Blood in the Water

Little room for error... Blood in the water... Insiders are dumping CoreWeave... The WeWork of AI... How every bubble ends...

We can't tell you when, or for certain...

Maybe the AI boom (or bubble, depending on your view) won't turn into a bust that leaves financial carnage in unsuspecting people's brokerage accounts.

But if it does, I (Corey McLaughlin) am willing to bet on a few names being at the center of a shakeout...

First, there's OpenAI, which set off the entire AI boom in early 2023 with the release of its ChatGPT-4. Millions of people use the platform and others like it. It has proved that machines are capable of much human-like activity, quicker than previously thought.

No doubt large language models ("LLMs") like ChatGPT are impressive. And we're already seeing real-world applications of such AI tools, some of which are already replacing tens of thousands of jobs in the U.S.

One example: I ordered food from an AI bot at a fast-food drive-through over the weekend, paid a human at the next window, and all went smoothly. The next day, at two other retailers, I was delayed at checkouts while dealing with human error.

But today, we're less concerned with AI's engineering prowess than we are about the hype that has been building around the tech. That's what has driven up a lot of related stock prices along the way.

When these prices become completely untethered from reality, we have ourselves a good old-fashioned bubble. Maybe we're there already, and maybe there's still road ahead. These things can go on longer than you think makes sense...

But as time has gone on and AI expectations have reached world-changing levels, AI companies now have little room for error with their earnings. We've seen several punished by the market lately, even with good results on the surface. Today marks the latest casualty...

CoreWeave goes down...

The data-center operator is a "neocloud" company for AI. CoreWeave (CRWV) reported third-quarter results after markets closed yesterday. On the surface, its business is thriving...

Revenue more than doubled to $1.36 billion – beating Wall Street's expectations – and the company's backlog surged past $55 billion. And during the most recent quarter, it announced a $14 billion deal with Meta Platforms (META), a $6 billion deal with OpenAI, and a $6.3 billion "collaboration" with Nvidia (NVDA).

But like we saw with other AI earnings this month, those numbers weren't good enough for investors who couldn't look past some flaws. CoreWeave shares fell more than 16% today and are down roughly 50% from its post IPO all-time high in June.

The sour sentiment spilled across the AI ecosystem today... It weighed on semiconductor stocks like Nvidia, Advanced Micro Devices (AMD), and Micron Technology (MU), and it kept a lid on most of the major U.S. stock indexes.

The Nasdaq Composite Index was down slightly, the small-cap Russell 2000 Index was little changed, and the benchmark S&P 500 was up a little, though the Dow Jones Industrial Average gained more than 1% to close at a record high.

Today's action wasn't for nothing... CoreWeave reported its fourth straight quarter with a net loss, and its operating expenses tripled to more than $1.3 billion. The company also raised another $1.75 billion in debt to fund its expansion.

That brings its total debt load to about $14 billion – against no earnings and only $1.8 billion in cash.

Moreover, CoreWeave lowered its full-year guidance. CEO Mike Intrator made the media rounds today to explain why that was – a delay he pinned to a "singular data-center provider," he said on CNBC.

Evidence suggests that provider is Core Scientific (CORZ), whose shares were down roughly 10% today. While Intrator suggested the delay had to do with just one of CoreWeave's 41 data centers, the market reacted like there was blood in the water.

Here's the point: All the revenue and backlog growth is great... but it looks like we're getting to the point in the AI boom where investors want to see how all this investment is going to turn into profits, even if we're still in the period of heavy investment.

The company was still optimistic. Intrator also said today...

CoreWeave's position as the essential cloud for AI has never been stronger as we drive growth through focus and innovation to power the next generation of AI.

Insiders are selling, though...

Since its IPO in March, CoreWeave insiders have sold $4.8 billion worth of shares on the open market, according to financial data firm Finviz.

Those sales have been led in recent weeks by Intrator himself offloading $38 million and his chief strategy officer selling about $120 million – all since October 1.

All in all, insiders at the company have not made one open-market purchase of CoreWeave shares since the IPO. And that comes even when the stock had fallen more than 50% from its post-IPO high (and it still sits about 40% below that level today).

We know insiders can sell shares for a variety of reasons, but the lack of buys even after this dip in the stock is a warning signal...

If the company is doing as well as it said in its earnings release (the "essential cloud for AI"), we'd expect insiders – who know the most about the company – to be buying the stock hand over fist.

They're not, though, and for good reason too...

CoreWeave shares get a "D" grade in both capital efficiency and valuation on our proprietary Stansberry Score. While the stock is still too new to get a complete rating, the early indications aren't in CoreWeave's favor.

We warned of CoreWeave all the way back in March...

In the March 31 Digest, Stansberry Research senior analyst Gabe Marshank gave his thoughts on the company's then-recent public debut. From Gabe back in March...

You may not have heard of CoreWeave, but you have certainly heard of its industry. It's a large data-center operator, housing the chips that host artificial intelligence ("AI"). The big expectations reflected its big ambitions: The company had aimed to raise $2.7 billion at a $32 billion valuation. It ended up raising $1.5 billion.

As Gabe noted, that IPO was "saved" by a last-minute $250 million investment from Nvidia – a key supplier to CoreWeave. The deal highlighted the "circular financing" in the AI space, a web that has gotten much more tangled since.

For example, now we know CoreWeave is building out its data centers with chips from Nvidia, which has invested $100 billion into OpenAI. And we know OpenAI is a privately held company that has agreed to roughly $1 trillion in AI partnerships overall in the years ahead, even though its CEO Sam Altman doesn't expect it to be profitable until 2029.

Back in March, Gabe broke down why CoreWeave had looked to go public, as more and more investors were getting giddy with AI. Put simply, CoreWeave needs more cash to fund its lofty growth goals. More from that Digest...

Today, CoreWeave has $8 billion in debt and posted a $900 million net loss last year – even as revenue jumped more than sevenfold to $1.9 billion. The company needs more cash to grow, so it's turning to the public markets.

Eight months later, and the worries haven't gone away. Gabe, a 20-plus-year Wall Street veteran, reiterated his concerns about CoreWeave last month in a special report for his new Market Maven newsletter.

He compared CoreWeave's model of renting out its data-center assets with that of another once-hyped business, office-sharing company WeWork... which filed for bankruptcy in 2023. Gabe wrote...

It's just like WeWork. CoreWeave is buying an asset, breaking it up into smaller assets, and renting it out. But the only reason it's generating revenues and taking market share is because it's doing it too cheaply.

CoreWeave will likely need to either raise prices for its AI cloud infrastructure and lose customers... or keep prices low and continue losing money. Even after its earnings dip today, CoreWeave is a stock to stay away from.

As for navigating the rest of the bubble – beyond today...

Stansberry Alliance members and Market Maven subscribers also now have access to another brand-new special report from Gabe: "A Veteran's Guide to Bubble Investing." It's a must-read. He starts here...

There are two quotes from famous investors about how to invest in bubbles... and they tell different stories.

George Soros – a multibillionaire hedge-fund manager who famously "broke the Bank of England" by shorting the pound sterling – once said, "When I see a bubble forming, I rush in to buy, adding fuel to the fire."

Warren Buffett – the value investor dubbed the "Oracle of Omaha" and head of Berkshire Hathaway (BRK-B) – once said, "When the tide goes out, you see who is swimming naked."

Taken together, these quotes tell us the truth about investing in a bubble: They will make you look stupid either before it pops... or after.

The truth is, nobody will ever perfectly time trading around a bubble, so it's pointless to even try...

But there are things you can do to "invest in bubble times" besides aiming to perfectly time the upside and the downside.

Gabe breaks down everything you need to know about navigating a bubble, in AI or anything else. He describes two types of bubbles, three lessons learned from past episodes, and where the biggest losers and winners are likely to emerge amid today's frenzy.

And here's his take on AI today: "Make no mistake – this is a bubble."

Stansberry Alliance members and Market Maven subscribers can and should read Gabe's special report here. We've already received some glowing feedback from subscribers about it.

New 52-week highs (as of 11/10/25): Altius Minerals (ALS.TO), Atmus Filtration Technologies (ATMU), Alpha Architect 1-3 Month Box Fund (BOXX), BP (BP), CBOE Global Markets (CBOE), Ciena (CIEN), Cencora (COR), Donaldson (DCI), Enel (ENLAY), EnerSys (ENS), iShares MSCI Italy Fund (EWI), iShares MSCI Spain Fund (EWP), Expeditors International of Washington (EXPD), Cambria Emerging Shareholder Yield Fund (EYLD), Alphabet (GOOGL), iShares Biotechnology Fund (IBB), JPMorgan Chase (JPM), Lumentum (LITE), Markel (MKL), Nuveen California Quality Municipal Income Fund (NAC), Omega Healthcare Investors (OHI), Roivant Sciences (ROIV), SandRidge Energy (SD), Sprott (SII), Tenaris (TS), United States Commodity Index Fund (USCI), and Valero Energy (VLO).

In today's mailbag, feedback on yesterday's Digest, which included a mention of President Donald Trump's idea of 50-year mortgages in the U.S... Do you have a comment or question? As always, e-mail us at feedback@stansberryresearch.com.

"There's so much to unpack from [yesterday's] journal that it's hard to know where to start, but I'll go with the 50-year mortgages, a truly terrible idea. Mr. Pulte should remember (but perhaps does not) that 40-year mortgages were a thing for a while, but faded into oblivion for good reason. We absolutely do not need 50-year mortgages now. Why stop there? Why not 100-year mortgages?" – Subscriber Sherwin R.

All the best,

Corey McLaughlin and Nick Koziol
Baltimore, Maryland
November 11, 2025