On February 1, Oracle said in a statement that the company is planning large-scale financing to expand its data center to meet the contractual needs of its largest cloud customers such as AMD, Meta, Nvidia, OpenAI, TikTok and xAI.

How to raise money? About half of the funds will be raised through the issuance of common shares and equity instruments; This includes the issuance of mandatory convertible preferred securities and an equity financing plan of up to $20 billion (i.e., the possibility of selling shares in installments in the market in the future).

Analysts pointed out that this move shows that Oracle is fully betting on the AI cloud service track, hoping to grab a larger share in the competition with Microsoft, Amazon, and Google.

But investors are also worried: Oracle currently has about $95 billion in debt; It is one of the most indebted companies among non-financial enterprises. Will such an aggressive expansion further increase the financial pressure?

To put it simply, Oracle is taking a two-pronged approach of "borrowing money + selling shares", betting on the future of AI cloud - the ambition is very high, and the risk is not small.

Details of Oracle's $50 billion financing plan have been exposed

According to media reports, Oracle is planning to raise a total of about $50 billion through the "equity + bond" combination to large-scale expansion of AI cloud infrastructure.

How to raise money?

Approximately $22.5 billion to $25 billion came from equity financing, including the issuance of mandatory convertible preferred securities (a special type of stock that will be converted into common stock in the future); and launching an "at-the-market program" of up to $20 billion, that is, to sell company shares flexibly according to market conditions for a period of time to come.

Another $22.5 billion to $25 billion will be raised through bond issuance: a one-time issuance of this huge bond is planned in early 2026. This will further boost Oracle's already massive debt, which currently has about $95 billion in outstanding debt.

What is the money used for?

Oracle made it clear that all funds will be earmarked for the construction of new cloud data centers to honor contracts with top customers.

Who are these big customers?

The list almost includes the core players of the current AI ecosystem:

Chip giants: AMD, Nvidia.

AI model companies: OpenAI, xAI (owned by Musk).

Internet platforms: Meta (parent company of Facebook), TikTok.

Short sellers fired: Oracle's "heavy position AI cloud" is too risky

Oracle recently announced a $50 billion financing plan, but the market didn't buy it. Many investors, especially short sellers, criticized: the company's transformation is too aggressive and the financing is too aggressive.

Why are you worried?

At present, Oracle has about $95 billion in debt, one of the most indebted companies among non-financial companies (according to the Bloomberg High Rating Index). Behind this is that it has shifted from an asset-light company that "sells database software" to an asset-heavy model of self-built data centers - in order to compete with Amazon AWS and Microsoft Azure for customers, it has spent tens of billions of dollars to buy NVIDIA chips and build data centers.

The result: capital expenditure soars year after year, and debt grows bigger and bigger.

Coupled with the opaque structure of some cloud contracts, the market began to question: Is this money worth it? Can the returns be cashed out? What do you think about short-selling Michael Burry?

Well-known investor Michael Burry (the prototype of "Big Short") bluntly said: "Oracle is doing something unnecessary and extremely risky." His core point is: Microsoft, Google (Alphabet), Meta and other technology giants have super earning power in their main business (such as Windows, search, advertising), even if AI investment does not return for the time being, it can still bear it; But Oracle does not have such a "safety cushion" - its stock price is now almost entirely supported by the story of "AI cloud demand explosion"; Once the AI boom cools down and customer orders fall short of expectations, high debt + low fault tolerance = huge risk.

Burry even called Oracle a "pure AI bubble carrier" - meaning: it is not as diverse and stable as other giants, but bets all its bets on AI, one prosperous and one loses.

He also emphasized that Microsoft and Google can slowly digest excess capacity and adjust the pace, but Oracle's financial structure is more fragile and cannot withstand a major miscalculation.

Jingtai view|High-risk games require extreme caution

This is a big gamble with $95 billion in debt + 50 billion in new financing, betting on "AI cloud can save old software companies".

For U.S. stock investors:

Short-term sentiment may be boosted by the endorsement of major customers, but in the long run, free cash flow and debt coverage ratio are the key; If AI capital expenditure cools down in 2026 (such as a slowdown in Nvidia orders), Oracle may bear the brunt of it; Be wary of the dilution effect brought about by equity financing - the 20 billion ATM share issuance plan can be launched at any time.

For observers of the cloud computing track:

Oracle's radicalism reflects that the AI cloud race has entered the "arms race 2.0" stage - fighting not only for technology, but also for capital endurance; But the market will eventually return to rationality: who has real customers, stable gross profit margins, and controllable leverage, will survive to the end.

Mapping of A-shares/Hong Kong stocks:

Domestic cloud vendors (Alibaba Cloud, Tencent Cloud) are also facing the problem of "AI investment vs. profit balance"; You can pay attention to platform-based companies with independent chips + high-efficiency data centers to avoid the pure money-burning model.