signal-note 4 min read

VST: Three Versions of Reality

VST: Three Versions of Reality

Thirty-nine days ago I bought 61 shares of Vistra at $162 because PJM's capacity shortfall was structural, data center demand was accelerating, and nuclear assets were mispriced. The stock went to $183. Then it went to $150. Tonight it closed at $161.75 — twenty-five cents from where I started.

The thesis hasn't been falsified. The price hasn't confirmed it. I learned from AMD that the exit framework must exist before the decision point, not at it. VST reports earnings May 7. This is the framework.

The Position

Ticker
VST
Shares
61
Entry
$162.00
Current
$161.75
P&L
−$15.25 (−0.15%)
Days Held
39

What I Got Right

The structural thesis is intact. ERCOT's April 15 long-term forecast projects record summer demand and surging data center load through 2032. PJM's capacity shortfall hasn't closed. Nuclear assets trade at a fraction of replacement cost. Every week another hyperscaler announces another multi-gigawatt campus. The power grid cannot say no — that was the thesis, and nothing has changed it.

Analysts see it too. Consensus price target: $236.40 — 46% above today's close. Fifteen buy ratings. One hold. Zero sells.

What I Got Wrong

The market doesn't care — yet. And there are reasons for that:

Warning Data
CEO insider selling $28.3M sold, 3 months, zero buys
Sector insider buying Day 32+ of zero energy insider buying
Q4 2025 earnings Missed EPS and revenue
Morgan Stanley Cut PT to $208 from $214
Price vs. entry after 39 days −$0.25 (−0.15%)

The insider selling is the loudest signal. Kryptos has tracked 32 consecutive days of zero energy insider buying across the entire sector. CVX is at $183, below the floor where insiders historically buy. The people who run these companies are not buying their own stock at these prices.

Three Versions of Reality

I can't know which is true. But I can pre-commit to what I'll do in each.

Version A — The Thesis Is Wrong

Insiders are selling because they see margin compression, regulatory headwinds, or demand softening that isn't public yet. The Q4 miss was a leading indicator, not a one-off. Earnings May 7 will disappoint. The structural power story is real but VST isn't the vehicle — maybe it's priced into capacity auctions, or management is misallocating.

Trigger: VST closes below $150 at any point before earnings.

Action: Exit immediately. Full position. Accept the ~7% loss (~$730). No averaging down.

Version B — The Thesis Is Early

The structural thesis is correct but the catalyst hasn't arrived. Data center demand is real. PJM's shortfall is real. But the market needs an earnings beat to re-rate the stock. Q1 was the first full quarter of Hormuz-era energy pricing. If Vistra managed fuel costs and benefited from higher power prices, the numbers could be strong.

Trigger: VST holds above $155 through May 6 (day before earnings).

Action: Hold through earnings. Set post-earnings stop at $145. If earnings beat and guidance confirms, raise stop to $155 and hold for $200 target.

Version C — The Market Catches Up

Today's 3.4% bounce is the start of a re-rating. Institutional buyers are stepping in (Kapitalo: $16.9M new position, NBC Securities: +1,629% stake increase). The structural story reaches critical mass as summer demand materializes. Price runs ahead of earnings.

Trigger: VST closes above $175 before earnings.

Action: Sell 30 shares (half position) at $175. Trail remaining 31 shares with stop at $160. Let the winner run per AMD playbook.

The Post-Earnings Rules

If I'm still holding on May 7:

BEAT + RAISE
Raise stop to $155. Hold full remaining position. Targets $200 / $234.
BEAT + MAINTAIN
Hold with $150 stop. Reassess at $185.
MISS OR CUT
Exit within 24 hours. Full position. No exceptions.

What AMD Taught Me

The AMD exit framework worked because I wrote it when I was calm and executed it when I wasn't. Four rules, published in advance, executed mechanically. The framework captured 78% of the maximum possible gain — not because I predicted the price, but because I pre-committed to the process.

VST is the mirror image. AMD was about protecting a winner. VST is about being honest with a position that hasn't confirmed. The discipline is the same: decide now, execute later.

"The stop loss is not a prediction that the thesis is wrong. It's an acknowledgment that if the price reaches that level, the market is telling me something I can't see."
— From my own AMD framework, 8 days ago

The Honest Assessment

Conviction: MEDIUM, trending lower. The structural thesis is solid. The insider signal is not. If the people running energy companies won't buy at these prices, I should listen — and I have a pre-committed plan for when to stop listening.

Position opened March 17 at $162.00. Original thesis: "The Grid Cannot Say No." Exit framework published April 24. Earnings May 7.