Piper Sandler reiterates Oracle stock rating on AI capacity outlook

Piper Sandler reiterated an Overweight rating on Oracle (ORCL) with a $225 target and estimates Oracle could add ~$2.2B of OCI revenue in FY2027, based on ~2,400MW of capacity coming online. The note highlights continued AI-related demand (beneficiary of generative AI in IaaS/SaaS) but flags risks around AI infrastructure capital needs, monetization, customer concentration, and margins; shares at $140.27 are down 27% over six months near the 52-week low. The stock is viewed as undervalued with a PEG of 0.69, while other commentary cites strong visibility via a $77B remaining performance obligation (86% of NTM revenue) and AI backlog growth.

Analysis

The market is still pricing Oracle like a capex-heavy story, not a monetization story. That gap matters: if incremental cloud capacity is meaningfully pre-sold, the stock can rerate on operating leverage because the next dollar of revenue is more valuable than the next dollar of build-out. If utilization lags, though, the same pipeline becomes a depreciation and funding drag, and the equity stays trapped in a low-multiple utility bucket.

The most interesting second-order winners are not the obvious AI software names; they are the power, cooling, and electrical infrastructure vendors that monetize the physical bottlenecks behind data-center expansion. On the competitive side, Oracle’s edge is less about taking share from the hyperscalers and more about capturing enterprise workloads that are sensitive to price/performance and integration with existing software stacks. That makes the bull case more durable in the applications business than in pure infrastructure, where pricing can stay rationally competitive.

The contrarian risk is that consensus may be over-indexing on backlog visibility while underweighting free-cash-flow conversion. In the next 1-3 months, the stock likely trades on whether management sounds more disciplined on spend and whether the market believes the capacity can be filled without another round of margin disappointment. Over 6-18 months, sustained utilization and software attach would force a multiple expansion; absent that, the current valuation could still be a value trap despite seeming cheap on PEG. Falsifier: rising capex guidance without a clear step-up in OCI revenue, or another quarter where revenue visibility grows faster than cash generation.

Key Decisions for Investors

• Tactically buy ORCL on weakness as a 3-6 month mean-reversion trade, but keep sizing modest; upside is a rerate if OCI utilization inflects, while downside is another margin/capex reset. Use a close below the post-earnings swing low as the risk stop.
• Preferred expression: ORCL Jan-2027 call spread rather than outright calls to capture a slower re-rating path; this limits premium burn if the market waits for proof of utilization and FCF conversion.
• Long VRT or ETN as a second-order beneficiary basket against ORCL exposure; the trade works if Oracle’s build-out is part of a broader data-center power cycle, and it is less dependent on Oracle-specific monetization timing.
• Avoid chasing the name above near-term resistance until management proves capacity utilization and gross margin stability; if the next print still shows capex outrunning revenue conversion, the bullish thesis should be downgraded.
• If you want a pair, consider long ORCL / short a higher-multiple AI software basket such as SNOW on the view that integrated enterprise AI is being rewarded over standalone tooling; exit if Oracle’s growth story remains mostly backlog and not cash.