--- title: "Bet-the-company for survival: Did ORCL's AI bet pay off? ---" type: "Topics" locale: "en" url: "https://longbridge.com/en/topics/39191852.md" description: "As the most debated of the Big Four cloud players, $Oracle(ORCL.US) reported 3QFY26 results (through end-Feb) after hours. Overall, the print was mixed.On the plus side, Oracle Cloud Infrastructure (OCI) — anchored by databases and compute leasing — maintained strong growth and even accelerated. On the flip side, gross margin pressure was worse than expected, while capex and borrowings climbed faster than anticipated.Net-net, the quarter screens neutral with positives offset by negatives. The real after-hours catalyst was management raising its FY27 full-year guidance..." datetime: "2026-03-11T03:51:36.000Z" locales: - [en](https://longbridge.com/en/topics/39191852.md) - [zh-CN](https://longbridge.com/zh-CN/topics/39191852.md) - [zh-HK](https://longbridge.com/zh-HK/topics/39191852.md) author: "[Dolphin Research](https://longbridge.com/en/news/dolphin.md)" --- > Supported Languages: [简体中文](https://longbridge.com/zh-CN/topics/39191852.md) | [繁體中文](https://longbridge.com/zh-HK/topics/39191852.md) # Bet-the-company for survival: Did ORCL's AI bet pay off? --- Among the four hyperscalers, $Oracle(ORCL.US) (ORCL) remains the most debated. Post-market, the company reported 3QFY26 (ended Feb) results. The print was mixed: database and compute capacity leasing under OCI stayed strong and re-accelerated, while GPM pressure was worse than expected, and both Capex and borrowings ramped faster than anticipated. Net-net, the quarter was broadly neutral as positives offset negatives. The post-market pop was driven by an FY27 revenue guide hike to 90bn from the prior 85bn. Details follow: **1) Core biz – OCI re-accelerated as expected:** IaaS OCI revenue reached 4.89bn, up 81% YoY in cc, versus 66% last quarter, a clear acceleration. The YoY net revenue add also jumped 36% QoQ to a record high. The momentum is strengthening, which is a constructive signal. However, given elevated expectations, this was in line rather than a beat. **2) AI mix weighed on margins more than expected:** As a corollary of OCI acceleration, Cloud + Software blended GPM (not disclosed separately) was ~68% this quarter, down 800bps YoY and worse than last quarter’s -590bps. This missed the Street. Assuming Software GPM held steady, Cloud GPM likely fell below 50%, with OCI GPM probably around the 30% area. **3) Leading indicator takes a ‘half-time’ breather:** RPO, a proxy for new order intake, rose by ~30bn QoQ this quarter. After a blowout in Q1, incremental orders have trended lower for two straight quarters, with fewer big surprises. That said, the absolute add remains sizable, still roughly 6x this quarter’s OCI revenue. The level remains healthy. **4) Capex accelerated higher:** A prerequisite for OCI growth, Capex hit 18.6bn this quarter, vs. 12bn last quarter, with another record QoQ net increase. The pace continues to climb. Capex exceeded total revenue for the first time. FCF has been negative for two consecutive quarters by roughly 10bn per quarter, implying heavy reliance on external financing under such cash burn. Management still guides FY26 Capex at 50bn, implying a notable step-down next quarter to ~12bn. 5) Of the previously announced 45–50bn financing plan, ~30bn was executed this quarter, including ~25bn of new debt and ~5bn of preferreds. Net interest-bearing debt climbed to ~95bn, with net debt-to-assets at 39%. Correspondingly, interest expense reached 1.18bn, up 32% YoY, with interest as a share of revenue up ~30bps QoQ. **6) Other segments remain muted:** Outside OCI, Cloud SaaS grew 11% YoY in cc, flat QoQ. Traditional Software and Hardware, stripping out FX tailwinds, were roughly flat to slightly negative YoY, showing no real inflection. Only Consulting improved meaningfully, benefiting from OCI-led deployment demand, with growth accelerating and surpassing 10% this quarter. The contribution is still small, so the overall impact is limited. **7) Overall results** **a.** Driven by OCI, total revenue growth broke the 20% threshold, a symbolic marker for the group. On a cc basis, revenue growth was 18%, in line with the prior high end of guidance, not a beat. **b.** Despite better GPMs in Software/Hardware, group GPM fell ~570bps YoY to 64.6% due to OCI mix, below the Street’s 67%. Margin pressure was heavier than expected. **c.** The three core Opex lines rose only 2.2% YoY, well below the Street’s ~6%. This suggests OCI growth and investment had limited impact on operating expenses. With tight cost control offsetting GPM pressure, Adj. OPM fell just 90bps YoY and beat by ~30bps. Adj. OP reached 7.38bn, up 19% YoY, slightly ahead of expectations. **8) Outlook — near-term inline, mid-term raised:** For 4QFY26, at the high end of cc guidance, total revenue is expected to grow 20% YoY, Cloud (IaaS + SaaS) 48%, and Adj. diluted EPS 17%. Trend-wise, the company guides continued acceleration in both total and Cloud revenue next quarter, with Adj. profit growth roughly stable. From a expectations perspective, the high end is essentially where the Street is. **Dolphin Research view:** 1) Before grading this quarter, we first frame ORCL’s current narrative and logic to pinpoint what matters in the print. Ex-OCI (IaaS), the legacy businesses have little structural growth, hovering around flat to low single-digit at best, and carry AI disruption risk similar to other software vendors. Therefore, OCI is the only segment that truly matters. Focus on the pace of OCI revenue acceleration and RPO as a read-through for new demand. While one quarter does not validate or invalidate the long-term guide, it does show near-term execution and capacity ramp. Beyond growth, the market’s two core concerns are: first, AI-heavy OCI revenue carries structurally lower margins given expensive GPUs and elevated Capex. Hence, track Cloud & Software blended GPM to back into OCI GPM levels and trend. Second, funding capacity matters as Capex is largely debt/equity funded. Key questions are whether funding is adequate, and how new debt/equity affects the balance sheet and P&L. **2) Against these priorities, this quarter shows:** a) OCI growth and implied next-quarter trajectory are broadly on track, accelerating within expectations. RPO adds were solid but not spectacular. b) OCI-driven margin drag is real and worse than feared. Pressure intensified versus last quarter. c) The worst-case funding risk has eased: the 45–50bn plan and 30bn already raised indicate credit markets and sentiment are not that tight. d) Still, elevated Capex and leverage mean lower GPM and a fast-rising interest burden as a share of revenue and profit. The adverse balance sheet and P&L effects are tangible. Overall, this print does not change the bull/bear narratives. The upside case of OCI growth and the downside case of margin erosion and rising leverage both showed up. The most important change is the FY27 revenue guide lift to 90bn from 85bn, implying total revenue growth accelerates to ~34%. This signals faster-than-expected capacity ramp, improving visibility on the long-term FY30 targets. **3) Valuation:** With growth and debate concentrated in IaaS, especially OCI AI, we consider two angles. First, strip out any impact from OCI AI (incremental revenue, margin drag, Capex, and debt/interest), and value the legacy stack (Software + Hardware + Services + Cloud SaaS + non-AI OCI) as the floor. Second, back-solve the market’s implicit valuation for OCI AI from the current market cap and compare with prior company guidance. **a) Legacy base:** SaaS + Software + Hardware + Services should stay in low single-digit growth (<3%). Even without AI, OCI grew ~50% in FY22–FY24, so non-AI OCI could compound ~35% through FY26–FY27. Together, total revenue could reach ~68bn in FY27. Assuming GPM eases slightly to 70% ex-AI and Opex grows low single-digit, FY27 net income excluding OCI AI would be a bit over 20bn. At 15x PE, EV would be ~301bn. Even ignoring AI-driven incremental debt, net interest-bearing debt was ~80bn by FY25-end, implying non-AI equity value of roughly 220bn. **b) Implied value for OCI AI:** With pre-earnings market cap at ~435.5bn, the market is implicitly valuing OCI AI, largely tied to OpenAI/Stargate, at ~210bn. The company guides FY30 OCI revenue of ~166bn, with ~140bn from AI, and steady-state OCI AI GPM at ~35%. Given limited marginal Opex for AI (few mega customers and high operating leverage, consistent with this quarter’s limited Opex sensitivity), assume marginal Opex at 15%. That yields OCI AI OP of ~28bn; net of ~6.5bn annual interest and tax, OCI AI net income is ~16bn. Discounted to FY27, the implied 210bn value equates to ~17x PE. Two takeaways emerge. First, if AI is disproved (OpenAI fails to fulfill, or the AI unit economics break and margins stay too low), legacy value alone cannot cover current market cap once higher debt and asset intensity are factored in, leaving downside risk. Second, if AI delivers as guided on scale and margins, the current implied multiples for both legacy and AI are undemanding at ~15–17x. Even without changing the long-term guide, any narrative or confidence improvement, or clearer delivery visibility (as this quarter suggests), could drive a meaningful rerating and rebound. **Key exhibits and business overview below:** **I. ORCL biz & revenue mix overview** Founded in the 1980s, ORCL’s historical core was databases and software (license model). With cloud transition and AI tailwinds, Cloud has taken the lead and is now the most important and watched segment. Since FY26, reporting has been realigned into four clearer buckets: Cloud, Software, Hardware, and Services. The disclosure is now cleaner by segment. In more detail: **a) Cloud:** Split into IaaS (OCI) and SaaS (OCA). OCA includes SaaS ERP/CRM and vertical tools, while OCI centers on database services and compute leasing. OCA used to dominate Cloud, but OCI’s rapid growth over the past 1–2 years has overtaken it. The mix is shifting toward OCI. **b) Software:** The on-prem software stack customers deploy and manage themselves, historically ORCL’s largest revenue contributor, now behind Cloud. It includes one-off Software License and recurring Software Support. **c) Hardware:** Similar to Software, it includes one-off server and related sales plus maintenance/support, and is the smallest segment by revenue. **d) Services:** Consulting and bespoke services outside the above, accounting for high single-digit percent of revenue in recent years. **II. Core focus: OCI performance** **1) OCI acceleration continued as expected** OCI IaaS revenue reached 4.89bn; in cc, growth was 81% YoY vs. 66% last quarter, a clear step-up. The YoY net revenue add was ~2.24bn, up nearly 36% QoQ. Momentum is plainly building, a positive sign. But with Street growth expectations also at 81%, there was no surprise. SaaS under Cloud grew 13% YoY this quarter, with apparent QoQ uptick entirely from FX tailwinds. In cc, growth was flat, and absolute growth remained modest. **2) AI mix dragged margins more than anticipated** With OCI scale-up, margin drag intensified. Cloud + Software blended GPM was ~68%, down 800bps YoY, worse than last quarter’s -590bps, and below expectations. Assuming Software GPM flat, Cloud GPM likely slipped below 50%, while sell-side work suggests OCI GPM near ~30%. Faster OCI growth amplifies the margin headwind, a concern reinforced by this print. **3) Leading indicator — RPO in a mid-game pause?** While revenue reflects capacity coming online, RPO as a demand indicator reached 553bn this quarter, up ~30bn QoQ. After the massive Q1 intake, adds have slowed for two straight quarters, looking quieter. Still, the absolute add remains large, at ~6x this quarter’s OCI revenue, roughly one and a half years of current OCI revenue in a single quarter. Management did not name customers, but the orders are again large AI deals. Importantly, they emphasized no additional financing or Capex is needed for these orders. Either customers already procured GPUs, or prepayments cover GPU purchases. **III. Capex and leverage ramping fast** As OCI revenue scales and capacity ramps, Capex and funding needs are rising quickly. Capex was 18.6bn this quarter (vs. a bit over 17bn in revenue), versus 12bn last quarter, with a record QoQ net increase. Capex exceeded revenue for the first time, a rare sight for a legacy leader with substantial non-cloud revenue, even if peers like CoreWeave show similar patterns. With operating cash flow up only ~20% YoY to ~7.1bn and Capex surging, FCF has been around -10bn for two straight quarters. To fund elevated Capex, management disclosed ~30bn has been raised toward the 45–50bn plan this quarter, including ~25bn of new bonds and ~5bn of preferreds. Net interest-bearing debt rose from ~81bn at FY25-end to ~95bn now, with net debt/assets at 39%. As a trade-off, interest expense reached 1.18bn, up 32% YoY, with interest as a percentage of revenue up ~30bps QoQ. Post-raise, cash stands at ~38bn (with ~10bn typically reserved for operations), enough for at most two quarters at the current burn. Further financing is likely (about ~25bn of equity issuance under the 50bn plan remains). Liquidity actions are set to continue. **IV. Other segments remain largely subdued** Beyond Cloud, growth in the other three segments remains tepid. Consulting keeps accelerating and topped 10%, aided by Cloud deployment demand. However, Software Support and Hardware sales, ex-FX, are still roughly flat YoY. No clear recovery is visible yet. **V. Overall performance** **1) Total revenue growth \>20%** Across the four segments, legacy remained sluggish, but OCI’s surge pushed group revenue growth above 20%, a notable milestone for several top-tier sell-side desks. In cc, growth was 18%, matching the high end of prior guidance, not a beat. **2) Group GPM under pressure** While Hardware and Services GPMs improved YoY, their weight is small. OCI’s drag pulled group GPM down ~570bps YoY to 64.6%, well below the Street’s 67%, underscoring OCI’s impact. **3) Expense discipline intact** The three core operating expense lines rose just 2.2% YoY, far below the ~6% expected, showing limited spillover from OCI growth into Opex and continued cost discipline. Non-operating items (e.g., amortization, restructuring) also declined. Total operating + non-operating expenses rose just ~1% YoY. With strong cost control, Adj. OPM fell only 90bps YoY and beat by ~30bps; Adj. OP was 7.38bn, slightly ahead of the Street. **Past Dolphin Research on ORCL:** **Deep dive:** Jan 22, 2026 initial coverage: [**Picking leftovers, courting whales — Is ORCL betting the house to survive?**](https://longbridge.com/zh-CN/topics/38018673) **Risk disclosure and statement:** [**Dolphin Research disclaimer and general disclosure**](https://support.longbridge.global/topics/misc/dolphin-disclaimer) ### Related Stocks - [Oracle Corporation (ORCL.US)](https://longbridge.com/en/quote/ORCL.US.md) - [ORACLE CORP DEPOSITARY SH REP 1/2000TH PFD SER D (ORCL-D.US)](https://longbridge.com/en/quote/ORCL-D.US.md) ## Comments (2) - **一缕阳光 · 2026-03-11T09:52:49.000Z**: The Oracle earnings report analysis is really out, awesome! - **Dolphin Research** (2026-03-12T01:42:46.000Z): Just got covered this year ✌️