GNG Research GNG Research

My 2026 Top 10 AI Power & Utilities Refresh

Core thesis validated - AI-driven data center demand intensified physical bottlenecks (transformers, substations, interconnection queues); AEP ~138, WEC ~119, ETR ~114, CEG fell to ~251 from a $412 high Scarcity failed to translate to market prices where regulators intervene - PJM capacity ran $28.…

Published: 2026-07-22 by GNG Research

Tickers: CEG, GEV, NEE, AEP

In December I published 2026 Top 10 AI Power & Utilities: The Real AI Trade Nobody's Talking About . The argument was that everyone was staring at chips and models while the actual constraint sat one layer down, in the substations and the interconnection queues and the transformer order books. Seven months later, let me show you the scoreboard before I show you anything else. You are entitled to that. The boring part worked. (AEP) was around $118 with a $130 to $140 target, and it trades near $138. (WEC) was around $100 against a $115 to $125 target, and it is near $119. (ETR) was around $89 against a $95 to $105 target, and it blew straight through to roughly $114. The exciting part did not. (CEG) was around $361 when I gave it WATCH status and wrote that if it traded above $400 without an earnings beat, do not add. It touched $412. It is around $251 today. And the name I told you was my largest utility weight, (NEE), returned 8.88 percent against the S&P 500's 11.84 percent over the same 140 trading days. It underperformed the index I was asking you to beat. So the defensive core did its job, the one name I flagged as dangerous fell 39 percent from its high, and my biggest position lagged. That is a mixed report card, and the pattern inside it is the reason this refresh exists. What did I actually get wrong?

This is a members-only GNG Research article. Read the full analysis with a GNG Research plan.

More GNG Research articles