Oracle gobbles $20 billion borrowing limit a week after setting it

Oracle plans to raise $20 billion at once through $10 billion of bonds and $10 billion of common stock, after saying it would cap new borrowing at $20 billion through fiscal 2030; CFO Hilary Maxson said the borrowing is expected during the remainder of fiscal 2027. The company spent $55.7 billion on capex in fiscal 2026 and plans $90 billion–$95 billion in fiscal 2027, with datacenter investment the likely use of proceeds. The article also notes potential financing costs above $100 million annually for a Wisconsin datacenter power commitment and reported payment-deferral discussions regarding Project Jupiter, which Oracle says remains on schedule.

Analysis

The key read-through is not simply higher leverage: funding with equity alongside debt makes the borrowing ceiling compatible with continued capex, but shifts part of the buildout cost to existing shareholders. It also signals that management is prioritizing investment capacity over buyback flexibility. The debt-cap statement is therefore not a cap on total external funding or investment.

For ORCL, the immediate pressure is likely issuance supply and uncertainty over dilution and bond pricing. Over the next 1–3 months, the market needs evidence that AI demand converts into billable usage and cash flow quickly enough to support the buildout. Large contracted commitments are not equivalent to realized revenue, particularly if projects slip or a concentrated customer delays deployment. Over 6–18 months, power availability, campus delivery and customer concentration matter more than headline backlog; any gap between capex and monetization raises the risk of lower returns on invested capital and multiple compression.

OWL is a watch item, not a clean short: the reported New Mexico payment dispute could matter to sentiment around data-center financing, but the supplied facts do not establish OWL’s exposure or likely loss. The contrarian point is that explicit use of equity may reduce hidden-financing concerns, even as dilution weighs on the stock. Verify offering terms, share count, bond maturity/coupon, and project-level payment obligations before sizing either security.

Key Decisions for Investors

• Keep ORCL underweight through pricing of the equity and bond offerings; reassess once dilution, coupon and use-of-proceeds detail are public. Falsifier: strong post-raise guidance for AI usage revenue and cash conversion without further financing needs.
• Avoid treating the stated borrowing limit as a cap on capital intensity: track total capex, operating cash flow, buybacks and any guarantees or other obligations in subsequent disclosures.
• Set a 1–3 month alert for evidence of AI backlog conversion—recognized revenue, customer usage and datacenter delivery milestones. Slippage or weaker monetization would strengthen the underweight; accelerating conversion would weaken it.
• Do not trade OWL solely on the reported Project Jupiter notice. First verify project-level financing exposure and whether any deferral changes expected cash flows; absent that evidence, the read-through is sentiment risk rather than a demonstrated earnings impairment.