Ken Griffin, founder and CEO of Citadel, declared at his firm's Future of Global Markets Conference on October 20 that investors are "almost certainly" in an AI bubble, drawing a direct parallel to the dot-com era and the telecom fiber-optic overbuild of 2000. It’s quite a hot take after Citadel's own amended second-quarter SEC filing revealed approximately $8.6 billion invested across five major AI companies, with Amazon (AMZN) and Nvidia (NVDA) ranking as the firm's second and third largest individual stock holdings respectively.
Just two days after Griffin's warning, Blackstone (BX) President and COO Jon Gray countered during a CNBC interview that investors should not fear the AI boom turning into a bubble, expressing confidence in the underlying strength of demand for computing power.
More Top Stocks Daily: Go behind Wall Street’s hottest headlines with Barchart’s Active Investor newsletter.
As with Griffin, Gray has a tremendous stake in the AI buildout, to the tune of about $200 billion in data centers alone.
Griffin's Three Vulnerabilities, and a $193 Billion Bond Binge
Griffin's concerns center on three specific vulnerabilities: GPU supply outstripping demand; the questionable useful economic life of AI chips that still command extraordinary rental rates five or six years on; and corporate America's highly uneven track record of extracting genuine productivity gains from AI spending.
The scale of capital pouring into AI infrastructure underscores the worry, as Amazon, CoreWeave, Google (GOOG) (GOOGL), Meta Platforms (META), Microsoft, and Oracle (ORCL) had issued approximately $193 billion in bonds by early August to finance AI-related infrastructure — a staggering leap from $16.7 billion in all of 2024.
UBS projects hyperscalers will spend $4.1 trillion on AI infrastructure between 2026 and 2028, nearly triple the preceding six years, while Anthropic's planned spending alone reportedly approaches $750 billion in additional computing capacity.
Gray's Counterweight: Pricing Power and Constrained Supply
Gray's counterargument rests on the observation that demand for computing power remains robust across Blackstone's private portfolio, while supply delivery stays constrained.
He noted that for neocloud companies like CoreWeave (CRWV), compute pricing has risen 50%, legacy chips are worth more than anticipated, and data center leases sit below market rates — a notable but underappreciated signal that the supply-demand imbalance still favors builders.
Gray acknowledged financing challenges facing CoreWeave, which carries roughly $51.6 billion in total debt, but expressed confidence that well-managed firms will navigate the pressure.
Again, it’s worth pointing out that the Blackstone exec is largely talking his book when he shares his bullish view on AI.
“On a call with analysts Thursday, the firm said that AI-related holdings represent about 7% of its roughly $1.4 trillion of assets under management,” reported Bloomberg of the asset manager’s quarterly results.
Citadel's Portfolio Tells a Different Story
Griffin's own behavior complicates his bubble warning in a striking way.
Citadel increased its Amazon stake by 34% and its Microsoft (MSFT) position by 43% during the second quarter, even while reducing Nvidia shares by 37% — a pattern suggesting the firm sees selective opportunity rather than indiscriminate froth.
Furthermore, Citadel profited handsomely when AI hedge fund Situational Awareness was forced to liquidate roughly $35 billion in positions at a steep discount in July, with Citadel subsequently shedding more than 80% of the acquired risk.
The Semiconductor Selloff Adds Urgency
The market digested these competing narratives in real time on October 22, as the Nasdaq Composite ($NASX) fell more than 2% after Meta's earnings report triggered concerns about massive capital spending plans and new non-voting share issuance.
The Philadelphia Semiconductor Index ($SOX) dropped 6.7% to its lowest level since late July, shedding more than $616 billion in market value, while Blackstone (BX) itself rallied 8% after its own earnings beat.
The Next Catalyst to Watch
The tension between Griffin's historically informed caution and Gray's conviction in constrained supply will face its most important near-term test when Nvidia reports earnings on November 17, with guided revenue of $108 billion representing the single clearest verdict on whether AI spending is producing returns commensurate with the unprecedented capital deployed.
This article was created with the support of automated content tools from our partners at Sigma.AI. Together, our financial data and AI solutions help us to deliver more informed market headline analysis to readers faster than ever.
On the date of publication, Sarah Holzmann did not have (either directly or indirectly) positions in any of the securities mentioned in this article. All information and data in this article is solely for informational purposes. For more information please view the Barchart Disclosure Policy here.
More news from Barchart
Tech Stocks Are Crashing. Don’t Make the Mistake of Selling All of Them ‘What a Conservative Is Today Is Far Different Than It Was 11 Years Ago,’ Says Mark Cuban. If It Were Up to Him, He’d ‘Get Rid of All Political Parties.’ Citadel's Ken Griffin Says AI is in a Bubble Right Now. Billionaire Data Center Backer Jon Gray of Blackstone Disagrees. Meta Approves New Non-Voting Stock Class Worth Billions to Zuckerberg — But the Real Price Was an 8-Year Residency Cage