Morgan Stanley estimates AI-related global debt issuance will hit $570 billion in 2026, with hyperscaler spending exceeding $1 trillion by 2027. Oracle’s crisis may be the first systemic warning sign.
The question Wall Street was reluctant to ask openly throughout 2024 and most of 2025 is now unavoidable: is the AI infrastructure buildout generating a debt burden that markets have not yet properly priced?

The numbers have become too large to dismiss as routine capital expenditure cycles. Morgan Stanley estimates that AI-related global debt issuance will more than double to nearly $570 billion in 2026, with aggregate hyperscaler capital expenditure projected to exceed $1 trillion by 2027. That figure encompasses spending by Amazon, Microsoft, Alphabet, Meta, Oracle, and a growing constellation of second-tier infrastructure providers building the physical layer of the AI economy.

How the Debt Stack Has Built

The trajectory of Oracle’s balance sheet is instructive as a case study in the speed at which leverage can accumulate. In fiscal 2025, Oracle carried a net cash deficit of approximately $394 million after free cash flow. By the end of fiscal 2026, that had deteriorated to negative $23.7 billion in free cash flow, with long-term debt reaching approximately $124.7 billion. Capital expenditures of $55.7 billion in a single fiscal year represent a 162% increase from the prior year.

Oracle is not alone, though its position is the most stretched. The structural dynamic across the hyperscaler complex is that the companies investing most aggressively in AI data centre capacity are simultaneously facing competitive pressure on their existing software and cloud businesses from AI-native tools — creating a margin squeeze that occurs precisely when cash demands are highest.

Credit Default Swaps as an Early Warning System

One underappreciated signal in this cycle is the behaviour of credit default swaps. Fortune reported that Morgan Stanley’s Lisa Shalett flagged Oracle’s CDS widening as a potential early indicator of broader AI trade stress. CDS spreads — which function as insurance premiums against corporate default — had reached record levels for Oracle by early 2026, even before the most recent earnings-related stock decline.

The concern Shalett articulated was systemic rather than company-specific: “If people start getting worried about Oracle’s ability to pay, that’s gonna be an early indication to us that people are getting nervous.” For a company whose debt is included in major corporate bond indices, the widening of Oracle’s CDS spreads has implications not just for Oracle investors but for anyone holding investment-grade credit exposure broadly.

Bank of America Research described “the lack of clarity on hyperscaler borrowing” as “the key risk going into 2026” — a view validated by subsequent events as Oracle’s stock collapsed and CDS widened even further.

The OpenAI Nexus

A critical vulnerability embedded in the current AI infrastructure cycle is concentration around OpenAI as both the defining customer and the primary justification for hyperscaler spending. Oracle‘s remaining performance obligations are concentrated at least $300 billion in the OpenAI relationship. OpenAI itself is burning cash at what one analyst described as “an insane rate” and has committed to more than $1.4 trillion in total AI buildouts — a commitment that depends on the company’s own ability to sustain fundraising and ultimately generate revenue at scale.

The logical chain from that dependency is a concern articulated plainly by Melius Research: “It is hard to know if Oracle can stick to this capex plan if incremental business arises from the likes of OpenAI and Anthropic. Also, its competitors are unlikely to slow spending and could use Oracle’s spending moderation as the means to gain share.” The competitive dynamic creates a collective action problem: no single hyperscaler can slow down without ceding ground, yet the collective pace of spending is generating balance sheet stress across the sector.

Second-Order Vulnerabilities: Data Centre REITs and Chip Suppliers

The debt accumulation in hyperscaler balance sheets has second-order effects that are not captured in the headline AI capex numbers. Data centre real estate investment trusts — which provide the physical infrastructure that hyperscalers increasingly lease rather than own — have their own exposure to counterparty concentration and lease extension risk. Reports that Blue Owl, Oracle‘s primary data centre financing partner, declined to back the Michigan facility highlighted the fragility of the supporting ecosystem even when the primary tenant appears solvent.

Nvidia, whose chips underpin the entire AI buildout, has been insulated from these concerns by persistent demand that exceeds supply. But if even two or three hyperscalers simultaneously scaled back data centre spending in response to balance sheet pressures, the chip demand outlook would shift rapidly.

The Memory Shortage as Collateral Signal

CNBC reported in late June 2026 that “the memory shortage shaking Apple and Microsoft is an ‘existential crisis’ for smaller players” — a reminder that supply chain bottlenecks are not yet resolved, adding cost and execution risk to projects whose timelines are already being stretched. The combination of persistent demand exceeding supply, expensive debt financing, and uncertain monetisation schedules creates a financial engineering challenge that may prove harder to solve than the engineering challenges of building the data centres themselves.

The AI infrastructure cycle is not necessarily a bubble in the sense of zero underlying demand — the use cases are real and adoption is accelerating. But the debt structure being used to finance it, and the concentration of risk around a small number of foundational relationships, has introduced systemic vulnerabilities that markets are only beginning to price.

China’s May 2026 data shows high-tech manufacturing up 15.1% while property investment fell 16.2%. How Beijing’s export-led gamble is reshaping global supply chains.

The National Bureau of Statistics’ May 2026 release confirmed what economists had begun calling China’s “industrial divergence.” Scale-above industrial value-added output grew 4.5% year-on-year in May, accelerating 0.4 percentage points from April, with high-tech manufacturing surging 15.1%. The semiconductor sector was the standout: domestic output jumped 87% from the prior year, while China’s exports of semiconductors were up 110% from a year earlier, exports of mobile phones climbed 44%, and automatic data-processing machines rose 66%.

The Export Engine Running at Full Throttle

China‘s May exports (denominated in US dollars) were up 19.6% from a year earlier — the second biggest monthly increase since January 2022. The first two months of 2026 had registered an extraordinary 39.6% gain. Over all of 2025, China recorded a trade surplus exceeding $1.2 trillion — the largest ever posted by any country — as manufactured goods, particularly in advanced technology categories, poured into global markets.

The strength carries a double driver. First, the global AI boom has generated extraordinary demand for semiconductors and related hardware, where China‘s manufacturing base has rapidly scaled. Second, as domestic demand softened, manufacturers redirected capacity toward export markets. Gary Ng, senior Asia Pacific economist at Natixis, characterised this as the operative dynamic: “China’s exports have decelerated as the Iran war starts to affect global demand and supply chains,” though he noted the moderation was from record levels.

China’s economy in mid-2026 resembles a dual exposure photograph — one frame showing a technology powerhouse outpacing global rivals, the other depicting a property market in structural retreat that is slowly draining household wealth.

Goldman Sachs had projected 5–6% annual growth in China’s exports and raised its 2026 real GDP forecast to 4.8% — above both IMF projections and Bloomberg consensus. That upgrade rested on the observation that Chinese exports demonstrated resilience even against elevated US tariffs that hit 100% in April 2025 before settling at 30% in May following a bilateral agreement. Chinese exports of chips, semiconductors, autos, and auto parts continued to expand despite the tariff headwinds.

The Property Hole That Will Not Close

The other side of the ledger is less encouraging. In the first five months of 2026, fixed-asset investment fell 4.1% year-on-year — the steepest decline since May 2020. Within that, property investment dropped 16.2%. Given that roughly two-thirds of Chinese household wealth is held in real estate, the wealth destruction is persistent and consequential. Consumers saving to restore depleted balance sheets rather than spending is the logical response — and it explains why domestic retail demand has been chronically soft despite headline economic growth of 5% in 2025.

The Economist Intelligence Unit’s Nick Marro captured the strategic bet underlying Beijing’s trajectory: “There’s a strong emphasis on doubling down on manufacturing and ensuring that China’s competitive positioning in global supply chains remains sticky.” China‘s 15th Five-Year Plan (2026–2030), approved in late 2025, explicitly prioritises advanced manufacturing, semiconductors, AI, renewable energy, and digital infrastructure — doubling down on supply-side transformation rather than demand-side stimulus.

The Global Spillover: China Shock 2.0

The US-China Economic and Security Review Commission flagged a “14 percent surge in China Shock 2.0,” noting that developing markets are bearing the brunt of an export deluge driven by China’s market distortions. Unlike the original China Shock of the 2000s — which displaced labour-intensive, low-value manufacturing in rich economies — China Shock 2.0 is crowding out high-tech, high-value manufacturing in Europe and Japan. Goldman Sachs estimates that for every 1 percentage point of export-driven boost to Chinese GDP, other economies may see a 0.1 to 0.3 percentage point drag, with tech-intensive producers facing acute pressure.

Meanwhile, China’s voracious appetite for advanced chips it cannot yet manufacture domestically has produced a paradox: China imported a record $135 billion in semiconductors in the most recent quarter as AI investment accelerates. The country remains dependent on foreign-made advanced logic chips dominated by ASML, creating a structural vulnerability that its Five-Year Plan is designed to remedy — but may not resolve within this decade.

The Endgame of the Xi Gamble

The Economist captured the existential dimension of Beijing‘s strategy by quoting Johns Hopkins University‘s Yuen Yuen: “At no time in modern history has a large country gone all in on investment in high-end technology while also navigating a slowing economy and a local-government debt crisis.” Xi Jinping’s wager is that the technology-driven growth model scales faster than the old property-and-construction model collapses. The data through mid-2026 suggest the race is closer than Beijing’s official narrative acknowledges.

China’s GDP growth target for 2026 is the lowest since 1991 at 4.5–5%. Meeting it will depend on whether AI and green technology exports can sustain momentum against an Iran-related global slowdown that is already beginning to weigh on overall demand. The outcome will shape global trade balances, supply chain geography, and the AI chip economy for the next decade.

Oracle’s stock collapsed 24% in 2026 as $130 billion in AI debt and negative free cash flow of $23.7 billion rattled markets. Inside the hyperscaler’s existential reckoning.
Larry Ellison’s audacious pivot to AI infrastructure is drawing comparisons to the dot-com implosion — and for good reason.

Oracle Corp. closed out the week of June 27, 2026 with a stock price of $148.53, down 19% in a single week — the worst weekly performance since the 2001 technology bust. The collapse has shaken not just Oracle shareholders but the entire ecosystem of AI infrastructure optimism that has dominated capital markets for the better part of two years. What began as a generational pivot into cloud computing has become a cautionary tale about how quickly leverage can transform ambition into crisis.

The Numbers Behind the Nosedive

The arithmetic is stark. Oracle’s capital expenditures surged 162% to nearly $56 billion in fiscal year 2026, leaving the company with negative free cash flow of $23.7 billion — a dramatic deterioration from just a $394 million deficit in fiscal 2025. Long-term debt ballooned to approximately $124.7 billion by the end of the third fiscal quarter, making Oracle one of the most leveraged technology companies in history relative to its operating cash generation.

Despite posting total revenue of $67.4 billion for fiscal 2026 — a 17% year-on-year gain — investors focused on what was missing rather than what was achieved. Cloud infrastructure revenue did surge 93% to $5.8 billion in the fourth quarter, and total cloud revenue climbed 47% to $9.9 billion, demonstrating genuine demand. But those gains are being funded by capital markets in a way that is testing the boundaries of investor patience.

Having already raised $43 billion in debt and $5 billion in equity during fiscal 2026, Oracle announced plans to secure a further $40 billion in fiscal 2027 — on top of a previously disclosed $20 billion at-the-market equity programme. The announcement sent shares tumbling roughly 10% in after-hours trading on the day of the earnings call.

The OpenAI Dependency Problem

Central to investor anxiety is Oracle‘s lopsided reliance on OpenAI. The ChatGPT developer accounts for the majority — at least $300 billion — of Oracle’s remaining performance obligations. The concentration risk is extraordinary for a company of Oracle’s scale. If OpenAI stumbles in its own fundraising or fails to monetise its products at the projected pace, the cascade effects on Oracle’s revenue backlog — which rose 325% to an eye-catching figure that initially thrilled analysts — could be severe.

D.A. Davidson analysts warned in a December 2025 note that, “considering Oracle is already barely hanging on to an investment grade rating, we would be concerned about Oracle’s ability to live up to these obligations without restructuring its OpenAI contract.” The concern is not hypothetical: the cost to insure Oracle’s debt against default on credit default swap markets has hit record levels, a signal that bond investors are demanding higher risk premiums.

Morgan Stanley estimates that AI-related global debt issuance will more than double to nearly $570 billion in 2026, with hyperscaler spending potentially exceeding $1 trillion by 2027. Oracle sits at the most precarious position in that ecosystem — large enough to be systemic, but without the balance sheet cushion of Amazon, Microsoft, or Alphabet to absorb multi-year cash burn.

The Margin Trap

There is a structural problem embedded in Oracle’s strategy that goes beyond near-term financing concerns. The company’s traditional enterprise software business carries gross margins of approximately 77%. Infrastructure — the business it is pivoting toward — runs at margins closer to 49% at maturity, according to FactSet analyst consensus. That is a punishing dilution for a company that has historically been valued on premium software economics.

Analysts estimate Oracle will burn roughly $34 billion in cumulative free cash flow over the next five years before the infrastructure business turns cash-flow positive in 2029. “Four or five years is a long time,” Eric Lynch, managing director at Suncoast Equity Management, told Bloomberg. “That’s just not within our investment discipline.” The concern is compounded by reports — which Oracle denied — that completion dates for data centres tied to OpenAI contracts had been pushed back from 2027 to 2028.

Meanwhile, headcount declined 13% to 141,000 employees in fiscal 2026, with pullbacks concentrated in sales and marketing — the exact functions needed to defend the existing software business from AI-native competitors. Larry Ellison, absent from the most recent earnings call, has been surpassed on the global wealth rankings by Larry Page, Sergey Brin, Jeff Bezos, and Michael Dell as the stock’s decline eroded the paper value of his stake.

What Evercore and the Bulls Are Still Saying

Not every analyst has abandoned the thesis. Evercore maintained a buy recommendation, noting that “financing/leverage and the pace of equity issuance” would remain the central investor debate “even as demand signals stay strong.” The company’s fiscal 2027 revenue guidance of $90 billion was left intact, and adjusted EPS targets were nudged higher to $8.05. Evercore analysts argue that the backlog growth and infrastructure demand pipeline are real — the question is whether markets will extend the runway needed to prove it.

The broader tech software sector offers context: the iShares Expanded Tech-Software ETF (IGV) is down 16% year-to-date in 2026, while Oracle has fallen 24% — worse than the index but not in isolation. The investor thesis on enterprise software has broadly softened on fears that large language models will automate away categories of software that have historically commanded subscription premiums.

The Systemic Warning

Oracle’s distress carries implications well beyond its own share price. Fortune reported that Morgan Stanley wealth management’s Lisa Shalett flagged Oracle’s credit default swap widening as an early warning indicator for the broader AI investment complex. If confidence in Oracle’s ability to service its debt erodes, it signals that markets are beginning to reprice the risk embedded in the entire hyperscaler debt stack — a reassessment that could spread to data centre REITs, AI chip suppliers, and enterprise cloud vendors.

The debt load, the leadership transition to dual CEOs Clay Magouyrk and Mike Sicilia, the OpenAI concentration risk, and the structural margin compression collectively make Oracle the most visible stress test of the AI infrastructure buildout in 2026. Whether it passes or fails that test will shape capital allocation across the technology sector for years to come.