The AI Infrastructure Boom’s Unsustainable Siren Song

The AI infrastructure boom is getting more leveraged, but at what cost? As Nvidia partners with Wall Street firms to mobilize $500 billion of third-party capital, the complex financing underpinning this boom is coming under closer scrutiny. Hyperscalers and their financial backers are turning to bond markets, joint ventures, leases, and other structures to fund an unprecedented infrastructure buildout. This includes about $1 trillion of “uncommenced” lease commitments, which won’t be reflected in financial statements until payments come due.

Goldman Sachs analysts estimate that hyperscalers have combined lease commitments for data centers, R&D facilities, offices, and equipment totaling $1.5 trillion – a 750% increase from five years ago. This scale of spending is staggering, with some experts comparing it to the 19th-century railway construction boom, adjusted for inflation.

The eventual returns from AI infrastructure are far from guaranteed. Nvidia CEO Jensen Huang touts his company’s chips as an “investable infrastructure asset,” but concerns over how much leverage is being taken on by investors are growing. The unravelling of Situational Awareness, after losses on its leveraged equity bets, highlighted the risks of amplifying returns through excessive borrowing.

The AI trade has become a perfect storm of elevated expectations and increased leverage. Sahil Mahtani at Ninety One warns that “high and rising earnings in the years ahead” are the main risk to markets – not leverage itself, but the inflated expectations driving it. He notes that equity concentration is historically high in tech-heavy markets, especially the US, and that a shift from companies buying back shares to issuing more stock could remove a crucial source of support for equity prices.

Some argue that leverage is simply a tool used by hedge funds to boost returns and provide market liquidity, but others warn that it poses a material threat to financial stability. The available evidence does not support treating hedge fund leverage as an inherent systemic risk – but the risks are real nonetheless. JPMorgan CEO Jamie Dimon recently noted that margin debt is “pretty high,” increasing the risk of amplified volatility.

The AI infrastructure boom’s siren song has been irresistible for investors, with promises of exponential returns on massive investments. But the music may soon be coming to an end – and the consequences will be dire if we fail to address the unsustainable leverage driving this boom.

The Leverage Trap

Leverage is a double-edged sword in finance. It can amplify returns and provide market liquidity, but it also increases the risk of amplified volatility and market crashes. As investors take on more debt to fuel their bets on AI infrastructure, they are playing with fire.

A Perfect Storm of Elevated Expectations and Leverage

Elevated earnings expectations drive up valuations, which in turn fuel more borrowing – creating a self-reinforcing cycle that is unsustainable in the long term. This cycle is fueled by investors’ desire for exponential returns on massive investments, but it ignores the risks of amplified volatility and market crashes.

The Risk of Amplified Volatility

Margin debt is “pretty high,” as JPMorgan’s Jamie Dimon warned – increasing the risk of amplified volatility. When heavily weighted stocks fall, market moves can become exaggerated, leading to sharp sell-offs and rallies. This was evident in the case of Situational Awareness.

The Unraveling of AI Expectations

Sahil Mahtani’s warning about inflated expectations driving the AI trade is crucial. If companies begin issuing more stock instead of buying back shares, it could remove a crucial source of support for equity prices – and exacerbate the risks already facing investors in this space.