Marketdash
Back to News

Goldman Sachs Says AI Infrastructure Boom Isn't Built on Debt—It's Cash-Funded Reality

MarketDash Editorial Team
12 hours ago
Goldman Sachs pushes back on fears of a debt-fueled AI bubble, revealing that 90% of infrastructure spending comes from operating cash flow. Sung Cho says CoreWeave's troubles are isolated tail risks, not systemic red flags.

The hand-wringing over AI infrastructure turning into a debt-fueled bubble? Goldman Sachs thinks you're worried about the wrong thing.

While CoreWeave Inc. (CRWV) has been making headlines with a spectacular nosedive, Goldman Sachs Asset Management is telling investors to take a breath and look at the actual numbers. The broader AI trade, they argue, is built on something much more boring and reliable than sketchy leverage: cold, hard operating cash.

Follow The Money

Sung Cho, co-head of public tech investing at Goldman, sat down with CNBC to address what he sees as a fundamental misunderstanding about how this massive AI buildout is actually being financed. Yes, the market is currently obsessing over "debt depreciation and doubts," but Cho says the funding reality through 2026 tells a completely different story.

Here's the math: The industry needs somewhere between $700 billion and $1 trillion for AI infrastructure over the next couple of years. That's a staggering number, and it's natural to assume companies are borrowing heavily to fund it. But that assumption would be wrong.

"You fear that 90% of that is being funded by debt," Cho explained, before flipping the script. "It's really only 10% that's being funded by debt. 90% is funded by operating cash flows."

In other words, the companies building AI infrastructure are paying for it out of their regular business earnings. These aren't speculative plays leveraged to the hilt—they're profitable giants with money to burn. Cho even pointed out that companies like Meta Platforms Inc. (META) hold credit ratings superior to the U.S. government itself. When Facebook has better credit than Uncle Sam, you're not exactly dealing with shaky ground.

The CoreWeave Problem Isn't Everyone's Problem

So what about CoreWeave's dramatic collapse, which the CNBC host described as a "staggering fall from grace"? And what about the concerns swirling around Oracle Corp. (ORCL)?

Cho's take: These are outliers, not omens. He characterized these situations as "tails" representing maybe 2-3% of the entire debt funding picture. They're edge cases, not evidence of systemic rot.

More importantly, Cho argued these specific companies aren't even dealing with a demand problem. The issue is supply chain bottlenecks that are slowing down their ability to deploy infrastructure. That's a very different problem than "nobody wants what we're selling."

The Leadership Shuffle

Beyond the funding mechanics, Cho highlighted something interesting happening in the perception game. Investor sentiment about who's actually leading in AI keeps shifting around like musical chairs.

Meta was previously seen as the dominant player, followed by a period where OpenAI seemed to hold that crown. But right now? Investors are betting on Alphabet Inc. (GOOG) (GOOGL) and its Gemini model as the frontier leader.

This perception shift has been very good for Google's stock price—the company has gained roughly $1 trillion in market cap over the last three months. But Cho warned that this kind of "model volatility" among the top players is likely to keep happening. Today's leader might not be tomorrow's, and the market will react accordingly.

CoreWeave By The Numbers

Speaking of market reactions, CoreWeave (CRWV) shares dropped 3.94% to $69.50 on Tuesday, capping off a brutal stretch. The stock has declined 23.34% over the last five sessions alone. Still, it's worth noting that CoreWeave is up 78.21% since its listing earlier this year, so perspective matters.

The technical picture isn't pretty, though. The stock maintains a weaker price trend across short, medium, and long-term timeframes, with a poor value ranking to boot.

Goldman's broader message seems clear: Don't confuse the specific troubles of a few leveraged players with the health of the entire AI infrastructure buildout. When 90% of spending comes from operating cash flow and the biggest players have fortress balance sheets, a couple of stumbles in the tail don't mean the whole thing is about to fall apart.

The AI trade might face plenty of challenges ahead, but according to Goldman, a debt crisis isn't one of them.

Back to News

Goldman Sachs Says AI Infrastructure Boom Isn't Built on Debt—It's Cash-Funded Reality

MarketDash Editorial Team
12 hours ago
Goldman Sachs pushes back on fears of a debt-fueled AI bubble, revealing that 90% of infrastructure spending comes from operating cash flow. Sung Cho says CoreWeave's troubles are isolated tail risks, not systemic red flags.

The hand-wringing over AI infrastructure turning into a debt-fueled bubble? Goldman Sachs thinks you're worried about the wrong thing.

While CoreWeave Inc. (CRWV) has been making headlines with a spectacular nosedive, Goldman Sachs Asset Management is telling investors to take a breath and look at the actual numbers. The broader AI trade, they argue, is built on something much more boring and reliable than sketchy leverage: cold, hard operating cash.

Follow The Money

Sung Cho, co-head of public tech investing at Goldman, sat down with CNBC to address what he sees as a fundamental misunderstanding about how this massive AI buildout is actually being financed. Yes, the market is currently obsessing over "debt depreciation and doubts," but Cho says the funding reality through 2026 tells a completely different story.

Here's the math: The industry needs somewhere between $700 billion and $1 trillion for AI infrastructure over the next couple of years. That's a staggering number, and it's natural to assume companies are borrowing heavily to fund it. But that assumption would be wrong.

"You fear that 90% of that is being funded by debt," Cho explained, before flipping the script. "It's really only 10% that's being funded by debt. 90% is funded by operating cash flows."

In other words, the companies building AI infrastructure are paying for it out of their regular business earnings. These aren't speculative plays leveraged to the hilt—they're profitable giants with money to burn. Cho even pointed out that companies like Meta Platforms Inc. (META) hold credit ratings superior to the U.S. government itself. When Facebook has better credit than Uncle Sam, you're not exactly dealing with shaky ground.

The CoreWeave Problem Isn't Everyone's Problem

So what about CoreWeave's dramatic collapse, which the CNBC host described as a "staggering fall from grace"? And what about the concerns swirling around Oracle Corp. (ORCL)?

Cho's take: These are outliers, not omens. He characterized these situations as "tails" representing maybe 2-3% of the entire debt funding picture. They're edge cases, not evidence of systemic rot.

More importantly, Cho argued these specific companies aren't even dealing with a demand problem. The issue is supply chain bottlenecks that are slowing down their ability to deploy infrastructure. That's a very different problem than "nobody wants what we're selling."

The Leadership Shuffle

Beyond the funding mechanics, Cho highlighted something interesting happening in the perception game. Investor sentiment about who's actually leading in AI keeps shifting around like musical chairs.

Meta was previously seen as the dominant player, followed by a period where OpenAI seemed to hold that crown. But right now? Investors are betting on Alphabet Inc. (GOOG) (GOOGL) and its Gemini model as the frontier leader.

This perception shift has been very good for Google's stock price—the company has gained roughly $1 trillion in market cap over the last three months. But Cho warned that this kind of "model volatility" among the top players is likely to keep happening. Today's leader might not be tomorrow's, and the market will react accordingly.

CoreWeave By The Numbers

Speaking of market reactions, CoreWeave (CRWV) shares dropped 3.94% to $69.50 on Tuesday, capping off a brutal stretch. The stock has declined 23.34% over the last five sessions alone. Still, it's worth noting that CoreWeave is up 78.21% since its listing earlier this year, so perspective matters.

The technical picture isn't pretty, though. The stock maintains a weaker price trend across short, medium, and long-term timeframes, with a poor value ranking to boot.

Goldman's broader message seems clear: Don't confuse the specific troubles of a few leveraged players with the health of the entire AI infrastructure buildout. When 90% of spending comes from operating cash flow and the biggest players have fortress balance sheets, a couple of stumbles in the tail don't mean the whole thing is about to fall apart.

The AI trade might face plenty of challenges ahead, but according to Goldman, a debt crisis isn't one of them.