The short version
- ORCL is the ticker for Oracle Corporation, the second-largest pure-play enterprise software company in the world by revenue.
- The price is no longer driven by database renewals. It is driven by the Remaining Performance Obligations (RPO) line, which crossed $130 billion at the last disclosure and prints first inside every quarterly release.
- Daily ranges run 1 to 2% on a normal session and 6 to 15% on the four annual earnings releases.
- At LHFX you trade ORCL as a CFD on MT5, leverage capped at 1:20, $3 per side commission, USD settlement, STP/ECN execution.
- You take a price exposure, not a shareholding. Quarterly dividends pass through as a balance adjustment on the ex-date.
What Oracle actually does in 2026
Founded in 1977 as a relational database vendor, Oracle now sits in three businesses that traders need to track as separate stories. The first is Oracle Cloud Infrastructure, branded OCI, a rentable GPU and CPU cloud that competes with AWS, Azure and Google Cloud. The second is cloud applications, where Fusion ERP and NetSuite carry the load. The third is the legacy on-premise database franchise, which still generates annuity-style support revenue at high margins.
What changed the equity narrative was the deliberate decision to position OCI as the cloud of choice for large training workloads. That decision sounded niche in 2022 when AWS was a hundred billion dollars ahead. By 2025 it was the loudest line on every Oracle earnings call. Big AI labs and frontier model builders sign multi-year compute commitments measured in tens of billions, and Oracle books those commitments as performance obligations. The commitments fall into RPO and unwind into revenue across the contract life.
Oracle did not catch this AI wave by accident. The company poured capex into GPU-dense regions earlier than most hyperscalers and signed multi-cloud agreements with Microsoft Azure, Google Cloud and AWS so the Oracle database could run natively inside their regions. That deal structure brought workloads back into the Oracle ecosystem instead of bleeding them away.
How the segments earn money
Investors who treat Oracle as a single revenue line miss the point. The reported segments behave differently in cyclical terms and different inputs drive each one.
Cloud Services and Licence Support is the largest reporting line at roughly three quarters of total revenue. Inside that line sits the cloud businesses (OCI plus cloud applications) and the legacy support revenue from perpetually licenced databases. Cloud applications are dominated by Fusion ERP, NetSuite and the smaller HCM and CX clouds. OCI is broken out separately in commentary even though it consolidates upward.
Cloud Licence and On-premise Licence is roughly 8% of revenue and captures perpetual licence renewals on the legacy stack. It is a runoff business in trend terms, but a high-margin one.
Hardware and Services together are roughly 17%. Hardware is the legacy Sun/Exadata engineered systems business. Services are mostly implementation consulting that wraps the database and ERP rollouts.
Gross margin profile is healthy because the licence and support lines run at software-style margins. OCI sits at a lower gross margin while the GPU buildout amortises, but the trajectory bends up as utilisation rises. Operating margin guidance reflects that mix shift, and any deviation from the guide is a single-session catalyst.
What actually moves the ORCL price
The RPO print
Remaining Performance Obligations is contracted revenue not yet recognised. At Oracle the RPO line has become the most leading indicator in mega-cap tech because it captures multi-year AI compute deals that lift the long-run revenue curve before any of that revenue flows. The headline RPO number and the year-over-year growth rate are the first two lines analysts read. Surprises of $20 billion or more on the absolute number have produced single-session moves in the 9 to 14% range.
OCI growth rate disclosure
Oracle prints an OCI growth percentage each quarter. A reading above 50% year over year keeps the AI bull case intact. A reading in the 40s on a tougher comp can be read as deceleration and brings out sellers within minutes. Management commentary on capacity constraints (the bottleneck has been GPU supply and grid power, not demand) usually sets the tone for the next two quarters.
Capex revisions
Oracle has revised its capex outlook upward several times during the AI buildout. Each revision is a referendum on confidence in the contracted backlog. The market reads upward revisions positively when accompanied by named customer wins and negatively when the only support is generic demand commentary.
Stargate and named customer milestones
The Stargate data centre programme, co-announced with OpenAI and SoftBank, plus large named deals with frontier labs, are visible deployment milestones. Each capacity announcement, site groundbreaking, or signed extension is a public catalyst with measurable price impact in the following session.
Fusion ERP and NetSuite take-rates
The cloud applications business is the slower-burn part of the story but it carries the long-run free cash flow estimate. SAP migration wins to Fusion ERP and NetSuite penetration in the mid-market both show up in the cloud-applications growth rate disclosed on the earnings call.
Earnings cadence and the off-cycle calendar
Oracle runs a fiscal year ending in May, which puts the four quarterly releases on a different calendar from most large-cap tech names.
Fiscal Q1 prints in mid-September, after the regular session closes. This is usually the most volatile release of the year because the RPO growth pattern at Oracle has been back-half loaded, meaning the first read of the new fiscal year sets the year-long narrative. The September 2024 print produced a same-day move of more than 13% as RPO came in above expectations.
Fiscal Q2 prints in mid-December, also after hours. The December release often gets attention from macro desks because it is the last meaningful AI-infrastructure tape in the calendar year, and Oracle's RPO line is read as a tell on enterprise AI commitments going into the new calendar.
Fiscal Q3 prints in mid-March and fiscal Q4 prints in mid-June. The June release includes a full-year wrap and forward guide. The off-cycle calendar means Oracle frequently sets a read-through tone for hyperscaler results that print weeks later, so positioning into and out of the ORCL release sometimes hedges across other AI infrastructure names.
Implied moves into the print typically sit between 7 and 11% from the at-the-money straddle in the options market. Actual realised moves have run higher than the implied through most of the 2024 to 2026 stretch, with the upside surprises larger than the downside.
| Fiscal quarter | Typical release window | What the market watches first |
|---|---|---|
| Q1 | Mid-September, after close | RPO growth rate, OCI growth rate, capex revision |
| Q2 | Mid-December, after close | RPO absolute number, AI customer commentary |
| Q3 | Mid-March, after close | OCI growth rate, Fusion ERP win count |
| Q4 | Mid-June, after close | Full-year RPO wrap, forward capex guide |
When ORCL trades
ORCL is a US-listed equity, so the CFD is available during NYSE regular session hours, Monday to Friday from 9:30 AM to 4:00 PM Eastern Time. There is no extended-hours quoting on the LHFX CFD, which matters because Oracle's earnings releases land after the bell. A trader holding ORCL into the print prices the move on the next session open, not in after-hours screens.
Liquidity inside the regular session is highest in the first 90 minutes and the last 60 minutes. The midday window between 11:30 AM and 2:00 PM ET typically sees tighter ranges and lighter participation, which is when many discretionary traders wait out the noise.
Economic releases and Federal Reserve decisions can push large-cap tech around regardless of company-specific news, so a calendar of the major US macro events is part of the planning before holding ORCL through scheduled risk.
CFD versus owning the share
An ORCL CFD at LHFX is a contract priced off Oracle's listed share price. The contract has no claim on the underlying equity. You take a directional exposure (long if you expect the share to rise, short if you expect it to fall) and your profit or loss is the price difference between entry and exit, multiplied by the number of shares notional, minus commission and any overnight swap.
The practical differences from buying the share through a brokerage account fall into five buckets.
First, leverage. The CFD allows leverage up to 1:20. Buying the share outright is a cash purchase unless your broker offers Reg T margin, and that margin tops out at 2:1 for retail accounts.
Second, shorting. Shorting an ORCL CFD is as simple as opening a sell order. Shorting the underlying share requires a locate, borrow cost and a margin account that supports short sales.
Third, ownership rights. A CFD position carries no voting rights at the annual meeting and confers no shareholder benefits. Buying the share makes you a registered shareholder.
Fourth, dividends. Oracle pays a quarterly cash dividend. The CFD passes the dividend value through as a balance adjustment on the ex-date, credited to long positions and debited from short positions, so the cash effect is similar but the mechanics are different.
Fifth, cost structure. The CFD carries a $3 per side commission and an overnight swap on positions held past the daily rollover. The underlying share trade carries a brokerage commission and no overnight finance, since you have paid in full.
ORCL CFD versus other ways to take Oracle exposure
Four common ways to express a view on Oracle, with the trade-offs each one carries.
| Vehicle | Leverage | Short selling | Holding cost | Best for |
|---|---|---|---|---|
| ORCL CFD at LHFX | Up to 1:20 | Open a sell order, no locate | $3 per side commission, overnight swap on rollover | Active directional trading, earnings setups, intraday and multi-day swings |
| Direct Oracle share purchase | Up to 2:1 on Reg T margin | Requires locate, borrow fee, dedicated margin account | Brokerage commission, no overnight finance | Buy and hold, dividend collection, shareholder rights |
| US-listed options on ORCL | Defined by strike and premium | Buy puts, sell calls, or build spreads | Premium decay (theta), no overnight finance | Defined-risk earnings bets, hedging, income strategies |
| Sector ETF holding ORCL (e.g. IGV, XLK) | 1:1 in cash, 2:1 on margin | Short the ETF, requires locate | Expense ratio, brokerage commission | Diversified software-sector exposure where ORCL is one of many holdings |
The CFD is the cleanest single-name expression for traders who want short selling without a borrow and leverage above what Reg T allows. Options are more capital-efficient for defined-risk earnings setups. A sector ETF dilutes the Oracle-specific signal and is a different trade.
Trading ORCL at LHFX
The specs
MetaTrader 5
STP/ECN
1:20
$3 per side
USD
Monday to Friday, 9:30 AM to 4:00 PM ET
CFD on US single stock
Cash adjustment on ex-date (credit long, debit short)
Orders route through STP/ECN execution on MT5. There is no dealing-desk intervention on ORCL flow. Spreads are raw and the only mark-up is the $3 per side commission booked separately in the deal record.
Worked example with different numbers
Assume a $5,000 account, ORCL trading at $140 ahead of an early-March session, and a thesis that the upcoming fiscal Q3 print will surprise on the RPO growth rate. You decide to take a small directional exposure.
- Open MT5, drop ORCL onto the chart from Market Watch and check the current ask.
- You size the position at 0.2 lots, which on a US single-stock CFD is 20 shares notional. At $140 that is $2,800 of exposure.
- The 1:20 leverage cap means margin used is roughly $140, or about 2.8% of the account. The remaining free margin absorbs adverse moves.
- You enter a buy order at $140 and place a stop at $133.42, a 4.7% adverse move. If hit, the loss is 20 shares times $6.58 plus $1.20 round-turn commission ($3 per side per standard lot), which is $132.80, or roughly 2.7% of the $5,000 account.
- Your earnings target is a move to $160 inside two weeks of the print, where the position would close for a profit of 20 shares times $20 minus $1.20 commission, or $398.80, about 8% of the account.
The point of running the maths before entry is to convert position size into a percentage of account at risk, not to predict the print. A 1:20 cap is a ceiling, not a recommendation. Most experienced single-stock CFD traders run effective leverage closer to 1:4 or 1:5 on ORCL into a scheduled release.
What can go wrong
Demand pacing on AI training compute
Oracle's recent quarters embed the assumption that frontier-model labs, sovereign AI programmes and large enterprises keep spending on training compute at the current trajectory. Any deceleration shows up first in the RPO growth rate, then in OCI revenue, then in the share multiple. Position sizing into a print should account for the asymmetric risk that one weak RPO line resets the narrative.
GPU and power supply bottlenecks
Building GPU-dense regions at the pace Oracle has committed to is constrained by GPU allocation from Nvidia, grid power availability at chosen sites, planning permission for new substations and skilled construction labour. Slippage anywhere in that chain pushes recognised revenue out of the current fiscal year and into the next.
Customer concentration in the largest deals
A small number of very large customers account for a meaningful share of the RPO line. A renegotiation, scope reduction or delay from any one of them moves the disclosed backlog by a visible amount. The market reads concentration as a risk factor even when each individual deal looks solid.
Multiple compression on a tape rotation
ORCL has rerated upward as the AI infrastructure story took hold. A broad rotation out of large-cap tech and into other sectors can compress the multiple without any company-specific bad news. Beta to the Nasdaq 100 is higher than it has been historically.
Practical mitigations: keep effective leverage well below the 1:20 cap, set stops outside the noise band of the expected daily range, scale size down ahead of the September fiscal Q1 print which has produced the largest historical gaps, and read both the headline RPO and the deployment-pace commentary in the prepared remarks before sizing into the next session.