Tesla tops Q1 EPS estimates as margins hit five-quarter high — but flags $25B+ capex year

This quarter's beat came not from selling more, but from earning more per sale. Revenue of $22.4B actually fell slightly short of the LSEG consensus ($22.64B), yet adjusted EPS of $0.41 topped the $0.37 estimate by 10.8%. The bridge was gross margin: 21.1%, a five-quarter high. Operating income surged 91% year over year to $0.94B — leverage nearly six times the pace of revenue growth (+16%).

GAAP net income of just $0.5B reflects a mark-to-market loss on Tesla's bitcoin holdings, an accounting effect rather than operational weakness. To read the quarter's true direction, operating income and adjusted EPS are the better guides.

The 358,023 deliveries are not the headline. What matters is automotive gross margin ex-credits improving 1.3 points sequentially to 19.2% from 17.9% — volume was earned while holding price, not bought with discounts. Deliveries in France and Germany jumped more than 150% quarter over quarter, reviving Tesla's weakest region, and the order backlog ended the quarter at its highest level in over two years.

The quiet grower is services and other: revenue of $3.75B, up 42%, far outpacing the core car business, driven by paid FSD subscriptions reaching roughly 1.3 million. The shift from selling a car once to earning recurring revenue from cars already sold is now showing up in the P&L.

Energy storage deployments of 8.8 GWh fell 38% from Q4's record 14.2 GWh, and segment revenue declined 12% year over year to $2.41B. Lumpiness is structural in a project-driven business, but the 'steady profit pillar' narrative took a dent.

Even so, storage gross margin exceeded 39.5%. That includes roughly $250M of one-time tariff benefits that may not repeat, but the underlying margin power — about double automotive — was confirmed. Whether Tesla delivers on its maintained guidance that 2026 full-year deployments exceed 2025 is the credibility test for this business.

The flip from +4% after hours to -3.56% ($373.60) the next day traces to two sentences on the call, not the results: 2026 capex of more than $25B (raised from around $20B), and negative free cash flow guidance for the remaining three quarters. The contrast stung precisely because Q1 FCF was a healthy positive $1.44B (+117%).

Where the money goes is clear — six factories moving toward operation, AI training compute, Cybercab and Semi ramps, the Fremont Optimus line, and a semiconductor research fab in Austin. In effect, Tesla declared it will burn today's automotive profits as advance investment in AI and robotics. The further out the payback, the more the stock can decouple from earnings — and this selloff was the first taste.

The AI segment carried progress and hedges in the same breath. Progress: unsupervised Robotaxi expanded to three cities — Austin, Dallas and Houston — on V14.3, Cybercab and Semi production began at Giga Texas, and FSD (Supervised) v4 won Dutch approval, opening the door to Europe. Hedges: management drew its own line — 'Robotaxi revenue will not be material in 2026' — and Elon Musk called Optimus 'probably the biggest product ever' while conceding this year's production rate is 'impossible to predict.'

The net read: this quarter marks the AI story's shift from demo to production-readiness, while revenue contribution remains a 2027-and-beyond promise. The $25B capex underwrites that promise, and the recovered automotive margin funds the capex — confirming how tightly the three are interlocked was the call's real takeaway.