⚠ Margins Are Thin and VolatileModerate threat

Bloom's margins have improved dramatically, but they remain thin by the standard of a real moat and have a long history of volatility, and much of the current level rests on temporary conditions. Mid-30s non-GAAP gross margins on hardware are decent for a manufacturer but modest next to a business with genuine pricing power, and they sit before the heavy operating costs of R&D, selling, and service — so the true profitability, while now positive, is not the fortress the valuation implies. More important, a substantial part of the recent margin gain reflects the premium pricing of the demand surge — customers paying up for scarce, fast power — rather than a structural cost advantage, which means the margins are exposed as that premium normalizes.

The volatility is the historical pattern. Bloom's margins have swung with volume, product mix, service and stack-replacement costs, and one-time charges throughout its life, sometimes sharply, and there is little in that record to suggest the current level is a stable floor rather than a cyclical high. As competition arrives, the grid catches up, and the speed premium erodes, pricing pressure would fall directly on margins in a business that competes, ultimately, to supply a commodity. And if volume becomes lumpier, the operating leverage that lifted margins on the way up would pressure them on the way down. The cost-down progress is real and continued improvement would strengthen the business genuinely. But an investor should not extrapolate the current, partly-peak-inflated margins as durable: they are improved but still thin, historically volatile, and substantially dependent on a premium that the same forces threatening the demand surge will compress — another reason the profitability, real as it now is, remains unproven as a structural feature rather than a favorable moment — fiscal 2025 still closed $88M in the red¹.