Euronet Worldwide, Inc. EEFT

Key data

The business

Euronet Worldwide sells electronic payments in more than 200 countries through three engines. EFT Processing owns and operates ATMs, heavily weighted to European tourist corridors, and licenses card-processing software to banks. epay distributes digital and prepaid content, the gift-card rack and mobile top-up at the supermarket till, through hundreds of thousands of retail points. Money Transfer is Ria and xe, plus the Dandelion cross-border rails, wiring remittances home for migrant workers. The moat is physical placement and a settlement network: a cash machine in a Lisbon airport and a remittance agent on a Manila street corner are both slots someone else already filled, and the money that flows through them earns a fee twice. The high-margin dollar is the conversion charge on a foreign tourist's withdrawal, which is exactly the dollar regulators and cardless travel both threaten.

The numbers

Revenue is up, profit is down. The last five quarters show the seasonal shape, Q3 peaks, Q1 troughs, but Q2 2026 profit fell 21% against Q2 2025 even as revenue rose 3%.

Revenue grew 3% in Q2 2026 against a year earlier, but operating income fell to $137.1M from $158.6M, down 14%, and net income dropped 21%. Adjusted EPS of $2.82 came in under the $2.93 the Street carried, the second miss in the last three quarters. The pressure sits in the richest segment, the European ATM fee.

Across FY2021 to FY2025 revenue compounded 9% a year, but EPS ran faster: from $4.41 in FY2022 to $6.84 in FY2025 is 16% a year, while net income over that same stretch grew 10%. The gap is the share count, bought down year after year. That is the whole engine, mid-single-digit revenue turning into mid-teens per-share growth through buybacks and operating leverage. First-half 2026 broke the pattern: EPS of $2.54 is down 19% from $3.12 a year earlier. What I believe the market is missing is that the ATM softness is travel-cycle and mix, not terminal decline, and the settling print is Q4 operating margin against the December quarter a year ago. If that comes in soft too, the discount is earned, not a gift, and I was wrong.

Management

The company reports Michael Brown as CEO with tenure dating to 1994; independent verification of current executive roster was not completed this run. The insider record this year points one way: director Thomas McDonnell bought $200,610 in the open market on May 26 and added $55,139 on August 7, both near the 52-week low, against a single $24,850 sale by another insider; plan status is not disclosed on the purchases, but open-market buys after a drop are the buys that mean something. The buyback is the capital-allocation story and it has done real work, shrinking the count enough to push per-share growth well above net income growth. The open question is price paid: repurchasing at 11x is far better than the 20-plus times paid in richer years, and management has kept doing it, which is the right instinct if the earnings hold.

How it fails or surprises you

ATM and conversion-fee erosion. The European ATM segment carries the highest margin and the most risk. Dynamic currency conversion, the markup added when a tourist's card is charged in home currency, draws regulatory attention, and cardless travel thins the volume. Q2 operating income already fell 14% on higher revenue. The print: EFT segment operating income next quarter.

The cheap multiple may be a value trap. The fact the 11x explains least is that revenue rose while profit fell across all of first-half 2026. If operating margin keeps sliding, the discount is deserved and you own a slow-growth cash machine at a fair price, not a gift. The print that would prove the read wrong: Q3 2026 operating income below the $195.0M of Q3 2025.

Money Transfer and buybacks re-rate it (right tail). Ria, xe and Dandelion move cross-border volume that compounds while the market pays nothing for it at the 8th percentile of a twelve-year multiple. If remittance keeps growing and management retires stock at these prices, EPS growth resumes off a smaller count and the multiple re-rates toward its old high teens. The print: Money Transfer revenue growth alongside a lower share count.

Closing thoughts

The stock sits at 11 times earnings, near the bottom of its twelve-year range, while first-half profit fell 19%. The fork is clean: either the ATM and conversion-fee pressure is structural, and the discount is earned, or it's travel cycle and mix, and the December quarter reverses it. The resolving print is Q4 operating income against last year's December quarter: if it falls again on rising revenue, the thesis breaks and the cheap multiple is a trap; if it recovers toward that $195M Q3 2025 level or better, the compounding resumes. The fatter tail leans up, insiders bought here, the buyback keeps grinding, and the share count keeps shrinking, but the regulatory risk on conversion fees is genuine and it could break left. What's at risk if the downside hits is the thesis itself: you'd own a slow-growth processor at a fair price, not a mispriced compounder.

The bet is still that Euronet's tourist ATMs, the gift-card racks, and Ria's remittance corridors keep throwing off cash, and that the fee pressure on those machines is cyclical, not terminal. It breaks the way it bent this year: revenue up, profit down. Watch operating income against revenue, quarter versus year-ago: the $137M on $1.11B this Q2 against $159M on $1.07B a year before. If the next print shows revenue up and operating income down again, the discount is earned, not a gift, and I was wrong.