Morgan Stanley wants you to believe AI power demand is about to go into hyperdrive, and that Bloom Energy and “many other stocks” are positioned as big winners. That’s the entire thesis on offer here. No numbers, no analysis of fuel costs, no discussion of what these systems actually cost to install and run. Just a bank’s research note and a rising tide narrative.
I’ve watched sell-side shops run this playbook for 35 years. Find a megatrend, attach a basket of tickers to it, and let the momentum crowd do the rest. Sometimes the trend is real—the underlying demand growth usually is. Whether the named stocks convert that demand into actual cash flow is an entirely different question, and it’s the one nobody answers in a one-line call.
Take Bloom. Solid oxide fuel cells are legitimate technology with real installations. But the economics hinge on natural gas prices, installation costs, and whether hyperscalers will actually sign the kind of long-term contracts that justify the hardware. Data center operators have plenty of options—grid upgrades, behind-the-meter gas turbines, renewables with storage—and they’re notorious for squeezing vendors. Being attached to a demand wave doesn’t mean you get paid like you own the wave.
The honest read: AI is genuinely reshaping electricity demand, and power generation equipment names will see order flow. Some will make real money. Many will see their valuations run far ahead of their earnings, which is where investors get hurt. The demand forecast is the easy part. Picking the company that actually keeps margin through the cycle is the whole game.
Buy the trend if you must, but size the position like a banker selling you the thesis—because that’s exactly what’s happening.
