RBC Capital maintains Oracle stock rating on mixed results

Oracle reported fiscal Q4 revenue growth of 21% year over year and EPS of $2.11 versus $1.96 consensus, though cloud revenue at 47% came in slightly below expectations. Remaining performance obligations rose to $638 billion, and management said AI-related bring-your-own-hardware contracts total $75 billion while the company plans to raise $20 billion in incremental capital. Analyst views were mixed, with several firms raising targets while RBC kept a Sector Perform rating and a $190 target.

Analysis

Oracle’s reaction looks less like a fundamental miss and more like a financing/optics reset after the market had already capitalized AI scarcity and margin expansion into the stock. The key second-order issue is that the business is shifting from software-like cash conversion toward infrastructure-like capital intensity, which compresses the quality of earnings multiple even if revenue growth stays high. That matters because a higher capex/financing burden can make each incremental dollar of AI backlog worth less than investors initially assumed.

The most important competitive implication is that Oracle is proving demand, but not yet proving that it can fund demand efficiently. In AI infrastructure, the winners over the next 6-12 months are likely to be vendors with the best access to cheap financing or the strongest balance sheets, while leveraged capacity builders risk multiple compression even on strong bookings. That creates a subtle headwind for adjacent high-beta AI infrastructure names if investors start discounting backlog by required capital rather than headline growth.

Near term, the stock is vulnerable to a “great quarter, worse model” trade: beats can fail to hold if the Street focuses on cash burn, dilution risk, and the absence of a clear revenue bridge for the added capital. A reversal would likely need either a visible acceleration in monetization of the AI backlog or evidence that bring-your-own-hardware meaningfully de-risks Oracle’s funding gap faster than expected. Over months, the debate is less about whether demand exists and more about whether Oracle can convert that demand into FCF at a premium multiple without repeatedly resetting capital needs.

Contrarianly, the pullback may be overdone if investors are underestimating how quickly operating leverage can reassert itself once this capex cycle rolls into revenue. The market is pricing Oracle more like a highly levered infrastructure story than a recurring software compounder, which may create room for a re-rating if management keeps execution clean for two quarters. The setup is asymmetric: weak sentiment now, but meaningful upside if the next update shows backlog monetization outpacing financing pressure.

Key Decisions for Investors

• Trade idea: buy ORCL on further post-earnings weakness only if it holds above the prior breakdown zone for 3-5 sessions; target a 10-15% rebound over 1-3 months if the market pivots back to backlog quality.
• Pair trade: long ORCL / short a basket of higher-duration AI infrastructure names that are more dependent on external capital markets; use this to isolate execution versus financing-risk dispersion over the next 1-2 quarters.
• Avoid chasing strength in ORCL until management provides a clearer cash conversion path; if capex guidance is revised higher again, expect another 8-12% de-rating risk in the following 2-4 weeks.
• For long-only accounts, consider a call spread in ORCL rather than outright stock to express upside from sentiment normalization while capping downside if the market re-prices the stock as a financing-heavy AI platform.
• Set a catalyst watch for the next quarterly update: if backlog growth remains strong but incremental capital needs stop rising, that is the signal to add aggressively; if not, fade rallies into analyst target upgrades.