PayPal Holdings, Inc. PYPL

Key data

The business

PayPal runs a two-sided payments network: on one side, roughly 430 million consumer and merchant accounts; on the other, the online stores that show the yellow PayPal button or run Venmo at checkout. It makes money by keeping a small cut of every dollar that crosses the platform, what the industry calls take rate. Total payment volume grew 10% last quarter. Three engines drive it: branded checkout, the higher-margin PayPal button consumers know; Venmo, the peer-to-peer app now pushing into merchant payments and debit cards; and unbranded processing through Braintree, the plumbing behind large merchants' own checkouts, which carries a far thinner margin. The mix matters more than the total, because a dollar shifting from the branded button to Braintree earns PayPal much less. The moat is the two-sided network and the checkout habit: consumers keep the button because merchants take it, and merchants take it because consumers use it.

The numbers

Read the five quarters as volume against take rate, the only lens that matters for a payments network. Net revenue is the cut the platform actually keeps.

The tension is in the top line against the volume line. Volume grew 10% year over year in Q2, but net revenue grew under 5%, from $8.29B to $8.68B. Strip out the cost of moving that volume and it is starker: transaction expense rose 10.5%, tracking volume exactly, so transaction-margin dollars, the true gross profit of a payments firm, fell 0.5%. Falling credit losses, down 17%, and flat marketing held operating income near $1.5B. On the adjusted line the company earned $1.38 against the $1.28 analysts modeled, three of the last four quarters beating estimates. The trailing net margin held at 14%.

Over four years revenue compounded about 7% a year and net income under 6%, yet diluted EPS compounded 11%, from $3.52 to $5.41. The gap is the buyback. PayPal spent $6.05B retiring stock in 2025, near 13% of today's market value, and has run $1.5B a quarter for the last two years, shrinking the share count about 7% annually net of stock compensation. Free cash flow was $5.6B in 2025, a 12% yield on the market cap. That is the machine: a business growing its cut in the mid-single digits, converting nearly all of it to cash, and using the cash to buy itself in size at ten times earnings. The market prices this as a melting ice cube, the cheapest multiple in its public life, on the view that take rate bleeds until volume growth stops mattering. The other side: at ten times earnings and a 12% cash yield, the share shrink alone compounds EPS near double digits even if net income merely holds. The print that settles it is the gap between volume growth and net-revenue growth, and whether it narrows over the next two quarters.

Management

Capital allocation is the clearest tell here and it reads well: $6B a year to buybacks at a trough multiple, funded entirely from cash flow, with net debt under one turn of EBITDA. The insider ledger is quieter and mixed. Over the last year one open-market purchase, $255K by chief financial officer Jamie Miller in June, against twelve sales totaling $1.5M, the largest a $500K disposition by Frank Keller in April. Plan status is not disclosed in the filings reviewed, so read the sales as routine compensation-driven selling rather than signal; the CFO reaching into her own pocket for stock at $54 is the more interesting data point. Pay is weighted to stock, which aligns management with the per-share compounding the buyback drives.

How it fails or surprises you

Take rate keeps bleeding (downside). Q2's tell: 10% volume growth produced a 0.5% decline in transaction-margin dollars, because unbranded Braintree volume, which earns a fraction of the branded button, grows fastest. If that mix keeps shifting, net revenue growth stalls in the low single digits and the buyback cannot outrun it. Watch the volume-versus-net-revenue gap each quarter; it widened, not narrowed, into Q2.

Branded checkout loses the consumer (downside, the hardest fact). The number this read explains least is the declining operating margin the company flagged itself: transaction expense grew exactly with volume while the premium branded slice did not keep pace. If Apple Pay and one-click rivals are quietly taking the checkout habit, the whole flywheel weakens. The print that proves the bear right is branded checkout volume growth turning negative.

The buyback re-rates the stock (right tail). At ten times earnings, $6B of repurchases retires about 7% of the shares a year net of stock comp. Hold net income flat and EPS still compounds double digits; let the multiple drift back toward even 15 times, far under its 30-to-50 history, and the stock is up about 45% with no operating heroics. The market pays nothing for this today. The print: share count in each 10-Q, plus any net-revenue reacceleration.

Closing thoughts

The payoff turns on whether PayPal can narrow the gap between payment volume growth and net revenue growth. That gap was five percentage points in Q2, volume up 10%, net revenue up under 5%, and transaction-margin dollars actually fell while volumes rose. If the gap narrows over the next two quarters, the market re-rates a cash machine it wrote off; you're looking at a move from ten times earnings toward even 15 times, about 45% upside with no operating heroics. If it widens, the branded button is quietly losing to one-click rivals and the buyback cannot outrun a melting margin. The upside is fatter, on judgment: the balance sheet is clean, the buyback is $6B a year at a trough multiple, and the 12% free cash flow yield pays you to wait while the share count does the work.

The bet is still that hundreds of millions of people keep paying with PayPal and Venmo, and that the company keeps enough of its cut for the volume to matter. What breaks it is simple to watch: net revenue growing at half the speed of payment volume, quarter after quarter, while branded checkout volume rolls over. Those two lines, the volume-to-revenue gap and branded checkout growth, tell you first whether you own a compounder buying itself cheap or a franchise quietly melting. Everything else is noise.