The AI bears present a compelling case for market caution, with Fred Hickey of The High-Tech Strategist comparing today's datacenter mania to the fiber-optic overbuild that crashed in 2000, calling it 'more dire than the great fiber-optic capacity overbuilds.' According to reports from Investing.com India, the Bank for International Settlements flagged roughly $1.65 trillion in off-balance-sheet obligations held by the largest hyperscalers, exceeding amounts on their books. Sequoia's David Cahn estimates the annual gap between AI infrastructure spending and ecosystem revenue at nearly $600 billion, while Allianz measured the capex-to-revenue divergence at about 46%, well past the 32% that marked the 2001 telecom bust. An MIT study in August suggested that most corporate AI pilots had produced no measurable revenue.
The AI bears are correct about significant market concentration, with the ten largest stocks now making up roughly 43% of the S&P 500, a record and past the 27% peak the index touched at the height of the dot-com boom. As reported by Investing.com India, this represents the most top-heavy market condition in modern history. The equal-weight index has already begun to diverge from the headline benchmark, which is exactly the kind of internal crack that tends to show up before megacap wobbles. Bank of America survey shows 54% of professional managers now calling AI a 'bubble', with Michael Burry watching Nvidia's credit-default swaps widen as the chipmaker turns into banker, landlord, and equity partner to its own customers.
The critical distinction between today's AI buildout and previous overbuilds lies in the funding mechanism. According to the analysis from Investing.com India, roughly two-thirds of the 2026 capex is funded directly from the operating cash flow and equity of Microsoft, Alphabet, Amazon, and Meta, four of the most profitable enterprises ever assembled. The existing borrowing is investment-grade and still a minority of spending, unlike the fiber-optic overbuild where companies had no business borrowing what they borrowed. Microsoft is sitting on approximately $80 billion of Azure orders it cannot fill for lack of electricity, with GPUs idle in inventory waiting on power. More than 60% of the data center capacity planned for 2027 is not yet under construction, indicating potential supply constraints rather than oversupply.
Despite bear concerns about zero revenue, Microsoft's AI business is past a $37 billion run rate, Amazon's AI revenue is growing in the triple digits, and Anthropic went from about $9 billion to a reported $47 billion run rate in a single year. As reported by Investing.com India, Nvidia's forward earnings have climbed so rapidly that its multiple has actually compressed as fundamentals caught up to expectations. The analysis notes that inference now clears roughly 70% gross margins and that the revenue trailing capex is a timing issue, not the zero-payback story the headlines imply. Anthropic's quarterly revenue surged more than 14-fold year-on-year to $11.5 billion in the second quarter of 2026, significantly outpacing the estimated fourfold annual growth in available computing capacity.
The analysis suggests a disciplined approach to AI investment rather than complete avoidance. According to Investing.com India, the self-funders and power bottleneck are part of the trade that looks least like 2000, while the levered edges are the part that looks most like it. The recommended strategy includes favoring the self-funders over the borrowers, spreading exposure across layers of the trade, and always insisting that price paid is backed by existing earnings. The analysis emphasizes position sizing as the one tactic that survives contact with drawdowns, noting that NVIDIA has fallen by 55% or more on five separate occasions since 2000 but has recovered to new highs after each. Fred Hickey himself holds his AI-bear book at roughly 1% of his portfolio in puts, suggesting he treats it as a hedge rather than conviction short, which the analysis recommends as the appropriate posture.
