AI Investment Frenzy Echoes Dot-Com Warning Signs as Stock Issuance Surges

SpaceX's $60 billion all-stock acquisition of Cursor — the AI coding tool that popularized 'vibe-coding' — is the latest and most striking example of a pattern rattling experienced market watchers: companies are increasingly choosing to pay for deals and expansion with their own expensive shares rather than cash or debt. That choice is a telling signal. Companies only issue stock cheaply when their valuations are inflated, essentially selling overpriced paper to fund acquisitions. Elon Musk's move is a textbook example — using a richly valued SpaceX stock to buy into the AI software space that his own Grok chatbot has failed to crack.

The broader trend is alarming. According to LSEG data, companies are currently financing nearly half of major deal costs with new equity in the current quarter — a level not seen since the dot-com era and the 2020–21 SPAC mania. In those two prior peaks, stock-funded M&A reached two-thirds and 45% respectively, figures that in retrospect marked the top of speculative cycles. Alphabet recently completed a record-breaking secondary share offering, and SpaceX's IPO launched with enormous fanfare. Investor demand for AI exposure is being met, in part, by companies manufacturing and selling new shares into that demand.

The contrast with the pre-2008 financial crisis is instructive. Back then, deal value peaked at $910 billion over the year ending September 2007, but only 19% was funded with stock — because cheap debt was the real bubble. The implosion came from credit markets, not equities, yet stocks were devastated anyway. Today's dynamic is different: equity is the vehicle of excess, not debt, which means the warning sign is flashing specifically in stock valuations.

Standard valuation metrics tell a muddled story. The S&P 500's forward price-to-earnings ratio sits just below 20 — down from 23 times in both 2020 and last year, and below the dot-com peak of 24.5 times — but Wall Street's rosy earnings forecasts, especially for cyclical chip stocks at record margins, make that number unreliable. Corporate financing behavior may be a cleaner signal precisely because it reflects what insiders actually believe about their own stock's value.

The core question hanging over all of this is which of three scenarios plays out for AI spending. Either the opportunity is genuinely vast enough to absorb hundreds of billions and still generate returns; or the opportunity is real but brutal competition will compress margins and destroy value; or — the most troubling possibility — the spending spree is being driven by shareholder enthusiasm rather than underlying economics, and the AI narrative is dramatically overhyped. The dot-com parallel is hard to ignore: then, as now, 'burn rate' was celebrated as proof of ambition, right up until the moment shareholders discovered their money had vanished.