Oracle Stock Is Down 60%: Why Its AI Cloud Explosion Makes It a Compelling Buy

Oracle shares have fallen more than 60% from their 52-week high to about $136, as investors weigh heavy data-center capital spending. However, Oracle’s cloud infrastructure growth nearly doubled last quarter, with cloud revenue accelerating to 47% year over year in the fiscal fourth quarter and a 93% increase in cloud infrastructure driven by AI services. The company issued $57B more debt over the last year and posted trailing-12-month free cash flow of negative $23.7B, though its debt-to-equity ratio declined from 5.3 to 3.9 and contracted remaining performance obligations totaled $638B. Valuation has improved (forward P/E down to 17) while analysts raised earnings growth estimates to 28% annually, leaving the story as a valuation-driven opportunity with financing/capex risk.

Analysis

Oracle is being priced like a capital-intensive utility with no visibility, but the mechanism is different: it is converting balance sheet capacity into a contracted AI infrastructure annuity. If management can keep utilization high, the market will eventually stop capitalizing the current FCF trough and start valuing the backlog as deferred earnings; that is a 6-18 month re-rate story, not a day-one squeeze.

The main second-order winner is not just Oracle itself but the AI hardware stack that feeds its buildout: NVDA and the broader accelerator supply chain benefit if Oracle continues to pre-commit capex to secure enterprise workloads. The competitive loser set is more subtle: hyperscalers and generic SaaS vendors lose budget share when regulated customers choose a security-isolated environment for production AI, because that shifts spend from experimentation to deployment. The risk is that this is still a financing trade as much as a product trade; if credit spreads widen or interest expense outpaces operating leverage, the equity can stay under pressure even with strong demand.

Contrarianly, the consensus may be underestimating how sticky contracted demand is, but overestimating how quickly backlog turns into free cash flow. The next 1-3 months matter for capex discipline, margin cadence, and whether spend remains matched to signed demand; the next 6-18 months matter for whether the asset base starts compounding rather than just consuming cash. Falsifiers are straightforward: cloud growth slipping materially, a capex guide that rises faster than bookings, or any sign the debt market starts demanding a higher premium for continued buildout.

Key Decisions for Investors

• Staged long ORCL on weakness: start a 25-50% position only if shares hold above the low-$130s after the open; target a 12-18 month re-rate as FCF inflects, with a tactical stop on a close below the mid-$120s or a guide that shows demand cooling.
• If you want defined risk, use a 6-12 month ORCL call spread rather than outright stock; the thesis needs time for capacity conversion, and the spread limits damage if negative FCF keeps suppressing the multiple.
• Pair trade idea: long ORCL / short a software index or broad tech multiple proxy if the next print confirms backlog conversion; this isolates the idiosyncratic balance-sheet-to-asset-base transition from sector beta.
• Add NVDA on any evidence that Oracle’s capex remains customer-funded rather than speculative; Oracle’s expansion is a cleaner read-through for accelerator demand than for ORCL equity alone, but the position should be smaller and trade as a supplier-beta expression.
• Set an alert, not a trade, if cloud growth decelerates or interest expense starts rising faster than operating income; that is the point where the bull case shifts from 'temporary cash burn' to 'structural value trap.'