Our top 10 things to watch in the stock market Thursday

U.S. stock futures rose as the S&P 500 looked set to recover its post-Fed decline, supported by lower Treasury yields and oil prices. Generac surged more than 30% premarket after securing an Amazon data-center backup-power agreement worth up to $8B, alongside Amazon warrants for up to $340M of Generac shares. Corporate updates were mixed: Boeing indicated 737 MAX production stabilization and free-cash-flow progress are lagging expectations, while Lennar missed on EPS ($1.19 vs. $1.28 expected) and revenue ($8.05B vs. $8.23B).

Analysis

GNRC’s re-rating hinges less on the headline contract value than on conversion timing, equipment scope, and working-capital burden. Data-center power projects can be low-margin if they require turnkey integration, inventory commitments, or performance guarantees; the Amazon warrant structure also creates dilution and gives AMZN an incentive to concentrate purchasing only after milestones are met. The more durable read-through is for electrical infrastructure and power-quality suppliers—ETN, VRT, PWR and CARR—where grid interconnection constraints and redundancy requirements support multi-year pricing power, rather than for pure backup-generation exposure alone.

BA’s cash-flow reset matters more than the production-rate commentary. A lower probability of reaching the upper half of free-cash-flow expectations extends deleveraging and delays the equity case from operational recovery to balance-sheet repair; suppliers with high 737 exposure, notably SPR, face a second-order risk if rate stabilization—not announced targets—becomes the binding constraint. Over the next 1-3 months, aircraft delivery cadence and inventory monetization are more relevant catalysts than a modest rebound in BA shares on a risk-on tape.

Cybersecurity is entering a more discriminating phase: AI-driven demand can remain real while valuation support weakens if platform consolidation reduces seat growth and customers rationalize vendors. PANW is relatively insulated through platform cross-sell, whereas OKTA and S have greater execution sensitivity to net retention and billings growth; a sector-wide multiple compression would therefore disproportionately hit the smaller, less profitable vendors. Separately, LEN’s miss should not be extrapolated into a broad housing collapse unless incentives rise materially: builders can protect volume via buydowns, but that shifts the issue from demand to gross-margin erosion and favors asset-light housing exposure over homebuilder beta.

The contrarian opportunity is MCD: investor-day commentary on traffic, value architecture and franchisee economics could catalyze a rerating if same-store-sales stabilization is credible. However, a promotional response that restores traffic only by sacrificing restaurant-level margins would validate the bear case and cap upside despite the stock’s prior drawdown.

Key Decisions for Investors

• Do not chase GNRC after the opening dislocation. Establish a 1-3 month watch for disclosed contract milestones, expected annual revenue cadence, gross-margin profile and warrant dilution; initiate only if the stock retraces materially while management confirms data-center revenue is accretive to consolidated margin. A failure to quantify these items makes the move vulnerable to reversal.
• Express the data-center power theme through a basket long ETN/VRT/PWR rather than incremental GNRC exposure over 6-18 months. These names capture electrical distribution, cooling and interconnection spend across multiple hyperscalers; reassess if hyperscaler capex guidance weakens or lead times normalize sharply.
• Maintain an underweight or short-biased BA versus long Airbus proxy EADSY over the next 3-6 months. The thesis is a widening delivery/cash-conversion gap; cover if BA demonstrates sustained monthly production and deliveries consistent with its cash-flow midpoint, rather than merely reiterating targets.
• Pair long PANW / short S for 3-6 months, sized modestly around earnings. The trade isolates platform consolidation against a higher-duration endpoint vendor; stop out on a material SentinelOne billings acceleration or PANW evidence of renewed large-deal discounting.
• Use MCD’s investor day as a catalyst watch rather than buy immediately. Add on evidence that U.S. traffic stabilizes without a meaningful decline in franchisee margins; invalidate the thesis if guidance implies promotion-led sales recovery or another downward revision to same-store-sales expectations.