Reports of SpaceX planning to borrow $40 billion to purchase Nvidia (NASDAQ: NVDA) chips have reignited debate over whether AI infrastructure spending has reached dangerous levels.
The deal structure reportedly involves $10 billion from banks and $30 billion handled by Apollo, carrying a triple-B credit rating that would allow insurers and pension funds to participate.
RSM Chief Economist Joe Brusuelas expressed serious concern, arguing that AI investment is beginning to crowd out both public and private capital across the broader economy.
“This is one of the reasons why rates are going up,” Brusuelas said, warning that the Federal Reserve is “really underestimating what it’s going to take to restore price stability” under current conditions.
Tematica Research CIO Chris Versace framed the spending as a necessary competitive move, noting that SpaceX must ramp capacity to retain clients like Anthropic or risk losing that business entirely.
Yahoo Finance’s Julie Hyman pushed back, pointing out that borrowing costs matter enormously, with the 30-year rate sitting at a 24-year high and negative free cash flow already plaguing many of SpaceX’s competitors.
POLITICO Economics Correspondent Victoria Guida highlighted that corporate borrowing costs are rising faster than Treasury rates, making SpaceX’s triple-B debt an especially interesting test of market appetite for AI-linked credit.
A detail from the Financial Times reporting drew particular attention: a source described a previous SpaceX fundraising memo as a two-page document featuring pictures of outer space and an arrow indicating the company planned to build data centers “somewhere in the universe.”
Brusuelas drew a sharp historical parallel, comparing current AI investment patterns to railroad overbuilding in the 1870s and the dot-com collapse, warning that malinvestment and malfeasance are inevitable at this stage of the cycle.
“There’s going to be some technological unemployment,” Brusuelas said, adding that a market selloff of around 20% is plausible, though he stopped short of predicting a systemic crash comparable to 2007 through 2010.
Hyman noted that existing AI data centers equipped with Nvidia chips are not running at full capacity around the clock, meaning current buildout is running ahead of actual demand rather than responding to it.
Versace pointed to the transition from inference to agentic AI as a key demand driver, arguing that usage patterns will eventually mirror the internet’s exponential data consumption growth over 15 years.
On the question of circular financing, Brusuelas acknowledged that the involvement of banks and debt markets rather than Nvidia directly was a mild relief, though the panel noted Nvidia does maintain a separate financing agreement with Apollo.
Guida flagged that a new task force led by Jay Clayton is focusing on financial market implications of AI, even as the broader administration pushes a no-guardrails posture toward the technology.
Versace singled out PepsiCo earnings as an early warning signal outside the tech sector, arguing that input cost pressures across sugar, cotton, and diesel will reveal whether inflationary pass-through is beginning to crack consumer-facing businesses.
Brusuelas said his own canary in the coal mine is tracking which banks failed to properly hedge interest rate risk, a pattern that historically leads from payment defaults to broader financial stress.
The panel also touched on Tesla (NASDAQ: TSLA), with Brusuelas describing the vehicles as poor performers as physical automobiles, while Hyman acknowledged that users of Full Self-Driving tend to rate the experience positively despite comfort comparisons favoring rivals.
Michael Burry’s recent social media post comparing current market sentiment to the denial stage of grief seen in 2000 and 2008 was cited, though the panel noted he has issued similar warnings repeatedly in recent years without a crash materializing.
The broader discussion reflected a market wrestling with two simultaneous and contradictory truths: that AI is genuinely transformative and productivity-enhancing, and that the investment cycle surrounding it carries the classic hallmarks of a boom destined for an eventual, painful correction.
