Trade Ideas September 30, 2026 08:30 AM

Buy Oracle on the Project Jupiter Sell-Off: AI Contracts Give You Asymmetric Upside

Core cloud momentum + record contract backlog, but heavy capex and execution headlines have driven fear into the tape — we buy the dip.

By Derek Hwang
Share
Twitter Reddit Facebook LinkedIn
ORCL

Oracle is trading at $137.90 with a ~$417B market cap after headlines around Project Jupiter delays and massive near-term capex. The company booked $30B+ in new AI cloud contracts, RPO of $664B and Cloud Infrastructure revenue up 121% to $7.4B. Negative free cash flow and heavy capex are real, but valuation metrics (EV/EBITDA ~14.3, P/E ~22) still leave room for upside if Oracle converts backlog into revenue. Trade plan: long at $137.90, target $190.00, stop $118.00, horizon long term (180 trading days).

Buy Oracle on the Project Jupiter Sell-Off: AI Contracts Give You Asymmetric Upside
ORCL
Summarize with
ChatGPT Perplexity Claude Grok Gemini

Key Points

  • Oracle has a record remaining performance obligation (RPO) near $664B, including $30B+ in new AI cloud contracts.
  • Cloud Infrastructure revenue accelerated 121% to $7.4B, with AI infrastructure utilization near 97.9%.
  • Near-term free cash flow is negative (-$28.72B trailing 12 months) due to heavy capex; management expects $90-95B CapEx for FY27.
  • Valuation is ~22x P/E and ~14.3x EV/EBITDA, which prices in some conversion risk but leaves room for upside if backlog converts to revenue and margins expand.

Hook / thesis

Oracle is a classic case of fundamental progress colliding with headline-driven fear. Over the last few days the market has punished Oracle on reports of Project Jupiter delays and a force majeure notice tied to AI data center builds. That reaction looks excessive. The company reported record AI cloud bookings and a remaining performance obligation (RPO) that now sits near $664 billion, while Cloud Infrastructure revenue accelerated 121% to $7.4 billion. Those are the kinds of revenue engines investors pay up for. Meanwhile, negative free cash flow and heavy capex have created short-term pain. That pain is priced in; we see an asymmetric risk-reward to the upside from current levels.

Why the market should care

Oracle sells infrastructure and enterprise software to large institutions, with three operating segments: Cloud and License, Hardware, and Services. The strategic pivot to build out AI-optimized data centers and capture a slice of enterprise AI workloads is the major fundamental driver today. Large multi-year cloud contracts are sticky - they create long-duration revenue and visible backlog. Oracle's reported $30B+ in new AI cloud contracts and RPO near $664B create a very different revenue profile than the legacy on-premises refresh cycle.

Put plainly: if Oracle can convert that backlog into realized cloud revenue at scale, margins and cash flow should follow. The market already pays for that possibility in the multiples; today's sell-off is an opportunity to buy optionality at a discount.

Key numbers that matter

Metric Value
Current price $137.90
Market cap ~$417B
RPO / backlog $664B (reported new AI cloud contracts $30B+)
Cloud Infrastructure revenue growth 121% to $7.4B
Free cash flow (TTM) -$28.72B
CapEx guidance / run-rate $90-95B expected FY27; Q1 reports included very large spend (>$28B cited)
P/E ~22x
EV / EBITDA ~14.3x
Debt / Equity ~1.88
Dividend $0.50 per share (quarterly) - ex-dividend 10/09/2026, payable 10/23/2026

Why this is actionable now

The sell-off following Project Jupiter headlines is largely about execution timing and cash flow volatility, not a sudden deterioration of demand. Oracle's cloud contracts and utilization metrics (AI infra utilization near 97.9% reported) argue the demand exists. The tension the market is pricing is between todays heavy capex and a future payout that may take quarters to materialize. That creates a window: you can buy high-quality enterprise contracts and cloud momentum at a multiple that assumes some conversion risk.

Valuation framing

At roughly $417 billion market cap and an enterprise value of about $507 billion, Oracle trades at ~14.3x EV/EBITDA and ~22x P/E. Those multiples are reasonable for a company that has both established software cash flows and a fast-growing AI infrastructure business. The company is not cheap on an absolute basis, but the multiples already bake in a transition where capex suppresses free cash flow for now. If Oracle converts a fraction of the $664B backlog into recurring revenue and achieves the margin expansion associated with cloud scale, a re-rate to tech-hyperscaler multiples would imply significant upside. Conversely, if execution slips materially and capex becomes permanent drag, the market will re-price the company lower. At $137.90 the market is nearer the latter case than the former - we think that's overly pessimistic.

Catalysts to watch

  • Quarterly results showing sequential improvement in cloud gross margins and a stabilization of free cash flow - early evidence of backlog conversion.
  • Operational updates that resolve Project Jupiter timing and power/infrastructure concerns at the New Mexico campus.
  • New large-scale AI contract announcements and customer references that demonstrate multi-year revenue visibility and utilization.
  • Signs of capex efficiency - lower quarter-to-quarter absolute CapEx or improved capital intensity metrics.
  • Analyst upgrades or institutions increasing weight as cloud revenue proves durable and margins expand.

Trade plan (actionable)

We recommend a long trade at an entry price of $137.90. Set a stop loss at $118.00 to limit downside if execution deteriorates or headlines worsen. Primary target is $190.00. This trade is oriented to the long term - explicitly the horizon is long term (180 trading days) to give Oracle time to convert backlog and for capex to start showing returns.

Why these levels? Entry at $137.90 matches the current price where headline fear is most concentrated. Stop $118.00 sits well below recent volatility but above the 52-week low of $114.50, giving room for noise while protecting against a deeper structural re-rating. Target $190.00 assumes a re-rate as cloud conversion becomes visible and multiples expand modestly - this is a ~38% upside from the entry, reasonable for a 180-trading-day horizon if at least some of the backlog converts into high-margin recurring revenue.

Risks and counterarguments

  • Execution risk on AI data centers - Project Jupiter delays and force majeure notices are not trivial. If Oracle cannot supply power or complete builds on schedule, customer deployments could be delayed or canceled.
  • Cash flow strain from capex - Oracle is spending heavily. Management expects $90-95B in CapEx for FY27, and free cash flow for the trailing 12 months is negative ~$28.72B. If capex continues to outpace returns, leverage and liquidity pressure could mount.
  • Competition from hyperscalers - Microsoft, Amazon, and Google are investing enormous amounts in AI infrastructure and have entrenched cloud businesses. Oracle must defend pricing and win share in an increasingly competitive market.
  • Balance sheet and leverage - Debt-to-equity around 1.88 and elevated CapEx mean the company has less financial flexibility if execution stalls or macro conditions tighten.
  • Market sentiment and headline risk - Oracle's stock may remain volatile while headlines about power, construction, or customer delivery continue. Even positive fundamental progress may take time to translate into a sustained rerating.

Counterargument: The bears are not wrong to flag free cash flow and capex. You can make a credible case that Oracle's aggressive build-out is a value-destructive cycle - pouring capital into specialized data centers could leave the company with stranded assets if AI demand is less durable or customers choose hyperscalers. In that scenario, multiples compress and the stock trades materially lower. That is why we use a protective stop and limit position size to what we can tolerate being wrong on.

What would change my mind

I would materially reduce the bullish stance if any of the following occurs: (1) management walks back revenue visibility and the reported RPO declines materially, (2) Project Jupiter issues cascade into cancellations from marquee customers, (3) capex guidance rises materially above $95B without clear milestones to show improved capital efficiency, or (4) free cash flow does not show a path back to positive within the next two to four quarters after the bulk of near-term builds are completed. Conversely, clear sequential margin improvement in cloud infrastructure and accelerating cash receipts from AI contracts would strongly reinforce the bull case.

Bottom line

Oracle is a high-quality enterprise franchise positioned to capture enterprise AI workloads. The headline-driven sell-off provides a buying opportunity for investors willing to accept shorter-term cash flow volatility in exchange for asymmetric upside as multi-year AI contracts convert into recurring revenue. The trade is not without risk - execution and cash flow dynamics matter - but at $137.90 the risk/reward favors a tactical long with a disciplined stop at $118.00 and a long-term horizon of 180 trading days to let the thesis play out.

Trade plan recap: Buy ORCL at $137.90, stop $118.00, target $190.00, horizon long term (180 trading days). Risk level: medium.

Risks

  • Project Jupiter execution delays could push out revenue and trigger customer cancellations.
  • Sustained negative free cash flow if capex remains elevated and returns on that capital lag expectations.
  • Intense competition from hyperscalers may pressure pricing and share gains in AI cloud.
  • Leverage and balance-sheet pressure - debt/equity ~1.88 increases sensitivity to cash flow shocks.

More from Trade Ideas

Buy SpaceX After Starship's Orbital Success - A Decade-Level Growth Bet Sep 30, 2026 STMicroelectronics: Recovery Is Broader — Add on Dips, We Remain Buyers Sep 30, 2026 MongoDB: Buy the Post-Departure Bounce - Tactical Swing to $470 Sep 30, 2026 Starling Oncology: Improving Cash Flow and Gross Margin Set Up a Mid-Run Re-rating Sep 30, 2026 Palantir: An AI Platform Thesis You Can Trade — Tactical Long with Defined Risk Sep 30, 2026