Oracle Rebounds to $137.45 From $133.48 Low as $664B Backlog Offsets Project Jupiter Power Delay

Oracle shares traded at $137.45 on Thursday afternoon, down $7.11 or 4.9% from Wednesday's close of $144.56, after a session that ran from a high of $143.17 to a low of $133.48. At the bottom of the range, ORCL was down 7.66%, its steepest intraday drop since the September 10 earnings reaction. The stock has since clawed back $3.97 off that low, which shows buyers stepping in at the $133 area rather than letting the selloff run.

Volume tells the same story. By mid-afternoon, 29.35 million shares had traded against a 38.88 million daily average. That is heavy participation for a partial session, but it is not panic volume. Institutions were repricing risk, not dumping the stock wholesale.

The trigger was a force majeure notice Oracle sent to the developer of Project Jupiter, its $165 billion AI data center campus in New Mexico. The notice gives Oracle the contractual footing to defer payments if the 2.45-gigawatt facility misses its 2028 launch date. Oracle says the project remains on schedule and that it is fully committed to New Mexico. The market sold first and asked questions later.

The bigger picture puts the move in context. ORCL trades 57% below its all-time closing high of $324.63, set on September 10, 2025, and sits in the lower third of its 52-week range between $114.50 and $322.54. The stock is below its 200-day simple moving average and has been in a technical downtrend since January. Thursday's low dragged it back to early-August levels.

The stock enters this selloff with fundamentals that have never been stronger on paper. Oracle's fiscal first-quarter 2027 results, reported September 10, showed revenue up 30% to $19.3 billion, cloud infrastructure revenue up 121% to $7.4 billion and remaining performance obligations at $664 billion. Management raised full-year guidance to at least $90 billion in revenue and $8.10 in non-GAAP EPS.

This forecast rests on one thesis: Oracle's backlog is real, but its stock price is set by the cost and timing of converting that backlog into cash, and Project Jupiter just made both look worse. With $125 billion in debt, negative free cash flow of $23.7 billion last fiscal year and a 10-year Treasury yield at 5.15%, investors are discounting every delay. The $133.48 intraday low is the line that decides whether ORCL rebuilds toward the $152.94 to $159.58 post-earnings range or retests the $114.50 52-week low.

The Force Majeure Notice: What Oracle Actually Did

The notice itself is narrower than the headline suggests, and the detail matters for the forecast. Oracle sent it to Stack Infrastructure, the Blue Owl Capital unit developing Project Jupiter. The purpose is to protect Oracle from higher costs and to allow it to defer payments on the campus if the facility fails to come online as scheduled in 2028.

Oracle is not walking away. It remains the site's anchor tenant and has not sought to exit the project. The move is a contractual positioning step, staking out Oracle's legal rights early in case the delivery date slips. Force majeure clauses normally excuse a party from obligations when events outside its control disrupt a contract. Whether the maneuver would actually free Oracle from its payment obligations is uncertain.

Both sides issued measured statements. Oracle said Project Jupiter remains on its planned schedule and that it is confident in its path forward. Blue Owl said the notice does not change the financial commitments behind the multi-year project. Neither statement disputed that the notice was sent, and neither addressed the specific cause cited in it.

The market reaction reflects what the notice signals rather than what it does. Companies don't send force majeure notices on projects running smoothly. The notice is the first formal step toward shifting the cost of a delay onto someone else's balance sheet, and investors read it as confirmation that the 2028 timeline is at risk. Concern over the timeline had been building for weeks before Thursday.

For Oracle's financials, the direct impact is limited in the near term. Project Jupiter was never scheduled to generate revenue before 2028, so a delay doesn't touch fiscal 2027 guidance. The indirect impact is larger. Jupiter is part of the Stargate AI infrastructure program with OpenAI and SoftBank, and a slip in the flagship campus raises questions about the timing of revenue tied to Oracle's largest customer commitments.

The notice also exposes the structure of Oracle's buildout. Oracle leases many of its AI campuses from developers rather than owning them outright. That model keeps some capital off Oracle's balance sheet but ties its revenue timeline to third-party construction, permitting and financing. When any of those links breaks, Oracle's backlog conversion slows. Thursday's selloff is the market pricing that dependency for the first time on a specific, named project.

The Power Problem Behind Project Jupiter

The root of the Jupiter delay is power, and it is the risk Oracle can't solve with capital alone. The campus is designed to run on up to 2.45 GW of fuel cells from Bloom Energy. Those fuel cells run on natural gas, and the gas has to arrive through a 17-mile pipeline that doesn't exist yet.

That pipeline has slipped badly. Its completion target has moved from August 2026 to February 2027 after the New Mexico State Land Office repeatedly refused right-of-way permits across state trust lands. Without the pipeline, the Bloom fuel cells can't operate, and the campus has no viable power source. A data center without electricity is a stranded building.

Oracle's choice of power source made the project vulnerable. The company had originally planned a gas plant for the site, then switched to Bloom Energy fuel cells and is seeking 2 gigawatts of renewable energy capacity in New Mexico. Oracle's own FY26 results release described its datacenters as intended to use clean energy from natural gas fuel cells to generate electricity with minimal emissions. Fuel cells can be deployed faster than a conventional power plant, but they still need a fuel supply, and that supply depends on state permitting.

The seven-month pipeline delay compresses Oracle's construction window. If gas doesn't flow until February 2027, the fuel cells can't be commissioned and tested until after that. The 2028 launch date leaves limited room for further slippage. Any additional permitting setback pushes the timeline past 2028, which is exactly the scenario the force majeure notice is designed to cover.

Bloom Energy's stock shows the linkage. Bloom fell 5.4% in premarket trading and traded at $257.25 later in the morning, down 6.52%. Oracle has a direct financial tie to Bloom: its fiscal 2026 results included a one-time gain from the sale of Bloom Energy warrants. Bloom is up 229.89% over the past year at a 57x forward multiple, so its valuation depends heavily on deployments like Jupiter.

The power problem extends beyond this one campus. Every major AI data center buildout faces the same constraint: grid capacity, fuel supply and permitting take longer than chips or buildings. Oracle has bet more aggressively than its peers on self-supplied power through fuel cells, which speeds deployment when it works but adds a new layer of execution risk. Jupiter is the first test of that bet at gigawatt scale, and it is running behind.

Financing: The $18 Billion Loan Package and Blue Owl

The financing structure around Project Jupiter explains why investors reacted so sharply. A group of 20 financial institutions assembled an $18 billion lending package to fund construction of the campus. That loan sits with the development entity, not with Oracle, but its health depends on Oracle's commitment as tenant.

Blue Owl Capital owns Stack Infrastructure, the developer, and supplies the equity. Oracle is the tenant and not the borrower. That arrangement is standard for large data center projects: the developer borrows against the long-term lease, and the tenant's creditworthiness underpins the loan. When the tenant signals it may defer payments, lenders reassess the collateral.

Blue Owl's stock reflects the stress. Shares traded at $9.18 on Thursday, down 4.39%, and the stock is down 46.06% over the past year. Blue Owl's exposure to data center real estate and private credit has turned from a growth story into a liability as AI buildout timelines stretch.

The lending risk matters for Oracle's broader funding. Oracle has leaned on project finance, private credit and leasing structures to expand capacity faster than its own balance sheet would allow. If lenders on Jupiter tighten terms or demand guarantees, similar demands could follow on other Oracle-anchored projects. That would push more of the financing burden back onto Oracle's own balance sheet, which is already stretched.

Stargate adds a layer of counterparty exposure. Project Jupiter sits inside the Stargate program with OpenAI and SoftBank, and OpenAI is among Oracle's largest customers. Stargate has hit turbulence before: in 2025, OpenAI scaled back its early Stargate plans to a much smaller initial data center. Any sign that the program's timeline or scope is shifting again raises questions about the timing of revenue behind a large share of Oracle's backlog.

The practical read for shareholders is that Oracle's growth depends on an ecosystem of partners: developers, lenders, power suppliers and anchor customers. Each has its own balance sheet and its own risks. Oracle controls the software and the cloud service. It doesn't control permitting, pipeline construction or lender appetite. The force majeure notice is Oracle protecting itself against failures in that ecosystem, and investors are pricing the chance that those failures spread to other projects.

Q1 FY27: Cloud Infrastructure Revenue Rips 121%

The operating business Oracle reported two weeks ago is growing at a pace few large-cap companies can match. According to Oracle's Q1 FY27 results filed with the SEC, total quarterly revenue rose 30% to $19.3 billion, driven by the delivery of 850 megawatts of additional datacenter capacity. Cloud revenue, combining infrastructure and applications, jumped 62% to $11.6 billion.

Cloud infrastructure is the engine. Oracle Cloud Infrastructure revenue surged 121% to $7.4 billion, the ninth consecutive quarter of acceleration. That followed 93% growth in the fourth quarter of fiscal 2026. CFO Hilary Maxson said the company expects the acceleration to continue through the rest of fiscal 2027 as more backlog converts into revenue.

Cloud applications grew 10%. Software revenue fell 3% to $5.5 billion as customers keep migrating from on-premises licenses to the cloud. Services revenue rose 5% to $1.4 billion, and hardware revenue climbed 15% to $0.8 billion. The mix shift toward infrastructure is stark: cloud infrastructure accounted for 91% of constant-currency cloud revenue growth in the quarter, and the Americas contributed 91% of the growth in the cloud and software business.

Capacity delivery is the constraint Oracle is racing to relieve. Since the end of the fourth quarter, Oracle delivered more than 300,000 GPUs to its AI cloud customers, almost triple the capacity delivered in Q4. GPU utilization stood at 97.9%, which means almost every chip Oracle installs is immediately put to work. Demand is running faster than supply.

The AI software layer is gaining traction too. Customers used Oracle's embedded AI capabilities more than 150 million times during the quarter, up 42% sequentially. AI agents executed more than 3.5 million times in production, and the number of customer AI agents in production rose 90% to more than 2,300.

The quarter was Oracle's first-ever sequential revenue increase in a fiscal first quarter. That matters because Q1 is seasonally Oracle's weakest period, when software license renewals dip. Posting sequential growth in that quarter shows cloud infrastructure has become large enough to overpower the old seasonal pattern.

The problem for the stock is that none of this growth answered the question investors care about most. The market already believed demand was real. It is worried about cost, financing and delivery risk. Q1 proved demand. Project Jupiter just raised the delivery question again.

$664 Billion in RPO and the Conversion Timeline

Oracle's backlog is the largest in enterprise software history, and its size is both the bull case and the risk. Remaining performance obligations, the contracted revenue Oracle has yet to recognize, stood at $664 billion at the end of Q1 FY27. That is up $209 billion from a year earlier and up $26 billion from the fourth quarter.

The backlog kept growing in Q1. Oracle booked more than $30 billion of additional AI cloud contracts during the quarter, according to its Q1 earnings release. The company said the structure of those new contracts has no incremental impact on its plans to raise capital.

The conversion timeline is the critical variable. Oracle expects roughly half of its RPO to convert to sales over the next 36 months. That implies more than $330 billion in revenue over three years from the existing backlog alone, against fiscal 2027 revenue guidance of at least $90 billion. The new RPO added in Q1 isn't expected to affect Oracle's capital expenditures or revenue until fiscal 2028 or later.

Contract structure reduces some of the financing burden. Most of the RPO added in the third and fourth quarters of fiscal 2026 came from large AI contracts where the customer prepaid Oracle for GPUs or supplied its own hardware. The prepaid and customer-supplied hardware portions of those contracts totaled $75 billion at fiscal year-end, which substantially reduces the capital Oracle must raise. In Q1, most new contracts used prepayments, bring-your-own-hardware arrangements or similar structures.

The risk is concentration and timing. A large share of the backlog is tied to a small number of AI customers, with OpenAI prominent among them. If any of those customers slows its buildout, the conversion timeline stretches. Project Jupiter is a direct test of that risk: it is a Stargate campus, and a delay there pushes out revenue Oracle had expected to recognize starting in 2028.

The market's skepticism shows in the gap between backlog and price. Oracle's RPO grew 46% over the past year, yet the stock has fallen 29% over the same period. Investors aren't doubting that the contracts exist. They are discounting how much cash those contracts will produce, how soon, and at what cost. A backlog that converts in 2030 is worth far less at a 5.15% discount rate than one that converts in 2027.

Guidance: $90 Billion Revenue and $8.10 EPS for FY27

Oracle's outlook gives the stock a clear earnings anchor. For fiscal 2027, the company now expects total revenue of at least $90 billion, representing 34% growth in constant currency, and non-GAAP EPS of $8.10. That EPS figure represents 18% growth after adjusting for one-time gains in fiscal 2026 from selling the Ampere chip business and Bloom Energy warrants. The $8.10 target is up from the $8.05 guidance issued with fourth-quarter results.

The second-quarter guide shows the acceleration continuing. Oracle expects Q2 fiscal 2027 total revenue to grow 30% to 34% and cloud revenue to grow 64% to 71%, with non-GAAP EPS of $1.85 to $1.93. Management noted that the prior-year Q2 included a one-time investment gain from the Ampere sale, which affects the EPS comparison.

The fiscal 2026 base shows how far the business has moved. Fiscal 2026 total revenue rose 17% to a record $67.4 billion. Cloud revenue climbed 39% to $34.0 billion, and software revenue slipped 1% to $24.5 billion. Guidance for fiscal 2027 implies revenue growth of at least $22.6 billion in a single year, more than a third of the prior-year total.

The stock's reaction to guidance has been volatile. On September 10, ORCL closed at $152.94, down 5.38% in the regular session, then rebounded to $159.58 after hours on the results and raised outlook, a 4.34% gain from the close. Since then, the stock has given back all of that rebound and more, falling to Thursday's $133.48 low. The fundamental outlook hasn't changed in two weeks. The price has dropped $26 from the after-hours peak.

That disconnect is the core of the forecast. Oracle's guidance doesn't depend on Project Jupiter. The campus isn't scheduled to produce revenue before 2028. Near-term numbers are driven by capacity already delivered and GPUs already installed. The market isn't questioning fiscal 2027. It is questioning fiscal 2028 through 2030, when the bulk of the backlog is supposed to convert and when Jupiter and similar campuses are supposed to be running.

For shareholders, guidance sets a floor on earnings power. At $8.10 in non-GAAP EPS, Oracle is on track for its strongest profit year ever. The question is what multiple investors will pay for that profit given the capital intensity behind it.

Balance Sheet: $125 Billion in Debt and Negative Free Cash Flow

The balance sheet is the reason the stock can't trade on backlog alone. Oracle's debt reached $125 billion after its first-quarter results. The company's free cash flow was negative $23.7 billion for fiscal 2026 as it poured cash into cloud infrastructure buildout. That is a business financing growth almost entirely with external capital.

The funding needs keep rising. In fiscal 2026, Oracle raised $43 billion in debt financing and $5 billion in equity financing. For fiscal 2027, the company expects to raise approximately $40 billion through a combination of sources. In Q1 FY27 alone, Oracle completed the sale of $20 billion of common stock through an at-the-market equity program, diluting existing holders to fund the buildout.

Leverage ratios reflect the strain. Oracle carries a debt-to-equity ratio of 3.89, and its share count has increased 1.68% over the past year before the full effect of the $20 billion ATM sale. The company's Altman Z-Score sits at 2.18. Scores under 3 flag elevated financial risk, though Oracle's cash generation from its software business gives it far more cushion than the score alone suggests.

The prepaid contract structure is Oracle's main defense. The $75 billion in prepaid and customer-supplied hardware across its large AI contracts reduces the capital Oracle needs to raise. Management has stressed that new contracts are being structured to avoid adding to capital needs. That approach shifts some financing burden onto customers, but it also makes Oracle more dependent on those customers' own funding capacity.

Founder Larry Ellison removed one overhang this month. On September 12, Ellison dropped a plan to sell up to $7.5 billion worth of Oracle stock. That decision eliminated a large source of potential supply and signaled insider confidence at prices near $150.

The dividend remains intact. Oracle declared a $0.50 quarterly dividend with an ex-date of October 9, 2026, an annual rate of $2.00. At $137.45, that implies a yield of 1.46%. The payout costs Oracle $5.8 billion a year across its share base, a small fraction of its capital program but a signal that management doesn't see the balance sheet under acute stress.

The balance sheet sets the stock's sensitivity to rates. Every point added to Oracle's borrowing costs on $125 billion in debt adds $1.25 billion in annual interest. With the 10-year Treasury at 5.15%, new debt costs more than anything Oracle issued during the 2020 to 2022 low-rate window.

A 5.15% Treasury Yield Raises Oracle's Cost of Capital

Oracle's selloff didn't happen in isolation. It landed on the second day of a bond market rout that pushed the 10-year Treasury yield to 5.15%, its highest since July 2007, and the 30-year to 5.446%, a peak not seen since June 2004. For a company financing a buildout with debt and equity, those yields are the price of growth.

The rate shock hits Oracle through three channels. First, new debt costs more. Oracle plans to raise approximately $40 billion in fiscal 2027, and every basis point on the benchmark adds to the coupon on that issuance. Second, the discount rate applied to Oracle's backlog rises. A contract that pays out in 2029 is worth less today when the risk-free rate is 5.15% than when it was 4%. Third, project finance partners like Blue Owl face higher funding costs, which makes them less willing to absorb delays.

The Fed path adds to the pressure. Fed funds futures price a 75.3% chance of an October hike and a 58.6% chance of another in December. New York Fed President John Williams said Thursday that another rate increase before year-end is a reasonable expectation. The policy rate already sits at 3.75% to 4.00% after the September 16 hike.

Oracle's high beta amplifies every macro move. The stock carries a beta of 1.73, which means it historically moves 73% more than the broader market. On Thursday, the Nasdaq Composite fell 0.49% to 26,804 in the morning session, and Oracle fell 15 times as much at its low. Most of that gap came from the Jupiter news, but the high-beta profile means Oracle will overshoot in both directions as rates move.

The AI infrastructure trade is repricing broadly. Bloom Energy fell 6.52%, Blue Owl dropped 4.39%, and technology and industrials led sector declines at the open as the Oracle notice rippled through AI-linked names. Investors are separating businesses that sell compute today at premium prices from those whose revenue depends on multi-year physical buildouts.

Oracle sits on both sides of that line. Its current cloud infrastructure business sells capacity at 97.9% utilization, generating revenue now. Its future backlog depends on campuses like Jupiter clearing permitting, power and financing hurdles. The rate shock punishes the second half of that profile. If the 10-year retraces below 5%, the pressure on Oracle's multiple eases and the stock's gap to its post-earnings high narrows fast.

Valuation: 17 Times FY27 Earnings

At $137.45, Oracle trades at 17.0 times its fiscal 2027 non-GAAP EPS guidance of $8.10. That multiple sits well below where the stock has traded for most of the AI cycle and below the market's broad forward multiple. Investors are paying less for each dollar of Oracle's earnings than for the average large-cap stock, despite cloud infrastructure growth of 121%.

The trailing numbers tell a similar story. Oracle earned $5.83 per share over the past 12 months on revenue of $67.36 billion and net income of $16.98 billion. At $137.45, the trailing P/E stands at 23.6. The gap between trailing and forward multiples reflects the 39% earnings growth implied by guidance.

The growth-adjusted picture is stark. Oracle's PEG ratio stood at 0.65 before this week's selloff, well under the 1.0 level that marks fair value on a growth-adjusted basis. Thursday's decline has pushed that lower. On a pure earnings-growth basis, the stock looks cheap.

The market isn't pricing Oracle on earnings, though. It is pricing free cash flow, and free cash flow was negative $23.7 billion last year. A company that burns cash to grow earns a lower multiple than one that funds growth internally, because shareholders absorb dilution and creditors take priority on cash flows. The $20 billion ATM sale in Q1 shows the dilution risk in action.

Market value frames the backlog. Using the 2.88 billion shares Oracle reported before completing its $20 billion ATM sale, the stock's market value at $137.45 comes to $396 billion. That is less than two-thirds of the $664 billion RPO. The market is valuing the entire company, including its software franchise, at a steep discount to contracted future revenue. That gap reflects the cost to deliver the backlog, the time value of money at 5.15% rates and execution risk.

Short interest is modest. The latest reading shows 47.77 million shares sold short, or 1.66% of shares outstanding. That is low for a stock with this much volatility, which means the selloff is driven by long holders trimming rather than a crowded short bet. It also means there's little short-covering fuel for a sharp rebound.

The valuation argues that much of the Jupiter risk is already priced. At 17 times forward earnings, the stock assumes meaningful execution risk on the backlog. If Oracle clears Jupiter's power and permitting hurdles or shows stronger free cash flow in coming quarters, the multiple has room to expand toward 20 times, which on $8.10 in EPS implies $162.

Competition and the Hyperscaler Race

Oracle competes in the most capital-intensive race in technology, against rivals with far deeper balance sheets. Amazon Web Services, Microsoft Azure and Google Cloud dominate the cloud infrastructure market, and each funds its buildout from massive internal cash flows. Oracle is the fourth hyperscaler, the physical challenger to the big three, and it has chosen to grow faster by borrowing more.

The competitive relationship is unusual. Oracle's multicloud strategy places its database services inside the data centers of Amazon, Google and Microsoft. That makes the three largest cloud providers both Oracle's competitors and its partners. The multicloud business grew at extraordinary rates through fiscal 2026 as Oracle rolled out datacenters inside its hyperscaler partners' facilities.

In AI infrastructure, Oracle's edge has been speed and pricing. Its networking design and willingness to build dedicated capacity for single customers won large AI training contracts, including with OpenAI. The 121% growth in cloud infrastructure revenue shows that strategy is winning share. Oracle bills itself as the world's fastest-growing provider of cloud datacenters.

Newer competitors are pushing into the same market. CoreWeave, a pure-play AI cloud provider, picked up an upgrade to Overweight with a $125 price target this week on the strength of compute pricing and premium short-term contracts. CoreWeave competes directly with Oracle for GPU cloud customers and faces similar capital intensity. The difference is contract duration: CoreWeave is leaning into short-term contracts at premium pricing, while Oracle has locked in multi-year commitments that push revenue further out.

The competitive threat that matters most is financial rather than technological. Microsoft, Amazon and Google can absorb delays at individual data center campuses without touching their balance sheets. Oracle can't. A permitting setback that costs a larger rival a few percent of annual capex costs Oracle a larger share of its capacity plan and forces it to renegotiate with developers and lenders. The force majeure notice is Oracle managing a risk its larger rivals carry more easily.

The sector backdrop is also shifting. The Nasdaq has pulled back from its record close of 27,244 on September 22, and AI infrastructure names are leading the decline. Meta's Muse agent and the broader AI application layer have drawn investor money this month, while capex-heavy infrastructure names are being sold. For Oracle, that rotation argues for a period of consolidation before the stock can reclaim leadership.

Technical Levels: $133.48 Support, $144.56 and $159.58 Resistance

The chart gives traders a clean map after Thursday's volatile session. The stock opened near Wednesday's $144.56 close, briefly traded as high as $143.17, then broke down to $133.48 before rebounding to $137.45. That intraday reversal, a long lower wick with buyers absorbing the selling, is the first technical signal of demand at the lows.

The first support is Thursday's low at $133.48. A daily close below that level would confirm that the Jupiter selloff has further to run and invalidate the intraday rebound. Below that, the $130 round number sits as the next reference, followed by the 52-week low at $114.50. A move to $114.50 would represent a 16.7% decline from current levels and would erase the entire rally from the spring low.

On the upside, the first resistance sits at the $143.17 to $144.56 zone, Thursday's high and Wednesday's close. The stock needs to reclaim that area to fill the Jupiter gap and signal that the notice has been fully absorbed. That level also marks early-August territory where the stock consolidated before the earnings run-up.

The second resistance is $152.94, the regular-session close on September 10, the day of Q1 results. The third is $159.58, the after-hours peak on the earnings release. Clearing $159.58 would put the stock back above its post-earnings high and signal that the market has moved past the balance sheet concerns. That level sits 16.1% above current price.

The trend structure is bearish on longer timeframes. ORCL has been in a technical downtrend since early January, with a series of lower highs from the $322.54 52-week peak. The stock trades below its 200-day simple moving average, which caps rallies on the daily chart. Until the stock closes above that average, rallies count as bounces inside a downtrend.

Volume provides the confirmation signal. Thursday's 29.35 million shares by mid-afternoon ran below the 38.88 million average, which means the selling lacked the capitulation volume that typically marks a durable low. A retest of $133.48 on heavier volume that holds would provide stronger evidence of a floor. A break on heavy volume would signal institutional distribution.

The beta of 1.73 means these levels will be tested quickly. Oracle's daily moves routinely run 3% to 5%, so the distance between $133.48 support and $144.56 resistance can be covered in a single session on the right macro or company news.

Catalysts: Jupiter Updates, the October 9 Dividend and Q2 Results

The next few months carry a clear sequence of catalysts that will decide whether Oracle breaks higher or lower. The most immediate is any update on Project Jupiter. Clarity on the pipeline permitting, a revised timeline from Oracle or Blue Owl, or news from the New Mexico State Land Office would move the stock sharply. A permit approval would remove the core risk behind Thursday's selloff. A further delay or a dispute between Oracle and the developer would extend it.

The October 9 ex-dividend date is a minor support factor. Holders of record by that date receive the $0.50 quarterly payment. Income-oriented buyers sometimes step in ahead of ex-dates, and the 1.46% yield at current prices is higher than it was when the stock traded above $150.

The macro calendar matters because of Oracle's 1.73 beta. Friday brings U.S. durable goods orders and consumer sentiment data. The late-September core PCE release will shape October Fed hike odds, currently at 75.3%. The Fed meets in late October. Any sign that yields are peaking would lift Oracle faster than the broader market. A further push in the 10-year toward 5.30% would weigh on it harder.

AI sector news will drive sentiment. Micron reports fiscal fourth-quarter results on September 30, a read on AI memory demand that will move the entire infrastructure complex. OpenAI's funding and buildout plans matter directly, given its role as a Stargate partner and major Oracle customer. Any change to Stargate's scope or timeline would hit Oracle's backlog conversion outlook.

Oracle's own fiscal second-quarter results, expected in December, are the next major fundamental test. The company guided to revenue growth of 30% to 34%, cloud growth of 64% to 71% and non-GAAP EPS of $1.85 to $1.93. Investors will focus less on those numbers and more on capital expenditure, free cash flow, new financing and any change in the RPO conversion timeline. A quarter that shows free cash flow improving would be the most powerful catalyst for a multiple re-rating.

Insider activity is worth tracking. Ellison's decision on September 12 to cancel a planned sale of up to $7.5 billion in stock removed supply. Any insider buying near current levels would reinforce confidence. Any revival of selling plans would add pressure.

Oracle Stock Forecast: $159.58 Upside, $114.50 Risk, $133.48 the Trigger

The forecast comes down to one level and two variables. The level is $133.48, Thursday's intraday low where buyers stepped in. The variables are the Project Jupiter timeline and the 10-year Treasury yield at 5.15%.

The bull case needs three things. Oracle or Blue Owl provides clarity that Project Jupiter's power supply is on track, or New Mexico approves the pipeline permits. The 10-year yield retraces below 5%. The stock holds above $133.48 on a closing basis. Under that path, ORCL reclaims the $143.17 to $144.56 gap zone, extends to the $152.94 September 10 close and targets the $159.58 post-earnings high, 16.1% above current price. A 20 times multiple on $8.10 in fiscal 2027 EPS implies $162. Assigned odds: 30%.

The base case is consolidation. Jupiter uncertainty lingers without new negative developments, yields stay near 5.15%, and Oracle trades between $133.48 and $152.94 into its December Q2 report. The stock's 17 times forward multiple limits downside while the balance sheet and rate backdrop cap upside. Month-end target in this path: $136 to $145. Assigned odds: 45%.

The bear case needs a further Jupiter setback, such as another pipeline delay or a formal dispute with the developer, alongside a push in the 10-year toward 5.30% and a broader AI infrastructure selloff. ORCL breaks $133.48 on a daily close, loses the $130 round number and slides toward the $114.50 52-week low, 16.7% below current price. At that level, the stock would trade at 14.1 times fiscal 2027 EPS guidance. Assigned odds: 25%.

The signals to track are specific. Daily closes relative to $133.48 and $144.56. Any Jupiter permitting news from New Mexico. The 10-year Treasury yield relative to 5% and 5.30%. Bloom Energy and Blue Owl share prices as real-time readings on Jupiter risk. Oracle's Q2 capital expenditure and free cash flow when it reports in December.

Verdict: Hold. Oracle's operating business is firing, with cloud infrastructure revenue up 121%, a $664 billion backlog and guidance calling for $90 billion in revenue and $8.10 in EPS this fiscal year. At 17 times forward earnings, much of the Jupiter risk is priced, and the intraday rebound from $133.48 shows buyers are active. But $125 billion in debt, negative free cash flow of $23.7 billion and a 5.15% Treasury yield make every delay expensive, and Jupiter's power problem won't resolve before February 2027 at the earliest. A daily close above $144.56 with Jupiter clarity flips the call to Buy with a $159.58 target. A daily close below $133.48 flips it to Sell with a $114.50 target. Until one of those triggers fires, ORCL is a range trade with the bond market and New Mexico's land office holding the key.

That's TradingNEWS