Good morning. Record revenue is only the beginning of today’s earnings story. Google’s cloud doubled, Microsoft squeezed more work from its GPUs, and Lilly lifted its sales outlook by $5.5 billion at the midpoint.

Also inside: ServiceNow’s contract outlook, DoorDash’s Instacart deal, Mastercard’s cash conversion and Caterpillar’s broader profit growth. Updated through the 8:30 a.m. ET inflation release on October 29.

Google’s cloud doubled. Its cash flow went negative.

Justin Sullivan / Getty Images, via ABC News. 2023 archive.

Google proved AI demand is real. Turning that demand into cash is taking longer. Alphabet’s Q3 revenue rose 25% to $127.6 billion. Cloud revenue doubled to $30.7 billion, while Search grew 12%, down from 17% in Q2.

Management called foreign exchange the largest reason for Search’s slowdown, alongside tougher advertising comparisons. It also acknowledged that AI Mode queries currently monetize at a lower rate. More questions create potential ad inventory; they do not immediately generate proportionate sales.

Cloud earned $11.1 billion in operating profit, a 36.2% margin. But TPU hardware sales contributed meaningfully to growth. Selling systems produces different revenue timing and inventory needs from renting computing capacity. Third-party capacity also arrived late in Q3, limiting rental costs; management expects a larger cost burden in Q4 as usage ramps.

The cash calculation: $42.3 billion from operations minus $51.4 billion of capital spending left negative $9.1 billion of free cash flow. A $5.2 billion European Commission fine payment and inventory buildup also hurt cash generation. Blaming the entire deficit on AI would miss those effects.

Operating income grew 30%, while investment losses weighed on EPS.

Our read: the next improvement needs to come from monetizing AI searches and converting cloud demand into cash. Management retained its $195–$205 billion annual capex range and expects a significant increase next year. Rising depreciation and rental costs mean today’s cloud margin is not a permanent entitlement.

Microsoft earns more from its GPUs. The next bill is coming.

Coolcaesar / Wikimedia. Redmond, 2016. · Creative Commons license

Microsoft beat its cloud outlook partly by getting more work from equipment already installed. Fiscal Q1 revenue rose 19% to $92.3 billion, above its $89.85–$90.95 billion guidance. Azure grew 48%, versus the earlier 44–45% constant-currency outlook; currency had no material effect on reported Azure growth.

Engineering improvements freed GPU capacity, better CPU utilization created additional billable work, and higher revenue per token helped. That is a stronger economic explanation than simply buying more chips: existing infrastructure produced more revenue.

Operating income grew 21% to $45.9 billion. Still, Microsoft Cloud’s gross margin fell to 66% as Azure became a larger share of the business. Better efficiency inside a product can coexist with a less profitable overall sales mix.

Capital spending including finance leases was $45.2 billion, below the prior above-$50 billion expectation because deliveries slipped. Cash equipment spending was $39.4 billion. Against $58.0 billion of operating cash flow, that left $18.6 billion of free cash flow.

The next-quarter outlook pairs 49–50% constant-currency Azure growth with capex above $50 billion. Spending then accelerates in the fiscal second half. Delayed equipment bills are still bills.

Smead Capital’s October capital-intensity critique questions applying software economics to infrastructure-heavy businesses. This quarter provides counterevidence through better utilization and positive cash generation. It does not yet establish the lifetime return on the full buildout.

Lilly adds $5.5 billion to its sales outlook. Pricing still bites.

Shutterstock, via WFYI. Headquarters archive.

Lilly’s treatment growth again overwhelmed falling prices. Q3 revenue rose 39% to $24.4 billion. Worldwide volume increased 53%, offset by a 15% pricing decline. Mounjaro sales reached $10.7 billion, up 65%; Zepbound reached $5.4 billion, up 49%.

The company lifted full-year revenue guidance to $91–$92 billion from $85–$87 billion: a $5.5 billion increase at the midpoint.

The geographic split explains the trade-off. U.S. volume rose 41% with prices down 7%; international volume rose 76% with prices down 29%, including China reimbursement effects. Wider access adds patients while reducing revenue retained per treatment.

There was also roughly $900 million of favorable milestone and rebate-related revenue, including a $410 million Jardiance milestone. These benefits should not be treated as a recurring prescription growth rate. Q2’s volume gain was 60%, with prices down 13%: growth remains strong, but the pricing pressure intensified.

The apparent EPS-guidance cut needs unpacking. The new non-GAAP range is $35–$36, versus $35.50–$36.50 previously. Newly included acquired-research charges cost $4.57 per share; underlying improvements offset most of that charge. Even Lilly’s non-GAAP earnings retain these acquisition-related research expenses.

The sales upgrade reflects genuine demand, but valuation should rest on repeatable treatment economics. More patients and manufacturing efficiency can support profits; milestone payments cannot recur indefinitely, and the volume-price gap still matters.

ServiceNow lifts its AI target. Contract growth cools.

ServiceNow office archive, via IT Brew.

ServiceNow’s AI opportunity is expanding faster than its near-term contracted revenue. Q3 subscription revenue rose 21% to $3.99 billion, slightly above its prior $3.975–$3.980 billion range. Constant-currency growth was 20.5%, versus 20% guided.

Management raised its year-end AI annual contract value target to $1.7 billion, from $1.5 billion in May. Annual contract value measures contracted business; it is not revenue already recognized. Renewals remained 98%, supporting the case that customers still depend on these workflows.

The forward signal is more mixed. Current remaining performance obligations—contracted revenue expected within 12 months—grew 21% in constant currency, versus 21.5% in Q2. Management expects 18.5% growth in Q4, even as it forecasts subscription growth of 21.5% on the same currency basis.

Contract structure explains part of that divergence: termination-for-convenience provisions can exclude business from remaining obligations while it still produces revenue. On-premise revenue timing also matters. Slower obligations growth therefore deserves scrutiny without automatically proving customers are leaving.

The full-year subscription midpoint increased only $15 million. Meanwhile, the adjusted operating-margin outlook rose to 32.5%, from 31%, and the free-cash-flow margin outlook reached 36%, helped by lower capex. Those are adjusted measures, not GAAP operating margins.

Embedded workflows and customer retention support the AI expansion case. But converting larger AI contracts into sustained revenue growth matters more than the contract-value headline alone. Lower investment can improve cash today without demonstrating stronger demand tomorrow.

The Mag 7 + three: Wednesday’s close.

Micron led this watchlist. SPY gained 0.34%; QQQ rose 0.60%. These are ETF price returns, not index returns.

October 28 adjusted regular-session closes versus October 27. C&C editorial watchlist, not an official index. Google and Microsoft reported afterward; these figures exclude their earnings reactions.

DoorDash’s $17 billion grocery deal comes with a reinvestment bill.

DoorDash signage, via Bloomberg. 2024 archive.

DoorDash is buying grocery scale, but the promised savings will not immediately flow to shareholders. It agreed to acquire Instacart for approximately $17 billion, or $65 a share, with at least 75% cash and up to 25% stock, subject to election and proration rules.

Management sees a path to at least $500 million of annual synergies, net of investment. It plans to reinvest more than all the initial synergies in the first year after closing, with positive net synergies from year two. The transaction is expected to close in late 2027 or the first half of 2028, subject to approvals.

The operating logic is delivery density: more grocery orders can spread fulfillment costs across a larger network. Cash on hand and new debt fund the cash consideration, so financing costs belong beside the savings forecast.

DoorDash enters the deal with improving operations. Q3 revenue rose 34% to $4.62 billion, but growth excluding Deliveroo was 23%. Adjusted EBITDA increased 41% to $1.06 billion; free cash flow rose a much slower 9% to $787 million.

This is a multiyear distribution investment, not an immediate earnings shortcut. Q4 adjusted EBITDA guidance of $900 million–$1.1 billion brackets Q3’s result. Integration, reinvestment and regulatory review could delay the financial payoff even if order density improves.

Mastercard’s services outgrow payments. Cash conversion lags.

Mastercard office photograph, via Built In.

Mastercard earned more from each dollar moving across its network, while services added another growth engine. Q3 revenue rose 14% to $9.81 billion; adjusted EPS increased 21% to $5.31. Payment-network revenue grew 11%, while value-added services and solutions grew 19% to $4.06 billion.

Underlying activity was healthy: purchase volume rose 9%, cross-border volume 11%, and switched transactions 9%. Faster revenue growth reflects pricing, mix and services rather than an equivalent acceleration in consumer spending. Rebates and incentives rose 18% to $6.37 billion, absorbing part of the gross revenue gain.

The adjusted operating margin still expanded to 60.5% from 59.8%. Buybacks contributed about 16 cents to EPS growth, while lower taxes also helped. The earnings increase therefore combines business performance with financial effects.

The cash statement is the counterweight. Nine-month operating cash flow fell about 6% to $11.86 billion. Larger prepaid-asset outflows and legal-settlement payments weighed on cash generation; this is not evidence that quarterly payment demand suddenly collapsed.

October’s early volume data remained constructive, but results do not support an unqualified consumer boom. Services and network economics are doing useful work, while incentives and cash timing explain why revenue, EPS and cash tell different stories.

Caterpillar’s profit surge reaches beyond data centers.

Caterpillar headquarters sign, archive via Mergr.

Caterpillar’s growth is broader than the AI power story. Q3 sales rose 17% to $20.68 billion, while adjusted EPS increased to $7.20 from $4.95. Higher volumes added approximately $2.2 billion to sales; pricing added another $663 million.

Power & Energy generated $2.14 billion of segment profit, up 31%. Construction Industries earned $1.83 billion, up 33%. All three primary segments increased sales, supporting a stronger industrial picture than a single data-center equipment boom would imply.

Volume spreads factory costs over more machines, while favorable pricing supports profit per sale. Together they helped lift adjusted operating margin to 20.6%, from 17.5%. However, operating profit also included $128 million of tariff recoveries. Treating that entire margin gain as permanent would overstate the improvement.

Enterprise operating cash flow reached $4.4 billion. That is not the same as machinery, power and energy free cash flow: the enterprise figure includes the financing business and precedes capital spending.

The quarter backs management’s earlier expectation for stronger volume and pricing. Cyclical demand and tariff costs remain the countercase; power equipment and construction both deserve credit.

Inflation eats September’s spending increase.

The new September PCE report shows prices up 0.5% monthly and 3.6% annually; core prices rose 0.3% and 3.1%. Nominal spending increased 0.5%, but real spending rose less than 0.1%. Real disposable income fell 0.3%. That is less purchasing power than the headline spending number suggests.

Advance Q3 real GDP grew 3.4% annualized, up from 2.2% in Q2. The Fed held rates at 3.75–4% yesterday, with two dissenters favoring an increase. Tonight: Amazon, Sandisk and Intel earnings. Amazon’s call is scheduled for 5 p.m. ET.

Prediction markets favor a December hike.

The December Fed market showed about 77% for a quarter-point increase and 21% for no change at 8:31 a.m. ET. These are market-implied probabilities, not Fed guidance; separately traded outcomes and rounding need not sum to 100%.

Microsoft’s target rises after earnings.

KeyBanc raised Microsoft’s price target to $630 from $600, retaining Overweight, according to StreetInsider’s October 29 report. This is a target increase, not a rating upgrade.

Crypto, briefly.

Bitcoin: about $79,937. Ether: about $2,317. Coinbase USD spot snapshots at 8:33 a.m. ET. Spot levels, not daily returns.

Forwarded this email? Get Cash & Chaos in your inbox.

Independent market commentary. Company forecasts and analyst estimates are not guarantees; C&C interpretations are analysis.