World Acceptance Corporation WRLD

Key data

The business

World Acceptance makes small, high-rate installment loans, a few hundred to a few thousand dollars, to borrowers banks turn away. It runs roughly a thousand branch storefronts across the American South and Mexico, where a customer walks in, signs, and often refinances the same loan again before it is paid off. The money is made on the spread between what it charges, rates that run into the triple digits where state law allows, and what it costs the company to borrow and to write off the loans that go bad. Insurance and fee products sold alongside the loan add a meaningful slice. The moat is the branch network and the file on a repeat borrower: a national bank will not open a storefront in a small Southern town to lend $1,500, and a first-time online lender does not know which of these customers pays. Revenue barely moves, near $585 million in both FY2022 and FY2026. This is not a growth story. It is a spread-and-credit machine dressed as one.

The numbers

Read the earnings line, not the revenue line. The top line has been flat for five years while profit has swung from $21 million to $89 million and back to $35 million. What moves it is credit. FY2026 profit fell by more than half against FY2025 on nearly identical revenue, the whole gap being loans going bad faster than the company had reserved.

The quarters show the seasonality that makes this stock hard to read one print at a time. Fiscal Q4, the January-to-March tax-refund window, is when borrowers pay down and the profit lands, $35 million in the March quarter alone. The two quarters before it posted small losses as provisions built. The June quarter's return to a $6 million profit and $1.33 a share, four times the year-earlier figure and well above the lone $0.58 estimate, is the first sign the bad-debt wave that halved last year's earnings is receding rather than spreading.

Across the cycle the share count, not sales, is the compounding engine. Revenue is flat, but the company throws off roughly $260 million of operating cash a year and spends it buying back its own stock, $132 million in FY2026 alone, shrinking the diluted share count to about 4.7 million. That is why EPS reached $16 in FY2025 while revenue sat where it was five years earlier. The single analyst covering the name looks for EPS to nearly double to $13 in the year to March 2027 and $15 the year after, which assumes credit costs normalize back toward the FY2024-25 run rate.

One honest gap: this bundle carries no delinquency buckets or clean full-year charge-off rate, so the credit read leans on the provision-driven swing in reported profit rather than the loss curve itself. The last carried figure, an 18.7% fourth-quarter-annualized charge-off, is the number to update at the next release. What the multiple is really paying for is 1.6 times tangible book for a lender earning about 11% on that book today but low-twenties in a clean year. It is priced for the depressed present, not the normalized one, provided the credit turn holds.

Management

Insiders are selling into the recovery, not buying it. Over the last twelve months there were no insider purchases and 31 sales worth $31.6 million, led by Prescott General Partners, a long-time large holder that sold $23.5 million across three blocks on a single August day, plan status not disclosed. The pay plan and the capital plan point the same way: this is run not for growth but for per-share value, and the FY2026 buyback, done at prices from the $110 low up toward today, was the most aggressive in years. The record there is mixed, the company also bought heavily in FY2022 near $180 before the stock fell by half. Weigh the buyback discipline against a board that keeps repurchasing into strength while selling personally into it.

How it fails or surprises you

Credit cycle turns down again. The June profit could be a seasonal head-fake. If the subprime borrower keeps weakening into a softer job market, provisions climb again and the mid-year loss quarters deepen instead of shrinking. Watch the next two prints: another quarter where charge-offs outrun reserves would push FY2027 EPS below the $13 the single analyst assumes and reset the multiple.

A rate cap or funding squeeze. The whole model rests on charging triple-digit APRs where state law allows and borrowing wholesale to fund it. A federal rate cap, a state law change, or a lender that pulls the revolving facility would break the spread directly. Long-term debt swung from $448 million to a $677 million peak over nine months (March to December 2025). The funding line is the pressure point.

Credit normalizes and the buyback compounds (right tail). If charge-offs revert to the FY2024-25 level, EPS snaps back toward $15 while the share count keeps shrinking 5% to 7% a year. At an unchanged 1.6 times book the stock re-rates on earnings alone, and the market is paying today for a depressed 11% return on equity, not the low-twenties this book has produced in clean years. The print that shows it first is a September quarter that stays in the black.

Closing thoughts

The credit trend needs several quarters to confirm, so survivability matters first, and World passes: it funds a short-duration loan book that self-liquidates, and even the halved-earnings year threw off $260 million of operating cash. The distribution is wide and roughly two-sided. The fat left tail is a deepening consumer-credit downturn stacked on a leveraged balance sheet at 1.8 times debt to equity. The fat right tail is a clean credit year against a share count shrinking faster than almost any lender its size. What is at risk if credit re-breaks is a return to loss quarters and a book multiple compressing toward the 0.9 times it saw at the last trough. What the upside is worth is a re-rating on doubled earnings.

The bet is still that World Acceptance keeps lending small cash to people with bad credit and that enough of them pay it back, with the wave of defaults easing rather than spreading. The June quarter says it is easing. Two more like it confirm it, and a September quarter back in the red breaks it. Watch two numbers together, the charge-off rate against the reserve build. When the first falls faster than the second, the earnings the buyback is compounding are real.