Oracle spent its last remaining publicly committed capital all at once on November 2, 2026, filing with the Securities and Exchange Commission to raise 10 billion dollars in new stock and 10 billion dollars in bonds five days after telling investors at the October 28 Investor Day in Las Vegas that 20 billion dollars was everything it would ever need to borrow again through fiscal 2030. The Oracle capital raise did not violate the numerical promise, but it tested the credibility behind it.
The pledge, announced by co-CEOs Clay Magouyrk and Mike Sicilia and CFO Hilary Maxson, covered all new capital-market raises across fiscal 2027 to 2030, split evenly between equity and gross debt and excluding refinancing of existing obligations. Oracle shares closed up 10.73 percent to 144.54 dollars on October 28, according to the reported market data. The market interpreted the cap as evidence that investors were less concerned about Oracle’s 455 billion dollar first-quarter AI backlog than about the balance sheet required to build it.
Five days later, Oracle deployed the entire four-year commitment. The stock offering and bond offering are separate and neither depends on the other closing. At the stock filing’s reference price of 130.67 dollars, the October 30 last sale, the 10 billion dollar offering would require roughly 76.5 million new shares. That would take the count from about 3.02 billion to about 3.10 billion, dilution of roughly 2.5 percent, and likely more if the price remains at the lower level reached after the announcement. Oracle said proceeds would go to general corporate purposes including capital spending, debt repayment, acquisitions, dividends and share buybacks.
Oracle capital raise turns a promise into a credibility test
The central issue is not whether Oracle stayed within its stated limit. It did. The issue is whether using the full allowance immediately served the purpose of a multi-year financing framework. A cap can reassure investors by setting a boundary, but it provides little comfort if the boundary is reached before the spending cycle has produced meaningful cash.
The share-price response was severe. The November 2 close was 119.22 dollars, 17.52 percent below the October 28 close, leaving a market value of about 361 billion dollars. Other coverage described the loss since Wednesday as roughly 60 to 65 billion dollars and the total decline as about 15 percent. That reaction suggests investors did not reject Oracle’s AI strategy so much as question the financing strategy that management said would support it.
The timing also undercut the apparent simplicity of the promise. Oracle used the entirety of its public capital allowance during the same trading week in which it was defending a sharply higher operating forecast. The balance between discipline and necessity now depends on whether the company’s revenue ramp arrives quickly enough to absorb the financing without requiring another raise.
The debt and cash-flow challenge behind the filings
Oracle entered the financing with substantial borrowing already on its books. As of August 31, 2026, the company held 125.34 billion dollars in notes payable and other borrowings. S&P Global Ratings downgraded Oracle to BBB- on July 9, citing an uncertain path to profitability and a negative free operating cash flow outlook. BBB- is the lowest rung of investment grade. A further downgrade to BB+ would make Oracle a fallen angel, moving its debt into the high-yield category and potentially forcing sales by pension funds and insurers that are restricted from holding such bonds.
Morgan Stanley credit analyst Lindsay Tyler said there was a risk of two low BBB ratings for Oracle, perhaps by calendar year-end. George Catrambone, head of fixed income at DWS Americas, said the market’s view of Oracle’s balance sheet had changed abruptly. Forced selling by constrained investors could widen Oracle’s bond yields and increase the cost of refinancing its existing debt.
The near-term financial arithmetic is demanding. Capital expenditure rose 162 percent to nearly 56 billion dollars in fiscal 2026, driving free cash flow to negative 23.7 billion dollars against revenue of 67.36 billion dollars. Oracle has guided to roughly 90 billion dollars of fiscal 2027 revenue, with capital spending expected to peak in fiscal 2027 and fiscal 2028. Management expects free cash flow to turn positive in the second half of fiscal 2028 and exceed 30 billion dollars in fiscal 2029, which ends in May 2029.
Those forecasts leave little room for a delayed customer payment, weaker infrastructure demand or additional construction costs. Oracle can remain within its fundraising cap while still needing more capital if its spending outruns the revenue conversion curve. That is why investors are focused on cash timing rather than only the headline backlog.
Bond investors price a higher risk premium
The new bond sale was structured across five fixed-rate tranches maturing between 2029 and 2046. Rates set on November 2 ranged from 7.000 percent to 8.650 percent, significantly above the roughly 6.5 percent pre-offering secondary yield on comparable ten-year paper. The pricing shows that Oracle had to pay for the possibility that heavy AI investment will strain its balance sheet.
The contrast with February is instructive. Oracle’s February 2026 bond sale of 25 billion dollars in eight tranches attracted 129 billion dollars of orders, the largest demand for a high-grade corporate bond deal on record. It also made Oracle the second-largest non-financial corporate bond issuer in the Bloomberg index, behind only Amazon. New Oracle bond spreads were reported as 145 to 175 basis points wider than those from the February deal on the closest maturities. Investors are charging more for Oracle debt because the risk profile has changed alongside the infrastructure build-out.
Stephens still upgraded the stock to Overweight and raised its price target to 213 dollars from 185 dollars, doing so 23 minutes before the filing. The split verdict illustrates the debate. Equity analysts focused on the large addressable market and long revenue runway, while bond investors focused on cash burn, ratings pressure and the cost of funding that runway.
For bondholders, the priority claim on Oracle’s cash flow provides some protection that common shareholders do not receive. Fitch affirmed its BBB stable rating and said it did not expect Oracle to raise material incremental capital over the rating horizon. S&P Global Ratings’ stable BBB- rating remained one notch above junk. Those positions offer some comfort, but they do not remove the risk that Oracle’s spending cycle could test the forecasts more severely than investors expected.
Backlog provides demand evidence, not cash
Oracle’s contracted business gives the capital plan a real foundation. Oracle Cloud Infrastructure revenue grew 55 percent year over year to 3.3 billion dollars in the first quarter of fiscal 2026, part of 7.2 billion dollars in total cloud revenue. By the fiscal fourth quarter, OCI infrastructure-as-a-service revenue had accelerated to 5.8 billion dollars, up 93 percent year over year.
Remaining performance obligations, which represent contracted revenue that has not yet been recognised, stood at 455 billion dollars in the first quarter. By the fiscal year’s end, that figure had grown to 638 billion dollars. These are separate figures and should not be treated as one number. The distinction matters because a backlog describes future contractual revenue, not cash already collected or profit earned.
Named OCI customers driving the build-out include AMD, Meta, Nvidia, OpenAI, TikTok and xAI. OpenAI alone is reported to have committed to a 300 billion dollar five-year cloud contract. The Stargate project with OpenAI and SoftBank targets 4.5 gigawatts of US data centre capacity. Its first facility at Abilene, Texas, is already operational with Nvidia GB200 racks drawing 1.2 gigawatts when fully loaded.
Building that capacity requires Oracle to purchase equipment, construct facilities and install power and networking before customers begin paying for billable consumption. A memory shortage can raise prices and delay deployments, but it can also increase capital costs and postpone cash receipts. The customers and contracts support Oracle’s long-term case, while the build-out sequence creates the near-term financing pressure.
What investors will watch next
Three questions will determine whether the strategy becomes credible again. First, contracted customers must pay on schedule. Second, Oracle must convert its backlog into revenue quickly enough to improve free cash flow. Third, the rating agencies must maintain investment-grade classifications without forcing a costly capital restructuring.
Jefferies said after the February financing announcement that the plan bought time for Oracle’s AI ambitions but warned that free cash flow was unlikely to turn positive until fiscal 2029. That timetable remains the pivotal assumption. Brad Smith, a Janus Henderson portfolio manager, described Oracle as shifting toward a clear policy of investment-grade credit and a balance between debt and equity. The immediate sale of the full amount leaves Oracle to demonstrate that this balance can work over time.
Management has also described the new financing as the final amount needed for the remainder of the current cycle. That statement provides a useful boundary. It also raises the standard for execution because Oracle has now answered the funding question before most of the relevant cash-flow evidence exists.
The market will look to the final stock pricing, quarterly free-cash-flow results, capital expenditure guidance and any rating action for confirmation. A stock price near 119.22 dollars may attract investors who believe AI infrastructure demand will remain strong. Bond investors will focus more narrowly on Oracle’s ability to service a growing debt stack while capital spending stays elevated.
Oracle capital raise was not an accounting breach, because the company stayed inside the disclosed cap and provided separate stock and bond offerings. It was a credibility test compressed into five days. The AI infrastructure backlog and memory shortage offer real strategic context, but only cash generation can settle whether Oracle’s balance sheet can carry the ambition investors rewarded on October 28. The Oracle capital raise now has to be followed by delivery.