Over the last months, I have written two Weeklies, criticizing the from my view not sustainable AI investment frenzy, featuring Meta’s stock as my main target. Maybe this was because Meta is a mainstream darling that on top has delivered a stellar return over the last three years. However, little did I know that there is an even more interesting case in the second row. With some interesting developments in the recent past, is this the Oracle in the AI coal mine for what to expect?
Summary and key takeaways from today’s Weekly
– I’d be very cautious now with buying any AI-related dips.
– The case of Oracle has become an all-in vabanque game.
– Risks are very high now.
For my newer readers, these are the two Weeklies you should know about (see here and here).
As often, over-exaggerations tend to last longer than one can imagine.
While skeptical voices like mine are not new to this topic, it looks like we reached now a stage were the one-directional ride might have come to and end. Popular AI stocks have entered a correction. This is nothing unusual so far. Until recently, every dip got bought and the climb higher continued as if nothing happened.
But indeed, much happened.
We all know well about the ever rising and rising investment budgets of the big AI players. I showed that serious cracks have started to appear on the surface of Meta.
However, this still compares favorably to what I am going to discuss today.
The Oracle of a disaster to happen. Potentially for good.
If you’re looking for attractive, overlooked stock ideas the majority has a blind eye on, look no further. I am offering compelling non-mainstream cases in concise 12-page member-exclusive reports.
Want to outsmart the market? Get active now! Become a supporting paid-member and receive my best stock ideas (plus updates).
The average total return of my best stock ideas is ahead of the S&P500 and the Dow Jones. With my risks-first approach (paired with high upside), I am able to find stocks with great returns.
Join me and my members on our journey to beat the markets!
both as per 19 November 2025 market close – since August 2022
The higher the flame, the shorter the burn
My long-time readers know that I like puns and jokes.
Referring to my headline, little surprisingly, today’s main topic is an AI play from the second row, Oracle (ISIN: US68389X1054, ticker: ORCL). While the company is well known as such, for long it has not been associated as much with AI like the Mag7.
But Oracle has hopped on the frenzy. This is no joke.
And how it did.
While most everyday users likely do not have direct contact or access to Oracle (or are not aware) unlike to Grok or ChatGPT, especially the last few weeks have blown my mind – in a negative way.
When you asked me about the more significant headlines concerning Oracle over the last few weeks and months, I’d say the following:
and:
The chronological order of course is the other way around, as Oracle first reported about the signed 300 bn. USD deal with OpenAI alongside Oracle’s quarterly results.
Oracle revealed the deal on 09 September 2025 (after the market close). The news sent Oracle’s stock up an unprecedented over 40%. As Oracle was and is no small-cap, this move made huge waves, with the two results pictured above.
The volume of 131 mn. shares was massive. Back the envelope, total trading volume in USD was almost 40 bn. (assuming 300 USD per share).
In one day.
This event not only fueled discussions about an AI bubble much more intensely, but it might have been the final nail in the coffin from today’s perspective. The final exaggeration or blow-off.
At least, when looking at Oracle’s stock.
The deal’s scale raised questions about OpenAI’s ability to fund it given its current revenue and lack of profitability.
In a nutshell, Oracle agreed to sell OpenAI up to 300 bn. USD in cloud computing capacity over many years – OpenAI currently generates only about 13 bn. USD in annual recurring revenue (ARR) as of late October 2025 (see here CEO Sam Altman on 06 November 2025). OpenAI seems to be on track to exceed 20 bn. USD in annualized revenue for the full year 2025.
But this is still peanuts.
The world’s currently most valuable private company would need explosive, unprecedented growth (or massive outside funding) to actually pay for that amount without going bust.
Imagination might not have limits, but certain real-world balance sheets do.
Speaking of balance sheets, Oracle once was known for strong margins and solid finances, frequently buying back its own stock.
We can see that sales are well on their way up, with some more dynamic over the last years, coinciding with the emergence of AI, starting in late 2022. However, operating earnings, while clearly on the rise, too, could not keep pace with sales growth.
Looking at operating margins, we can see they are falling. A plausible explanation is quickly found – they are sacrificing short-term profitability for long-term gains.
At least, that’s what we are told or made to believe.
Sounds plausible, yes.
But why are gross margins tanking so much, then?
Short answer: the same shenanigans we know from Meta.
Oracle is engaging in significantly lower-margin activities that better reflect the true dynamics. A level below, concerning operating margins, they know how to polish that for a better look.
Michael Burry as of writing a week ago has posted the following. He shows, practically all the major AI infrastructure players have bumped up their depreciation times, extending the estimated useful lives of their equipment.
As a result, Oracle boosted its earnings by 27% through accounting gimmicks.
Quickly checking Oracle’s annual report, Burry even used a too-conservative number in Oracle’s case. Estimated useful lives jumped from four to five, then to six years, not just from five to six.
50% higher estimated useful life of quickly depreciating hardware in a race where no one wants to be the first loser.
I’d like to remind my readers what I wrote and showed about amazon in this context. They have REDUCED their estimated useful lives of their assets (per 2024 annual report) and even scrapped some hardware prematurely due to intensifying competition and quicker-than-thought obsolete hardware.
You be the judge who’s right.
Like its peers, Oracle is heavily investing now.
But there’s a catch.
While the Mag7 companies still have relatively robust balance sheets, we cannot say that about Oracle. Financial debt went up to more than 90 bn. USD (before leases) and cash drifted lower from almost 70 bn. USD at the peak to only 11 bn. USD per last count (purple line).
That’s quite a shift.
But it does not seem we’re close to the finish line.
This makes Oracle an extremely leveraged do-or-die play on AI. I am sure many people will see it differently, maybe referring to the successful past of the company. But numbers often tend to speak for themselves.
With the massively worsened balance sheet, the market started to ask a few questions.
source: twitter, see here
Oracle’s bonds suddenly dropped by 7%.
At the same time, so called credit default swaps – insurance derivatives – went through the roof. Purple is the stock price, white is the CDS.
Unfortunately, this is not exclusive to Oracle…
The turn in the narrative about Oracle is not only due to its noticeably higher debt in absolute terms. It is also about relative terms.
Free cash flow is currently negative, even before the 5 bn. USD dividend it is still paying.
With negative free cash flow and the unnecessary dividend (except for the founder who owns 40% of equity), this has turned into a dangerous vabanque game.
At first glance, Oracle seems to have spread their debt maturities way into the future which of course would be great.
But I am talking in conditional, as first of all there’s plenty of low-cost debt maturing every year, likely to increase interest payments. If they want to invest, they need the money, making principal repayments unlikely.
Debt will need to be rolled, increasing cost of debt.
The other things is, with a debt downgrade and exploding CDS, it is unlikely Oracle will have an easy hand to issue much more debt at will.
I would be very cautious here.
Two more things to close this chapter. The first is about the proclaimed 300 bn. USD OpenAI deal that suddenly seems to have been re-valued.
And the other thing is, while I myself do not know how to assess it, the news from China regarding AI are not necessarily supportive for the AI Capex theme.
Even if only half of it were to be true.
I don’t know how the proclaimed and prolonged estimated useful lives of these core assets will be held up plausibly. I learnt that accounting should be conservative and cautious.
But this is extremely aggressive, almost ridiculing those who read and believe it.
In light of all this, isn’t it surprising to see what I wrote in my first AI Weekly happening? The often as-failure criticized Apple (ISIN: US0378331005, ticker: AAPL) sees its stock hovering around its all-time high. Despite the current market correction and despite the suddenly come-to-the-surface AI bubble fears, Apple does not care.
As I said, the winner might be the one who does the least stupid things.
This doesn’t mean I’d buy AAPL stock now. It’s too expensive.
There are clearly much more attractive opportunities out there. My Premium and Premium PLUS members frequently receive my best ideas, circumventing mainstream themes like the above.
And it’s worth it. Especially for downside protection.
Even if it sounds contradictory in the first moment, it is not. We need to take care of the downside, minimizing risks. Many people who are sitting on gains and who until recently were celebrating big wins often forget one thing:
It is not about paper gains, but about realizing them.
My members and I do that. My closed cases are ahead of the market.
Conclusion
I’d be very cautious now with buying any AI-related dips.
The case of Oracle has become an all-in vabanque game.
Risks are very high now.
By becoming a Premium or Premium PLUS Member, you get instant access to all my already published research reports as well as several updates.
Likewise, you qualify for eight, respectively four more exclusive reports with my best investment ideas plus updates on the featured businesses over the next twelve months.