Sezzle Inc. SEZL

Key data

The business

Sezzle is buy-now-pay-later. A shopper splits a purchase into four interest-free payments at checkout, Sezzle pays the merchant, and the consumer pays Sezzle back over six weeks. The money comes two ways: a fee from the merchant, and a growing stream from the consumer, who increasingly pays a monthly subscription that lets them use Sezzle anywhere rather than only at partner stores. It runs 3.2 million active consumers, 81% of them Gen Z and Millennial, and has bolted on a cash-advance product, SezzleCash, averaging about $165 per draw. WebBank originates the loans; the switch that matters is that Sezzle now retains those loans on its own balance sheet up to $150 million, up from $30 million. The moat is a two-sided habit: consumers who checkout with it repeatedly and merchants who need its younger buyers.

The numbers

Growth is real and the beats are consistent, but the profit line is where the story turns.

Revenue rose 52% year over year in Q2 and 10% over Q1, yet net income fell 21% from Q1. That drop is at the operating line, not taxes: operating margin sank from 51% in Q1 to 37% in Q2. Some is seasonal, Q1 is always the peak as tax refunds arrive and losses run light, but the tell is year over year: Q2 net margin of 27% is a hair below the 28% of a year ago, on 52% more revenue. Scale is not dropping to the bottom line, and the reason is loss provisioning rising as the book grows. The watch-item flagged three days ago, credit cost eating the margin, did not hold: Q2's sequential margin break is that provision showing up in the print.

Revenue quadrupled in four years, a 41% compound rate, and the company crossed from loss to $133 million of profit in three. Consensus wants $609 million and $5.25 in EPS this year; 1H already did $285 million and $2.64, so the second half is priced to slow, and the four-quarter beat streak says analysts keep setting the bar low. What the market does not believe, and I think is right not to, is that the 40%-plus operating margins of early 2026 are the run rate. The print that settles it is the provision line: the variant here is whether retaining loans lifts revenue faster than it lifts losses.

Management

Co-founder Paul Paradis runs it, and the insider ledger is one-sided: 69 sales worth $23.5 million against a single $15,000 buy over the past year, with Paradis himself selling roughly $5 million across June and July near $120. Plan status is not disclosed in the filings I have, so I cannot say how much rode a preset schedule. Against that, the company returned capital hard, $65 million of buybacks in 2025 and another $26 million in Q1 2026 alone, though some of that was spent well above today's price. Guidance has been beaten four quarters running.

How it fails or surprises you

Credit turns as the book grows. Sezzle now keeps up to $150 million of loans on balance sheet, so a downturn in its young, thin-file borrowers hits earnings directly through provisioning, not someone else's book. Q2 already showed margin bleeding to loss costs. Watch net charge-offs and provision as a share of revenue next quarter; a second sequential margin drop confirms it.

The multiple compresses to its own history. You pay 26x trailing against a 16x-to-18x band every year since 2023. Heavy insider selling into that multiple, with growth decelerating per consensus, is how a re-rate starts. The print that triggers it is any quarter where revenue growth slips under 40% while losses rise.

Subscription monetization compounds (right tail). If the monthly-fee base and SezzleCash lift revenue per consumer faster than losses across the 3.2 million users, 2027's $6.68 EPS estimate is low. The market is pricing a slowdown, not a step-up. The first sign is subscription revenue outrunning consumer loan losses two quarters straight.

Closing thoughts

The provision line decides this, and you find out within one quarter whether the margin roll from 51% to 37% is seasonal noise or the first crack in a credit book management just chose to hold more of. The left tail is fatter near-term: a 26x multiple on a lender carrying its own loans does not survive rising charge-offs, and what is at risk if credit turns is not a bad year but a re-rate. The right tail is real too, subscription income that compounds faster than losses, but the burden of proof sits with the next two prints, and the insiders are selling into the wait.

The bet is still that younger Americans keep splitting purchases with Sezzle and keep paying it back, now with more of those loans on Sezzle's own books. It breaks the day provisioning outruns revenue: watch net income against revenue, and if net income falls again while revenue rises, the credit cost, not seasonality, is the story.