
Yeah we’re making more money but we’re gonna have less cash at the end of it dw about it.
Why is this happening?
TLDR: Oracle is spending billions on its AI infra buildout, to satisfy its insane deal with OpenAI. This means HUGE capex investment upfront, assets which the company will depreciate over multiple years. Hence, free cash flow goes down in the early years (‘26 and ‘27), but accounting net profit goes up, per GAAP.
Whether this makes sense or not, and whether these investments will pay off is essentially the crux of the debate in markets right now.
This chart is basically a Rorschach test on whether you think we’re in an AI bubble or not.
Source: Bloomberg
Tool: Excel
Posted by chartr
![Oracle’s Free Cash Flow & Net Profit Are Set To Wildly Diverge, As It Splurges On An Enormous AI Infrastructure Buildout [OC] Oracle’s Free Cash Flow & Net Profit Are Set To Wildly Diverge, As It Splurges On An Enormous AI Infrastructure Buildout [OC]](https://www.byteseu.com/wp-content/uploads/2025/12/bkuvkliewz5g1-1018x1024.png)
18 Comments
Source: Bloomberg
Tool: Excel
For those who don’t live in the finance world, profit and cash flow are not the same thing. Companies with huge capex requirements, or those that are experiencing tremendous growth, always have cash flow issues.
Not surprising that a software company generally doesn’t have huge capex requirements, but now with the advent of AI and the requirement for huge infrastructure to support the energy needs, that’s no longer the case.
“Whether this makes sense or not, and whether these investments will pay off is essentially the crux of the debate in markets right now”
My bet is on no, but that most of the CEO’s of these respective companies don’t really care much. if they crash the economy and their companies, they just get to retire to some mansion in whatever part of the world they want with their billions of dollars.
Oracle is the government; they’ll be fine.
Reddit: capitalism is broken. Businesses do not invest anymore. They only care about short term profits.
Reddit: why are companies spending all this money? This is very broken. What of these investments are not profitable?
This is how it looks like when a company is making a huge investment
Oracle’s Cash On Hand is $11.2 Billion. So this is not exactly an existential bet on their part.
Source: [https://www.macrotrends.net/stocks/charts/ORCL/oracle/cash-on-hand](https://www.macrotrends.net/stocks/charts/ORCL/oracle/cash-on-hand)
I’m more surprised at Oracle’s growth. In the tech field they have a pretty poor name. I can’t think of a circumstance where I’d use Oracle again voluntarily. Fusion and their DB platforms have better modern offers at lower price points. Going all out on AI seems hi-cost and hi-risk.
Thanks, this is some very beautiful data. As you mentioned, nobody makes capital investments that they don’t expect to be profitable – the true risk, of course, is whether the revenue will materialize. This is, for Oracle, a question around OpenAI’s funding capabilities.
Anywho. I’ve worked in investment banking. At the moment, the appetite for an IPO from companies like Anthropic, OpenAI, is very high, and they’re rumored to be planning one next year. Of course, this could change for many reasons, so they should capitalize while they can.
what early years? they would have to replace the depreciated assets to stay competitive, or is there another possibility?
I genuinely hope Oracle gets toppled by the dept it’s incurred.
Not to worry – a 5 pound golden bar with “Oracle” engraved on it as a “gift” to US President will ensure the bailout in case the numbers go negative.
The end of oracle if the whole AI thing doesn’t pan out?
Are they buying their own stock, too?
FCF divergence is the least of it. Oracle’s $18B debt raise to fund this OpenAI capex spree is the real tell. They are betting the entire short-term balance sheet on an AI bubble payoff. High cost, high risk, zero margin for error. That’s not growth; that’s leverage maximalism.
Oracle Cloud Infrastructure is late to the party. As a result, they were able to observe what worked and what didn’t with Azure/GCI/AWS. They made some really good design decisions and it’s arguably now the best cloud. You get higher performance at lower cost. This can be compelling for some use cases.
I don’t know if there is an AI bubble or not, but Oracle seems to be putting their money where their mouth is in the bet that there is not.
Like it or not, Oracle is showing their massive balls.