The Watson Weekly is sponsored by Avalara — the agentic AI platform automating global tax and compliance for leading eCommerce brands with simple plans, transparent pricing, and complete coverage across Shopify, Stripe, WooCommerce, BigCommerce, and more.

Both of these companies grew earnings this quarter by getting smaller in the places that matter most, and both stocks got hit for it anyway.

UPS reported $22.8B in revenue, up 7.4%, with adjusted EPS of $1.76 against a $1.65 consensus. It raised full-year revenue guidance to roughly $91.2B and adjusted EPS to about $7.22. GAAP net income fell to $604M, or $0.71 a share, from $1.28B a year ago. Shares fell close to 7% on the day.

The domestic beat came off one line. U.S. revenue rose 6.0% on a 9.3% increase in revenue per piece. That is price, not demand. Volume is moving the other way, and Carol Tomé told analysts U.S. average daily volume declines mid-single digits in the third quarter, with smaller per-package revenue gains in the back half as the fuel surcharge benefit fades. GAAP domestic operating margin came in at 0.1% against an adjusted figure of 8.0%. Sitting in that gap is $891M of after-tax separation cost from the Driver Choice Program, which is what it costs to buy out drivers you no longer need. The Amazon glide down is finished, and the pricing power that came with walking away from it has a shelf life that management just put a date on.

PayPal ran the same play on a different asset base. Revenue came in at $8.68B, up 5%, with adjusted EPS of $1.38 against $1.28 and total payment volume of $486.4B, up 10%. Full-year adjusted EPS guidance moved to $5.38 and transaction margin dollars to roughly $15.6B, which reverses a prior call for a slight decline.

The margin lines run the other direction. Transaction margin dollars grew 1%, or 3% once you strip out interest on customer balances. Adjusted operating margin compressed 248 basis points to 17.4%. Active accounts finished at 439M,, up 0.3% year over year and down 0.2M from the prior quarter.

Online branded checkout grew 2%, and full-year guidance for that line is now low single digits, which the company frames as a modest improvement. Everything carrying the quarter sits outside the checkout brand. BNPL volume grew 26%, Venmo TPV grew 14%, and Braintree volume came in up mid-teens, which is a lending book, a peer-to-peer wallet, and a merchant processor. The yellow button that every PayPal bull thesis has rested on for a decade is close to flat, and Enrique Lores is financing a five-pillar rebuild around it out of at least $1.5B in gross run-rate savings, roughly $400M of which lands this year.

If you ship physical goods, the number that matters is 9.3%. Your per-package cost keeps climbing inside a network that keeps shrinking, and the fuel-driven portion of that increase is guided lower from here. Model the back half yourself rather than taking the carrier's word on it, because the carrier has already told you the tailwind is thinning.

If you run a checkout page, PayPal's own disclosure now points its growth somewhere other than the button. That should change what you ask for in the next contract, and who inside PayPal you go to for it.

Both companies raised guidance on the strength of what they stopped doing. Neither has shown me the quarter where the raise comes from something they started.

Crown Brands Group bought a trademark and some inventory, and the number that matters is the one the release leaves out.

Calida paid $80M for Cosabella in 2022, earn-out included. At the time the brand generated $29M in US sales and $4.8M in EBITDA. In the first half of 2026, Cosabella turned over CHF 4.3M, down 36.6%, about 4.6% of group revenue. Calida has not said what it recovered. Its stated reason for selling is focus, which is what sellers say.

The Italian framing does some work here. Valeria and Ugo Campello founded Cosabella in Miami in 1983, with production contracted in Italy.

Raymond Dayan calls the Hanky Panky and Cosabella pairing a clear leadership position in premium intimates. Look at the operating structure before accepting that. Crown directs brand strategy and licensing. Rafar Group, the Gelmart parent that already holds the Hanky Panky license, runs design, development, eCommerce, and distribution. Dreamwear takes sleepwear and loungewear. Crown collects royalties and manages a name. It says more heritage brands are under evaluation.

The model works when the licensee protects price. It fails quietly when the licensee fills open wholesale capacity. Calida spent four years learning how hard that distinction is to hold