#甲骨文 With $300 billion in AI orders, why is the stock plunging?
Oracle $ORCLB once surged 36% in a single day, briefly making its founder, Larry Ellison, the richest person in the world.
The biggest story behind the rally was its AI partnership with OpenAI, worth as much as $300 billion.
Many people thought this was a tech giant’s ultimate comeback in the AI era.
But Oracle’s stock later suffered a sharp pullback.
With orders this big, why is the market getting worried?
Put simply, it comes down to four words: the business model.
Oracle used to make most of its money selling software.
Develop the code, sell it to customers, and the marginal cost stays relatively low. Gross margins can reach around 70%.
That’s a classic asset-light business.
But the AI era is different.
To rent out computing power, you first have to spend big on chips, build data centers, secure enough electricity, and take on enormous financing costs.
It’s a bit like this:
In the past, you sold recipes—print one, earn money from one.
Now you have to borrow money to build an apartment block, then wait for tenants to pay rent over time.
Here’s the biggest problem: before the building has paid for itself, the equipment inside may already be obsolete.
AI chips are replaced rapidly, creating enormous depreciation costs.
If the upfront investment is too large and cash comes in too slowly, debt and interest payments can become a heavy burden.
Even a huge order doesn’t mean the cash is already in the bank.
And impressive revenue doesn’t guarantee a profit in the end.
So when looking at Oracle’s AI story, don’t focus only on the size of its OpenAI orders. Pay attention to three metrics:
1️⃣ Capital expenditures: How much more will it need to invest in computing infrastructure?
2️⃣ Cash flow and debt: Is it earning enough to cover interest payments and future investment?
3️⃣ Return on investment: Will these expensive chips ultimately earn back their cost?
In the AI era, the most valuable company won’t necessarily be the one with the most orders. It will be the one that can turn computing power into profit—and profit into cash flow.
Don’t get distracted by how big the story sounds.
What the market ultimately cares about is how much money you can make—and when that money will actually be in your hands.
Oracle $ORCLB once surged 36% in a single day, briefly making its founder, Larry Ellison, the richest person in the world.
The biggest story behind the rally was its AI partnership with OpenAI, worth as much as $300 billion.
Many people thought this was a tech giant’s ultimate comeback in the AI era.
But Oracle’s stock later suffered a sharp pullback.
With orders this big, why is the market getting worried?
Put simply, it comes down to four words: the business model.
Oracle used to make most of its money selling software.
Develop the code, sell it to customers, and the marginal cost stays relatively low. Gross margins can reach around 70%.
That’s a classic asset-light business.
But the AI era is different.
To rent out computing power, you first have to spend big on chips, build data centers, secure enough electricity, and take on enormous financing costs.
It’s a bit like this:
In the past, you sold recipes—print one, earn money from one.
Now you have to borrow money to build an apartment block, then wait for tenants to pay rent over time.
Here’s the biggest problem: before the building has paid for itself, the equipment inside may already be obsolete.
AI chips are replaced rapidly, creating enormous depreciation costs.
If the upfront investment is too large and cash comes in too slowly, debt and interest payments can become a heavy burden.
Even a huge order doesn’t mean the cash is already in the bank.
And impressive revenue doesn’t guarantee a profit in the end.
So when looking at Oracle’s AI story, don’t focus only on the size of its OpenAI orders. Pay attention to three metrics:
1️⃣ Capital expenditures: How much more will it need to invest in computing infrastructure?
2️⃣ Cash flow and debt: Is it earning enough to cover interest payments and future investment?
3️⃣ Return on investment: Will these expensive chips ultimately earn back their cost?
In the AI era, the most valuable company won’t necessarily be the one with the most orders. It will be the one that can turn computing power into profit—and profit into cash flow.
Don’t get distracted by how big the story sounds.
What the market ultimately cares about is how much money you can make—and when that money will actually be in your hands.