The $300 billion agreement between Oracle and OpenAI, first revealed on September 10, has gone deep into the red; negative $74 billion to be exact.

The market wiped out $315 billion from Oracle’s value since the announcement, a loss that not only erased the entire worth of the deal but also torched the equivalent of General Motors plus two Kraft Heinz.

While major tech benchmarks like the Nasdaq Composite, Microsoft, and the Dow Jones US Software Index stayed mostly unchanged, Oracle got smoked.

The brutal fall wasn’t because of overall market pressure, it was all about Larry Ellison’s bet. Oracle threw billions into a deal built on credit, hoping it could become OpenAI’s main infrastructure partner. Investors didn’t like that. Not when the entire business model is now tied to one customer and one moonshot mission: artificial general intelligence.

Oracle doubles debt to chase OpenAI, bets future on one client

Oracle positioned itself as the go-to compute engine for Sam Altman’s OpenAI, offering what it claims are lower upfront costs and faster income potential than other cloud giants.

But here’s the problem; Oracle isn’t flush with cash like its rivals. It’s not Amazon, it’s not Google, and it’s sure as hell not Microsoft. Instead of big operating profits, Oracle is pouring every last drop of capital into the OpenAI pipe dream, with the hopes it’ll pay off later.

At Oracle’s analyst meeting held last month in Las Vegas, the company said it wants to hit $166 billion in cloud revenue by 2030. To get there, it’s throwing $35 billion into capital spending this year alone.

By 2029, annual capex is expected to hit $80 billion. That’s just to keep pace. And from 2027, Oracle is banking on OpenAI to become its top revenue source. Everything, servers, software, and cash, is being stacked on OpenAI’s shoulders.

But Oracle’s financials are already strained. The company’s net debt has more than doubled since 2021 and now sits at 2.5 times EBITDA. And it’s forecast to nearly double again by 2030. Even worse, cash flow is projected to stay negative for five straight years. That means Oracle is quite literally borrowing to gamble.

The deal might already be priced out of Oracle stock, but the company’s debt situation hasn’t calmed. Credit-default swap (CDS) costs, used to insure against debt default, just reached a three-year high.

Liquidity in Oracle’s CDS market isn’t great either, but demand jumped following $18 billion in bond sales in September. A CDS premium hovering around 100 basis points isn’t off-the-charts bad, but it shows that investors are uneasy. Not every firm shorting Oracle debt is clueless.

OpenAI hype fades, other partners see similar stock slides

The question now is whether announcing an OpenAI deal is even worth it. A few months ago, just saying “OpenAI” could spike a company’s shares. In October, OpenAI landed warrants in AMD as part of a chip deal, and AMD stock jumped 24%.

That was then.

Now?Not so much. Oracle isn’t the only name in free fall. Broadcom and Amazon both dipped after their own OpenAI announcements. Nvidia didn’t move much after its OpenAI investment either.

The so-called OpenAI glow-up is starting to look like a fading trick light.

So now Oracle’s stuck. It’s got a $300 billion partnership that investors have already punished with a $315 billion sell-off. It’s facing negative cash flow, rising debt, and a CDS market flashing warning signs. All for one company, one strategy, and one longshot shot at AGI.

The AI capex arms race is climbing toward $1 trillion, but the market no longer rewards participation trophies. If Oracle can’t get a stock bump out of its biggest-ever deal, then Wall Street’s message is clear: no faith, no lift, no gain.

The smartest crypto minds already read our newsletter. Want in? Join them.