"Fervo Energy's Geothermal Gamble: First Power On, Stock Down 64%"

There is a particular kind of silence that follows a company doing exactly what it promised. On 30 September, Fervo Energy declared commercial operation for the first 33-megawatt block of its Cape Station plant in Beaver County, Utah, a day ahead of its contractual deadline, and the block began earning revenue under its power purchase agreement. The plant draws heat from rock some 8,000 feet down, rock that holds no natural reservoir of hot water, by drilling sideways through it and deliberately cracking it open, using techniques borrowed wholesale from the shale oil industry. Water pumped through the fractures returns at about 190 degrees Celsius, hot enough to drive a turbine day and night, whatever the weather. It is, by any technical measure, the moment enhanced geothermal stopped being a demonstration and started being a power plant.

The market answered by cutting the company's value by nearly two-thirds. Fervo priced its Nasdaq debut in May at $27 a share, jumped to $36 on its first day of trading, and by mid-October was changing hands at just over $9, a record closing low and roughly 64% below its IPO price. The stock market, in other words, is pricing the same project in opposite directions at once: the engineering works, and the business may not.

Both readings have evidence behind them. On the engineering side, the first block synchronized with the grid on 24 September and met its full 33-megawatt output threshold, and the company says the other two blocks of the roughly 100-megawatt first phase are scheduled to follow by 1 January 2027. A second, 400-megawatt phase is already under construction with commercial operation targeted for 2028, which would take Cape Station to about 500 megawatts. Google signed a 396-megawatt purchase agreement for the project on 1 September, with an option to expand its offtake by about 600 megawatts more by June 2030. And the wider sector is quietly winning price arguments it used to lose: in NV Energy's 2026 resource plan, three geothermal offers came in at $105 to $107 per megawatt-hour, below the roughly $114 average for solar-plus-storage projects in the same filing, a gap that matters because geothermal runs around the clock while a solar-plus-battery project in that filing mostly covers four hours of storage.

On the sceptical side, the numbers are stark. Fervo reported just $138,000 of revenue for 2025 against a $57.8 million net loss, and its prospectus puts current development costs at roughly $7,000 per kilowatt, with a long-term target of $3,000. In August the company warned of transmission curtailment at Cape Station, meaning power it can generate but cannot always deliver. And on 6 October, short seller Morpheus Research alleged thermal decline, water loss and seismicity problems at its sites. Those allegations are unproven, and short sellers have their own incentives, but they landed on a stock already falling, and they name exactly the risks that would matter most for an engineered reservoir: that the heat fades, the water escapes, or the ground objects.

What I find most interesting is what this disagreement is actually about. It is not about whether the first block works; it demonstrably does. It is about whether an engineered underground reservoir behaves like a machine, with predictable output over years, or like a geology experiment that degrades in ways no short operating history can rule out. A solar farm's output can be modelled from weather records. An enhanced geothermal reservoir has no weather record, only months of its own behaviour, and every additional block is another bet that the rock will keep cooperating. The US Department of Energy's "Enhanced Geothermal Shot" goal, $45 per megawatt-hour by 2035, assumes the learning curve bends the way it did for shale drilling, where Fervo's latest Cape Station well was finished in 21 days after a 143% improvement in drilling rates. Shale, though, never had to prove that its reservoir would still be there in twenty years. It only had to prove that it could be drained once.

So the question that will settle this argument is not dramatic. It is administrative, almost boring, and it arrives within weeks: do the other two blocks of Phase 1 reach commercial operation by 1 January 2027, taking Cape Station to about 100 megawatts of operating, revenue-earning capacity? If they do, the short thesis loses its best near-term evidence, and the conversation moves to reservoir longevity, where Fervo will have years of operating data to argue with. If they slip, the market's verdict hardens into something closer to a consensus, and the company's $7.2 billion contract backlog starts to look like a list of promises rather than a pipeline. Either way, the next two months will tell us more about enhanced geothermal than the last two years of fundraising did, because a plant that sells power is a fact, and a plant that only promises power is still a story.