

But you may also reduce the number of good loans.

And eventually, if you aren't replacing customers, the portfolio ages.

Here's the number I want operators to study.

World Acceptance originated approximately $11.49 in refinancing volume for every $1 of new-customer origination volume.

That is a ratio of dollars originated, not a ratio of customers or profit.

Refinancing volume can include replacement of existing principal. It should not be confused with incremental new money advanced or incremental revenue earned.

But it demonstrates the enormous role existing-customer relationships play in this business.

Now consider credit performance.

World Acceptance's 0–60-day recency delinquency ratio improved from 20.0% to 18.9%.

Its 61-plus-day ratio improved from 5.8% to 5.4%.

However, its annualized net charge-off ratio increased from 17.1% to 17.6%.

So no, every credit metric did not improve.

Still, the company reported $8.8 million in quarterly net income, compared with a $1.7 million loss in the prior-year quarter.

That improvement cannot be attributed solely to tighter underwriting. Revenue, loan mix, expenses, and other factors matter.

The repeat-customer question

Every lender knows acquiring a new borrower costs money.

Advertising. Lead generation. Employee time. Verification. Fraud screening. Declined applications.

And you pay many of those costs before collecting a dime.

A properly qualified returning borrower may have a known payment history, a lower acquisition cost, and a more predictable risk profile.

But repeat status alone does not make a borrower profitable or a new loan appropriate.

Past repayment is evidence. It isn't a guarantee.

And refinancing can become a problem when lenders focus on transaction volume rather than the customer's financial condition and the economics of the replacement loan.

World Acceptance's bet: Accept reduced new-customer volume while managing portfolio quality and relying heavily on existing-customer activity, then carefully reopen the new-customer funnel.

The lesson: The customer you already know may be worth more than the stranger you just paid to acquire. Prove it with the numbers.

ENOVA: SCALE DIGITAL ORIGINATIONS WHILE KEEPING CREDIT PERFORMANCE UNDER CONTROL.

Enova International reported third-quarter results on October 22.

The numbers are hard to ignore.

Enova originated approximately $2.6 billion in loans and finance receivables during the quarter.

Its combined loans and finance receivables reached approximately $5.9 billion.

Revenue increased 26% to $1.01 billion.

Net income increased from approximately $80.3 million to $107.3 million.

And its consolidated net charge-off ratio declined from 8.5% to 8.1%.

That's growth accompanied by stronger reported earnings and a lower consolidated charge-off ratio.

But here's where operators need to read beyond the headline.

Enova's reported figures include different products and both consumer and small-business lending. Its combined loan measures also include certain third-party-originated loans associated with credit services organization programs.

Its consolidated net charge-off ratio cannot be directly compared with World Acceptance's annualized net charge-off ratio or OneMain's consumer-loan ratio.

Different products. Different denominators. Different accounting.

And Enova reports changes in fair value for its loan portfolio, making a simple comparison of revenue with another lender's interest income misleading.

Here's the part that interests me.

Enova reported approximately $229.7 million in quarterly marketing expense, compared with $147.4 million a year earlier.

That's an increase of roughly 56%.

Originations increased 31%.

That doesn't prove customer acquisition became less efficient. Marketing spending and current-quarter originations don't necessarily match the same customer cohorts, and product mix can change.

But it certainly gives management a question worth answering.

How much incremental lifetime profit is each additional marketing dollar buying?

That is the number I would want before increasing the advertising budget.

The storefront comparison

Enova doesn't need a branch on every corner.

Its online model offers different distribution economics.

But that doesn't automatically make the online model superior.

Storefront lenders have advantages that may include local recognition, face-to-face customer service, repeat relationships, and employee judgment.

Those advantages are only economically valuable if they produce better conversion, better repayment, lower acquisition costs, stronger retention, or some other measurable benefit.

Otherwise they're just expenses with a good story attached.

Enova's bet: Use online distribution, analytics, diversified products, and funding capacity to scale originations while maintaining acceptable risk-adjusted economics.

The lesson: Technology is useful when it makes better decisions, reduces costs, or improves outcomes. Otherwise it's another bill.

THREE COMPANIES. THREE DIFFERENT BETS. NONE OF THEM IS YOUR BUSINESS.

Here's the comparison.

These are my interpretations of the companies' reported operating choices, not formal strategy labels adopted by their management teams.

There is no honest way to declare one company the winner from a few headline KPIs.

But there's a very useful way to apply their experience to your own business.

Stop asking how much you funded.

Start asking how much those loans are expected to earn.

THE $1 MILLION ORIGINATION TRAP

Let's put some real operating math on the table.

Suppose your lending company funds 1,000 additional loans averaging $1,000.

Your origination report proudly announces:

$1,000,000 IN ADDITIONAL LOANS FUNDED.

Marketing is celebrating.

The stores are celebrating.

The CEO is thinking about opening two more locations.

Now let's examine the expected lifetime economics of those additional loans.

A million dollars funded.

Seventy thousand dollars in expected contribution profit.

Not bad, assuming the forecast is accurate and the return adequately compensates the capital and risk.

Now suppose the underwriting team misjudged credit losses.

Instead of $180,000, losses reach $260,000.

Everything else stays the same.

That's what happens when a lender confuses production with performance.

The employees got their funding numbers.

The marketing company got paid.

The loan-management vendor got paid.

The payment processor got paid.

And the owner got the bill.

One more important distinction: this is a lifetime contribution model, not a cash-flow forecast. A growing lender also needs sufficient liquidity to fund principal advances while waiting for collections and repayments.

THE NUMBER YOUR DAILY REPORT PROBABLY DOESN'T SHOW

Most lenders can tell you yesterday's applications.

They can tell you approvals.

They can tell you funded loans and average loan size.

Some can tell you FPD.

Fewer can tell you what the last 100 loans are expected to contribute to profit after every meaningful variable cost.

And that's the number management ought to be working toward.

Start with seven measurements.

1. Applications and funded conversion. Where are the customers coming from, and how many actually become funded borrowers?
2. New versus repeat borrowers. What are the differences in acquisition cost, repayment, and lifetime contribution?
3. First-payment default. Which stores, employees, lead sources, and borrower cohorts produce early payment failures?
4. Delinquency migration. How many accounts move from current to 1–30, 31–60, and later stages?
5. Net credit losses. How much principal is ultimately lost after recoveries?
6. Fully loaded variable costs. What do acquisition, verification, servicing, payment processing, collections, and funding actually cost?
7. Contribution profit by vintage. Which groups of loans produce an acceptable return, and which destroy value?

FPD should be measured on loans that have actually reached their first-payment due date, using a consistent definition.

And when comparing loans over time, use comparable vintages and observation periods.

Otherwise you'll mistake immature loans for good loans.

THE QUESTION I'D ASK EVERY STORE MANAGER TOMORROW MORNING

Imagine walking into a store and asking the manager:

"Of the last 100 loans you funded, which 20 are most likely to lose money?"

Most managers would struggle to answer.

Not because they're stupid.

Because nobody built a system that gives them the answer.

Now imagine the manager can see:

• Which new-customer cohorts are producing excessive FPD.
• Which lead providers are sending borrowers who don't make their first payment.
• Which employees are making too many policy exceptions.
• Which loan sizes are producing poor risk-adjusted returns.
• Which repeat-customer segments produce reliable repayment and acceptable profit.

That manager can make better decisions tomorrow.

Not next quarter.

Tomorrow.

Of course, no risk model can know in advance exactly which individual borrowers will default. These are probabilistic assessments that must be validated and used in compliance with applicable lending laws.

But even an imperfect, properly tested management system can be more useful than a beautiful report nobody acts on.

THE REAL BET: KNOW YOUR NUMBERS BEFORE YOU PUT MORE MONEY AT RISK.

OneMain is demonstrating what growth looks like when it requires substantial capital and credit-loss provisions.

World Acceptance is demonstrating the difficult balance between protecting the existing portfolio and bringing new customers through the door.

Enova is demonstrating how digital distribution and scale can support rapid growth when credit performance and operating economics cooperate.

Different strategies.

Different customers.

Different risks.

But one question connects all three.

If you can't answer that, your origination target is only half a business plan.

And probably the easier half.

WANT TO PUT THESE NUMBERS TO WORK?

I spend my time helping lenders understand the operating decisions behind the numbers: underwriting, customer acquisition, collections, portfolio performance, and the systems that connect them.

There are three ways to go further.