◆ What the Market Isn't Pricing In

Bloom Energy is the purest test in this collection of a single question: how much should you pay for a thrilling story attached to a thin moat? Here is a company that lost money for twenty years, makes a commodity output (electricity) with a differentiated but contested technology, and has a genuine but unproven service annuity as its best moat — and it is valued at $80 billion, around twenty-six times sales and, on its first profits ever, some three hundred and thirty times trailing earnings. The market is not pricing what Bloom is; it is pricing what Bloom might become. Whether that is visionary or reckless is the whole debate.

The bull case is that the market is right to look past the history to an enormous future. The AI-data-center build-out has created a power crisis that Bloom is uniquely positioned to help solve, with a real product, a genuine speed advantage, a $20 billion backlog, a $25 billion financing partner¹, and — at last — real profits and expanding margins. If the AI boom sustains demand for years, if the grid stays behind, and if Bloom's growing installed base and service annuity mature into a durable recurring-revenue business, then Bloom could grow into and beyond its valuation, and the current price will look like an early entry into a company powering the defining infrastructure build-out of the age. On this view, the thin moat is thickening in real time — the service annuity growing, the profitability proving out, the scale advantages compounding — and the market is capitalizing a genuine transformation the skeptics are too anchored to the past to see.

The bear case is that the market has mistaken a demand tailwind for a moat and priced a thin-moat company as a wide-moat one. Everything driving Bloom's surge — the grid bottleneck, the speed premium, the power scarcity — is a temporary condition the whole energy industry is racing to relieve; its advantages (technology, speed) are perishable and contested; its best moat (the annuity) has unproven margins; its profitability is peak-tested only; its demand is concentrated in a few powerful customers building their own power strategies; and its history is one of cash burn and dilution. Pay twenty-six times sales for that², and you have priced in durable dominance the company has not earned, with no margin of safety if the surge behaves like surges do. On this view, Bloom is a good business enjoying a great moment, valued as though the moment were permanent and the moat were wide.

What the market may be under-appreciating, in either direction, is how completely the outcome depends on the durability of the AI-power demand — a variable no one can forecast. Unlike a wide-moat compounder whose value rests on advantages that will persist across many scenarios, Bloom's value is a leveraged bet on one: that the AI build-out sustains extraordinary power demand for years and that Bloom stays ahead of a catching-up grid and a crowded field. If that bet is right, the thin moat matters less because the tailwind carries the business a long way; if it is wrong, the thin moat offers no protection and the valuation has nowhere to stand. The honest verdict is that Bloom is a genuinely improved company — real product, real demand, real profits at last — trading at a price that assumes the best of a wide range of possible futures, with a moat too thin to defend the downside if the future disappoints. The technology is real and the story is thrilling. The moat is thin and the price assumes it is wide. Reconciling those two facts is the entire investment decision, and the market, for now, has decided the story is worth more than the moat — a bet that the AI-power boom will make the thinness of the moat not matter, which it may, until the day it does.