Bloom Energy Falls Harder Than Its Peers With Earnings Due Tomorrow

Bloom Energy dropped 15.5% on the week to $222.65, a steeper fall than any of its closest correlated peers and one that arrives the day before the company reports Q3 results.

The peer comparison is worth pausing on. NVT fell 13.4% on the week, VRT lost 14.2%, and ETN slid 10.8%. The whole power infrastructure group has been under pressure, but BE's decline exceeded them all. That context matters: the previous two notes in this series flagged that the drawdown from a 172% year-to-date run looked more like sector rotation than a BE-specific breakdown. That read holds for most of the week. BE's short interest, at 8.6% of the free float on roughly 20.4 million shares, is actually up fractionally on the week (about 1%), but remains well below the September peak near 27 million shares. The borrowing market is not pointing to aggressive short-side conviction: availability is at 1,203% of outstanding short interest, and the cost to borrow is 0.44%, neither number suggesting the lending pool is under any strain.

Options positioning adds texture to the picture. Traders are more defensive than usual: the put/call ratio is 1.18, about 1.6 standard deviations above its 20-day average of 1.13. That is elevated but not extreme, especially measured against the 52-week high of 1.43. Demand for downside protection has drifted higher through the month, consistent with investors hedging into a print rather than making a directional bet.

The analyst community has been moving the opposite direction from the stock. Over the past two weeks, UBS raised its target to $350, Baird lifted to $351, and Truist bumped to $240 (maintaining Hold). Barclays raised to $308 in early October. The mean Street target stands at $285, around 28% above Tuesday's close. The divide between bulls and bears is substantive. The bull case centres on a backlog above $20 billion, first-half 2026 revenue of $1.82 billion, and a FERC ruling last June that validates on-site generation as a structural solution for AI data centre power demand. The bear case points to high-capex risk, execution exposure tied to Project Jupiter and Project Tembo, and customer concentration: the top two customers accounted for 39% of 2024 revenue combined. Valuation gives the bears something to work with too. The price-to-earnings multiple has compressed roughly 14 points over the past month to about 49 times, and EV/EBITDA has eased to 41 times, but neither is cheap by any relative measure.

One notable ownership detail sits in the activist register. SK ecoplant Co. filed a Schedule 13D/A last August, confirming a 4.7% stake. A 13D filing signals active intent rather than passive holding, and SK ecoplant has filed three times since July 2025. As the disclosure note requires: stakes are as last disclosed around the 5% threshold, and a holder dropping below 5% may not file again. On the insider side, the net picture over 90 days is modestly negative: around $21.8 million in net selling, led by the Chief Commercial Officer under a 10b5-1 plan in September and October. Plan-driven sales are routine, but the absence of any open-market buying from senior management ahead of a major print is worth noting.

The last two earnings reactions add context without providing a template. The July 30 print produced a 26% one-day gain and a 40% five-day gain. The prior release, on July 28, fell 13% on the day before recovering 21% over five days. The five-day pattern has been strongly positive in both cases, though the day-one reaction has swung sharply in each direction. With the stock already down sharply, the question tomorrow is less about the AI demand narrative and more about whether the company's own cost and margin trajectory is running ahead of or behind what the target increases implied.