Strong earnings from the world’s biggest tech firms have sparked headlines, yet Michael Burry—renowned for profiting from the 2008 housing collapse—has doubled down on a bearish outlook, warning that the market could suffer a “1987‑type fall.” His latest Substack post links today’s AI‑driven rally to the valuation excesses that helped trigger the 1987 crash, suggesting a similar correction may be imminent.
Burry’s argument rests on three observations. First, the speed of recent stock rallies outpaces any clear return on the AI investments that now dominate corporate capital budgets. Companies such as NVIDIA, which supplies the chips that power large‑language models, have seen valuations soar despite limited evidence that AI will generate sustainable cash flow at current levels. Second, central banks are quietly shifting assets toward gold, a move that historically signals waning confidence in fiat currencies and can precede market stress. Finally, the growing reliance on private‑credit financing for AI‑related capex has strained balance sheets across the sector.
The warning matters because it surfaces a potential disconnect between headline earnings and underlying fundamentals. If AI spending continues to erode free cash flow, firms may struggle to fund ongoing research, data‑center expansion, and talent acquisition without resorting to dilutive equity raises. Google’s recent decision to launch an equity fundraise—its first in decades—illustrates how even cash‑rich giants feel pressure to shore up their balance sheets.
Institutional investors are already feeling the ripple. Large‑cap equity funds that overweight tech have reported higher tracking error as AI‑centric stocks diverge from broader market trends. Meanwhile, hedge funds with short positions on AI‑related equities have seen volatility spike, prompting risk‑off behavior among pension funds and sovereign wealth entities that traditionally seek stable returns. The market reaction could manifest as a rapid rotation from growth‑oriented tech stocks to defensive sectors, a pattern reminiscent of the post‑1987 sell‑off when investors fled high‑beta names.
Beyond the balance sheet, the AI boom raises structural questions about the industry’s supply chain. Ed Zitron, an emerging AI skeptic, notes that OpenAI and Anthropic’s parent company (Claude) together command more than half of data‑center demand. This concentration creates a systemic risk: if either firm faces a profitability crunch, the ripple effects could impair the broader ecosystem, from chip manufacturers like NVIDIA to cloud providers such as Meta and Google.
Real‑world implications are already visible. Companies are beginning to rent compute capacity rather than own it, a shift that reduces capital intensity but increases operating expenses. Meta’s recent partnership with Anthropic to embed advanced language models into its platforms underscores how firms are betting on AI to drive user engagement, even as advertisers grow cautious about brand safety in AI‑generated content. The environmental footprint of massive data centers also fuels public backlash, adding a reputational layer to financial risk.
Historically, market tops have often been preceded by a blend of exuberant earnings, speculative technology narratives, and macro‑policy shifts. The dot‑com bubble of the early 2000s, for instance, saw companies post strong top‑line growth while cash flow lagged, culminating in a sharp correction once investor sentiment turned. Burry’s current stance suggests a similar alignment of factors: inflated AI valuations, tightening credit conditions, and a subtle loss of confidence in the dollar.
While no single indicator can predict a crash, the convergence of these signals warrants close monitoring. Institutional players may adjust exposure, regulators could scrutinize AI‑related disclosures, and investors might demand greater transparency on how AI projects translate into earnings. Whether the market experiences a gradual pullback or a sudden plunge, the underlying dynamics highlighted by Burry point to a pivotal moment for technology‑driven automation and workflow transformation.
