Economist Steve Hanke has issued a stark warning about the AI market entering bubble territory, drawing direct parallels to the dot-com bubble as massive capital flows drive unprecedented valuations. According to reports from Milk Road, Hanke's bubble detector shows clear bubble territory, with the warning supported by the bond-stock yield spread as a second confirmation signal. The Johns Hopkins applied economics professor, who previously served as a senior economist for President Ronald Reagan, has flagged similar overvaluation concerns through 2025 and 2026. Hanke's core concern centers on a disconnect between massive infrastructure spending and immediate commercial returns, warning that the market's exuberance hinges on whether AI firms' spectacular revenue forecasts hold water.
The AI boom has generated $10 trillion in capital flow with the combined market cap of top AI-driven firms surpassing $10 trillion, led by NVIDIA and Nvidia at around $4.8 trillion alone. This massive valuation is backed by concrete corporate spending, with firms like Alphabet, Amazon, Microsoft, and Meta investing hundreds of billions into AI infrastructure. The flow creates a self-reinforcing cycle where corporate spending funds infrastructure buildout, validates chipmakers' capacity and margins, and supports soaring valuations across the entire ecosystem. As reported by Milk Road, the Magnificent Seven traded sideways for months before this recent leg higher, with the ultimate test being whether the $10 trillion flow translates into promised productivity gains and profits.
TSMC's valuation surge past $1.65 trillion in January serves as critical validation for the AI sector, making it the sixth most valuable company globally. The company's record-breaking 2025 results, driven by 2nm production and massive capacity expansion, provide tangible proof of the AI buildout's scale and longevity. However, Hanke warns this creates a 'Capacity Bubble' model where historic capital expenditures like $520 billion forecast by Meta, Amazon, and Alphabet for 2026 were not yet tied to current utility or profit. The issue may not be stock prices being inflated, but rather the utility of the product being questioned, as the infrastructure buildout is ahead of commercial payoff, creating a dangerous mismatch between investment and returns.
Market analysts attribute the AI-driven rally to strong earnings and heavy AI capital expenditure, with the next major catalyst being TSMC's investor conference scheduled for January 15 to set the tone for the 2nm supercycle. The market's exuberance will be deemed irrational if AI companies fall short of aggressive growth forecasts, with Hanke predicting a sharp repricing is likely if commercial reality fails to match the $10 trillion flow. The primary macroeconomic risk identified by Hanke is inflation, pointing to a $3.5 trillion increase in M2 money supply over the past five years as a key metric that could undermine Fed control. Whether Hanke's warning proves prescient will depend on whether the infrastructure buildout yields promised returns, as the bubble's fate hinges on whether spectacular revenue projections hold water.
