⚠ The Valuation & the Peak-Demand QuestionHigh threat

The single largest risk in Bloom is its valuation, because the price assumes answers to questions the company has not yet answered. After a stock that has risen more than fourfold¹ from its 52-week low to about $273 and an $80 billion market capitalization, Bloom trades at around twenty-six times sales and — on its brand-new, barely-positive earnings — roughly three hundred and thirty times trailing profit, about seventy-seven times forward. The shares have also been tested: on 8 July 2026 Hunterbrook Media, which disclosed it may profit from a fall, published a report alleging that Bloom's scandium supply runs through China, and the stock fell about 12% that day;² Bloom rejected the report's claims as false and misleading,³ an investor filed a proposed class action,⁴ and on 21 September 2026 Bloom joined the S&P 500.⁵ These are the multiples of a company the market expects to become a large, durable, dominant, fast-growing franchise. Bloom might become that. But it is today a thin-moat maker of a commodity-output product, profitable for only a few quarters, riding a demand surge that may or may not last — and the gap between what it is and what the price assumes is the investment's defining risk.

The valuation is really a bet on the peak-demand question: will the AI-data-center power surge that has transformed Bloom prove durable and defensible, or is it a spike that competition and a catching-up grid will erode? If the demand persists for years and Bloom holds its position, the current price could look reasonable in hindsight, because the addressable opportunity — powering the AI build-out — is genuinely enormous. If the demand cools, or if the grid catches up and competition (gas turbines, batteries, rival fuel cells, eventually small modular nuclear) competes away the speed premium, then the peak earnings the valuation capitalizes never grow into the price, and a thin-moat company reverts to a thin-moat multiple — a long way down from twenty-six times sales. The asymmetry is stark: the upside requires the boom to be durable and Bloom to stay ahead, while the downside requires only that a demand surge behave the way demand surges usually do.

What makes the valuation especially demanding is the quality of what it is applied to. This is not a wide-moat compounder whose durable advantages justify paying up; it is a company whose best moat mechanism (the service annuity) is of unproven margin quality, whose current profitability is peak-tested only, whose advantages (technology, speed) are contested and perishable, and whose demand is concentrated and possibly temporary. Paying a wide-moat, high-growth price for a thin-moat, newly-profitable, narrative-driven business leaves no margin of safety and prices in success as though it were already achieved. The demand is real, the profits are real, and the opportunity is large — the bull case is not foolish. But the valuation demands that nearly everything go right, for a long time, in a business with a thin moat and a short history of profitability, at the peak of a boom. That is the central risk: not that Bloom is a bad company — it is a genuinely improved one — but that its price assumes a durability and dominance it has not earned, and the peak-demand question on which that assumption rests is precisely the one no one can yet answer.