AI data centers are signing up power companies while one mega-utility deal hits a political wall — long Black Hills, hedge with NextEra

Verdict: A dated catalyst, an unconfirmed setup — wait on BKH

**Verdict: Wait — a real catalyst, an unconfirmed setup, and a thin edge.** The strongest point for this idea is that its core tension is concrete and dated: per Reuters on October 6, 2026, Black Hills committed $1.8 billion to power a Google data center, while Bloomberg reported the same day that Virginia's Lieutenant Governor opposes NextEra's $67 billion pursuit of Dominion. The strongest point against is that the completed 12-month backtest produced only five trades, a 40% win rate, and roughly +1.1% total return against a ~5% maximum drawdown — a thin, statistically noisy margin — and the fundamentals cut against the fade: BKH's Q2 free cash flow loss narrowed from -$91.2M to -$23.0M and Dominion just posted its best quarter of the group with revenue up 20.9% to $5.1B. Right now the setup itself is not live: BKH closed at $70.74, its ADX (14) at 14.9 sits 5.1 points below the required 20, and the 9-day average at $70.56 still trails the 50-day at $71.87 by about $1.31, so the short cannot trigger today. What would flip this verdict is a confirmed entry (ADX through 20 with price above $70.81) alongside a fundamental catalyst on the fade side — a guidance cut or a delay in the $1.8B data-center buildout — or conversely, any sign the NextEra-Dominion deal revives, which would lift the short leg and kill the thesis.

Trade now

This is a short-reclaim strategy on BKH, and right now you wait. BKH closed at $70.74, essentially sitting on its short-term momentum line (the 9-day average at $70.56, just $0.18 away) but still below the middle of its 20-day band at $70.81 by about $0.07. The RSI (14) is 50.4, comfortably inside the required 45-65 window, so the momentum filter is live. What is not live: the 9-day average has not crossed above the 50-day average (it trails by about $1.31), and ADX (14) at 14.9 is still well below the required 20 — the single biggest missing condition. Until price closes back above the 9-day line, clears the band middle, and the short-term average flips above the 50-day while ADX reaches 20, the entry cannot trigger. Once triggered, the plan is a short with a fixed risk profile: a 2.5% adverse-move stop, which from the $70.74 close is roughly $72.52, against a 5% profit target near $67.18 — about a 2:1 reward-to-risk. Position sizing is capped at 25% per name on a fixed-risk basis. Structural exits also apply: a close above the second-ranked support structure level, or a reclaim of the 78.6% retracement, would cut the trade early, and a 90-bar holding limit forces a time exit. On evidence: this setup is backtested, and the completed 12-month test produced 5 trades with a 40% win rate, roughly +1.1% return and a maximum drawdown of about 5.0% — modest, consistent with a selective fade setup, not a high-frequency signal. Note that no robust parameter setup was established because the frozen strategy could not be fully evaluated for tuning; the thresholds above are the published rules as written. 'Wait' means concretely: set alerts at $70.81 (band middle) on BKH and watch the 9-day versus 50-day average gap close from $1.31 toward zero. If ADX keeps fading while price chops sideways, the setup stays dormant — there is no action today other than monitoring, not scaling in early.

A demand story with a completed backtest behind it

The core tension this idea trades is real and well-documented: on October 6, 2026, Reuters reported Black Hills' $1.8 billion commitment to power a Google data center, while Bloomberg reported Virginia's Lieutenant Governor opposing NextEra's $67 billion pursuit of Dominion the same day. That is a concrete catalyst on both legs of the spread — small-cap utility demand expansion on one side, and a large-cap deal facing political friction on the other — landing within hours of each other, not a vague thematic claim. The backtest, while modest, actually supports the thesis direction: over 12 months, the short-BKH / long-D / short-NEE configuration produced a positive 1.1% return with a contained 5.0% maximum drawdown across 247 evaluated daily bars. Critically, the strategy did this through a period when the equity curve was persistently negative for most of the window before recovering — suggesting the setup has some mean-reversion or thesis-realignment quality rather than pure market-direction dependence, which is exactly what the idea claims (a spread that works regardless of broader market direction). The fundamental data partially supports the contrast the idea is drawing. Dominion's Q1 2026 results were genuinely strong — revenue up 20.9% to $5.1B, operating margin up nearly 10 percentage points to 27.3% — but this is precisely the kind of strength that gets priced into a deal target if the deal is at risk of falling apart. Meanwhile, Black Hills' Q2 2026 free cash flow improved materially to -$23.0M from -$91.2M the prior quarter, showing the capex burden is easing even as the data-center buildout ramps. The thesis is not asking for BKH to be a bad business — it is asking for a spread between a company spending $1.8B on contracted AI demand and two large utilities whose relative valuations are distorted by deal and scale dynamics. Position sizing and risk controls are disciplined: 2.515% fixed risk per position with a hard 5.03% take-profit and structural stop-losses, meaning the strategy is built to survive the kind of choppy, sector-wide moves utilities routinely make without a single trade destroying the account.

A thesis fighting strong fundamentals with a thin sample

Start with what the completed 12-month backtest actually produced: a total return of 1.1% across just 5 trades with a 40% win rate, meaning two of the five were losers, and the strategy sat mostly flat or underwater through the entire evaluation window before recovering at the very end. The maximum drawdown of roughly 5.0% is not catastrophic, but the equity curve spent the better part of a year in negative territory — the position spent most of the period between -0.4% and -4.4% underwater before finishing at +1.1%. That is a thin margin of comfort for a setup that requires both the short leg and the market to cooperate. The backtest was computed on daily bars…

Scores

• Conviction score breakdown: 45
• Thesis support: 55
• Trade readiness: 30
• Risk quality: 60
• Backtest evidence: 35
• Fundamentals trend: 45

Watch items

• BKH — ADX (14) on BKH
• BKH — Price vs 20-day band middle on BKH
• BKH — EMA (9) vs EMA (50) on BKH
• BKH — BKH quarterly revenue trajectory
• D — Dominion dividend event
• D — NextEra-Dominion merger status
• NEE — NEE relative momentum