Stop gambling with earnings…Start trading them predictably
Earnings used to feel like gambling until I built a checklist.
Now they're the most predictable profit windows on the calendar. Four times a year, like clockwork.
We logged 40 triple-digit winners over 12 months. Fast entries, faster exits, zero burnout.
If you're ready to trade smarter instead of harder, the Earnings Flips replay walks you through everything.
👉 [GET ACCESS NOW]
Don here...
Where's the actual cash?
Brandon Chapman asked a simple question this morning that nobody else is asking about Oracle's $38 billion debt offering.
Oracle is borrowing $38 billion to build data centers for OpenAI. The contract promises $300 billion in payments over five years starting in 2027.
OpenAI generated $12 billion in total revenue this year.
Blake Young did the math live. Oracle expects $60 billion annually from a company currently making $12 billion total. That's 5x current revenue. Every single year. Starting in two years.
The numbers don't work. But the banks are lining up anyway.
In today's free session replay, you'll discover:
- Why AI companies show billions in revenue with zero cash changing hands. Microsoft "invested" in OpenAI by giving them Azure platform credits. OpenAI uses those credits. Microsoft books it as Azure revenue. No actual dollars ever moved. Brandon walked through the circular structure destroying real valuation.
- The data center build-out that doesn't pencil. Oracle's borrowing $38 billion to build data centers for OpenAI. The contract promises $300 billion in payments over five years starting 2027. OpenAI generated $12 billion in total revenue this year. Blake asked the obvious question: where's the other $48 billion per year coming from?
- How Core Weave turned $235 million into phantom billions. Investment firm gives Core Weave $235 million. Core Weave returns $50 million in preferred stock plus $230 million in platform credits. Those credits get passed to new AI startups who use them for "free" compute. Core Weave books it as revenue when used. Brandon called it legalized accounting fraud.
- Why Oracle's jumping from high-margin software to low-margin infrastructure. Software licensing requires zero manufacturing. Gross margins hit 70%. Data centers are completely different. Massive capital expenditure. Ongoing utility costs. Brandon built data centers firsthand. He explained why the margin compression threatens Oracle's valuation.
- The Texas power problem nobody's planning for. Two of Oracle's three planned data centers go in Texas. Blake noted Texas power infrastructure collapses when it gets cold. Oracle promises standalone power generation separate from the grid. That infrastructure doesn't exist yet and costs billions more to build.
Brandon structured this session like a loan officer evaluating Oracle's credit worthiness.
The conclusion? Oracle's a terrible credit risk.
They're taking on 40% more debt to build infrastructure for a tenant that can't possibly generate the promised revenue. The whole structure depends on AI companies eventually paying real cash instead of shuffling credits around.
Blake compared it to Dumb and Dumber. Lloyd trades a Lamborghini for an IOU note saying "as good as cash." Then uses that IOU as investment capital. Eventually someone has to collect the actual Lamborghini.
The AI economy has IOUs stacked on IOUs. Microsoft gave OpenAI credits. OpenAI promises Oracle $300 billion. Oracle's borrowing $38 billion from banks. NVIDIA's financing chip purchases with credits that get returned if unused.
Brandon asked: where's the actual cash?
Revenue shows up on income statements. But free cash flow tells the truth. Oracle's free cash flow went negative as capital expenditures exploded 9x since 2022.
The session revealed something bigger than Oracle. Brandon and Blake walked through similar patterns at NVIDIA, Core Weave, and multiple quantum computing stocks with zero products trading at $11 billion valuations.
Brandon told a story about his Spanish class. The teacher gave out Bueno Bucks as rewards. Students could spend them at the Bueno Bucks store for candy.
The currency worked inside that closed system. But you couldn't take Bueno Bucks to the grocery store and buy real food.
That's AI revenue right now. It works until someone asks for actual dollars instead of more credits.
→ Watch Brandon expose the AI revenue structure and explain why Oracle's debt offering signals a top
Brandon's not predicting an immediate crash. He's showing why the fundamentals don't support current valuations. When markets care about valuations again, these companies face catastrophic repricing.
Blake noted Jeff Bierman's principle: fundamentals don't matter on the way up. They matter on the way down.
We're still on the way up. But understanding the structure tells you what's coming when sentiment shifts.
To your success,
Don Kaufman
Chief Market Strategist, TheoTRADE
Post a Comment
Post a Comment