Earnings Season Surprises

Tesla beat adjusted EPS at $0.41 and topped expectations on vehicle deliveries and revenue, while signaling higher 2026 capex of $25B from $20B previously. Commentary was constructive on Tesla's AI/robotics push, including Terafab, Optimus factory plans, and lower-cost Model 3/Model Y versions, though energy storage deliveries slightly missed. Elsewhere, IBM and GE Vernova were described as solid AI-linked earnings stories, Texas Instruments posted 19% revenue growth with data-center revenue up 90%, and Progressive remains under pressure amid cyclical insurance margin concerns.

Analysis

The common thread is not “AI beneficiaries” in the abstract, but a tightening of the bottlenecks around power, grid gear, and control electronics. GEV and TXN look better because the spend cycle is migrating from speculative model-buildout to concrete electrical infrastructure with multi-year backlogs and pricing resilience; that tends to compress downside because cancellation risk is low once capacity is booked. The second-order winner is anyone selling upstream components into data centers and electrification, while the loser is the market’s assumption that AI capex is purely a hyperscaler/software story.

Tesla’s most interesting signal is not the headline-grabbing robotics roadmap, but the mismatch between ambition and current capital deployment. If management is telegraphing a much larger spend step-up later this year while current capex remains subdued, the stock is vulnerable to a “show me” period where narrative outruns execution. The hidden risk is supply-chain and construction friction: domestic buildouts for chips, robots, and power hardware all compete for the same contractors, tools, and permitting bandwidth, so delays can cluster across seemingly unrelated initiatives.

Progressive looks less like a stock-specific problem and more like a classic late-cycle underwriting trade. The market is discounting that peak margins in property/casualty usually mean mean reversion in 12-24 months, not immediately, so the setup is better for patience than urgency. Contrarian angle: the move may be overdone if investors are extrapolating near-term profitability decay without respecting Progressive’s structural share gains and pricing tech advantage, but the catalyst to turn it higher is not valuation — it is evidence that loss trends stay benign into the next renewal cycle.

Key Decisions for Investors

• Long GEV on pullbacks over the next 2-6 weeks; the setup is best as a multi-quarter backlog/visibility trade, with downside protected by scarcity value in grid and gas-turbine capacity. Risk/reward favors owning rather than chasing after the recent rerating, because book-out visibility should support estimates through 2026-2027.
• Maintain a tactical long TXN versus a basket of more cyclical industrial semi names for the next 1-2 quarters; the thesis is pricing resilience plus data-center pull-through, not broad-chip beta. Use any industry-wide AI digestion selloff to add, as TXN’s demand signal is more durable than the average semiconductor cycle.
• Avoid fresh TSLA adds ahead of the next capex/production milestone; consider selling downside puts only after confirmation of spending acceleration or production evidence. The trade-off is upside optionality on robots/Terafab, but the near-term risk is narrative compression if execution lags the announced timeline.
• For PGR, prefer a staged long over 6-12 months rather than a chase here; use weakness if the stock continues to price in peak earnings. The risk/reward improves if loss-cost trends stabilize, but the next 2-3 quarters likely remain a valuation swamp rather than a catalyst-rich window.
• Pair long GEV / short a broad market industrial ETF for 1-3 months if you want cleaner AI-infrastructure exposure; GEV has direct backlog monetization, while the short leg hedges macro and valuation beta. This isolates the power-grid capex winner from the rest of the industrial complex.