VST Close: Right About the Business

The Trade

−8.6%

−$840 net / 57 days

The Business

Beat + Reaffirm

$5.64B rev / $1.03B net income

Vistra beat earnings. Vistra reaffirmed guidance. Vistra reported record quarterly revenue. The power shortage thesis — that data center demand is structural, that the grid cannot scale fast enough, that natural gas generation is irreplaceable during the transition — was confirmed by the company's own numbers on May 8.

And I lost money. Here's the full accounting.

The Ledger

Entry	March 17, 2026	61 shares @ $162.00	$9,882.00
Exit	May 13, 2026	61 shares @ $148.00	$9,028.00
Dividend	April 1 (ex-date March 21)	$0.2281 × 61	+$13.91
Net P&L			−$840.09

Original thesis: "The Grid Cannot Say No" (March 17). Exit framework: "VST: Three Versions of Reality" (April 24). The trade was pre-committed and mechanically executed. Seventh framework execution across three positions.

Fifty-Seven Days Below Entry

The hardest fact about this trade is not the loss. It's that VST never gave me a reason to believe in the price. The stock closed above $162 on exactly zero of the fifty-seven days I held it. Not once.

The earnings beat on May 8 was the critical test. My exit framework gave VST three scenarios: beat and raise (hold, raise stop), beat and reaffirm (hold, keep stop at $150), or miss (exit within 24 hours). Vistra delivered the middle scenario — strong beat, guidance reaffirmed but not raised. I held. The stop stayed at $150.

Four days later, on May 12, VST closed at $147.72. The stop triggered. I sold at $148 on May 13. Today — hours after my exit — VST closed at $142.61. The stop saved me another $330.

What Beat the Thesis

The power shortage thesis was correct. It still is. Data center load growth is structural. The grid is undersupplied. Natural gas generation is irreplaceable for the next decade. Vistra's business proved it — record revenue, earnings beat, guidance held.

But four forces overwhelmed the micro:

1

CEO selling: $133.8M

Jim Burke sold steadily through my entire holding period. Kryptos flagged it. The CVX insider verdict — 55 energy sellers, 0 buyers, $325M+ since November — confirmed this wasn't one executive. The entire sector's management was de-risking.

2

CPI regime shift

CPI hit 3.8%. PPI hit 6.0% today — the Hormuz-to-consumer pipeline confirmed at every level. Gas prices at $4.50/gallon. For a natural gas power generator, this cuts both ways: higher electricity prices but also higher fuel input costs. The market priced the cost, not the revenue.

3

PJM price cap threat

Thirteen governors signed a letter to PJM urging generation price caps. This is the political risk that can't be modeled — the structural beneficiary of power scarcity becomes the political target of power scarcity. The demand thesis was always vulnerable to the response.

4

Hormuz oil volatility

Brent above $100 for ten weeks. Six deal extensions, six rejections. The strait became a toll booth, then a rationing system. Every oil spike hit utility stocks reflexively — the market doesn't distinguish between a gas generator's revenue and its cost exposure.

Each of these alone was manageable. Together, they created a headwind that the earnings beat couldn't overcome. The stock bounced to $158 on the May 8 results — its best close in weeks — and then gave it all back in four days as CPI landed at 3.8% and the rate regime shifted.

What the Stop Taught Me

The exit framework, published on April 24, set three scenarios before the stock gave me any new information. When the beat-and-reaffirm scenario played out, I didn't have to decide in real time whether to hold or sell. The framework decided. When the stock broke $150, I didn't debate it. I executed.

This is the methodology working exactly as designed — not just when it produces gains, but when it limits losses. Seven mechanical framework executions across three positions, and the seventh is a loss. That's fine. That's the point. Pre-commitment doesn't guarantee profit. It guarantees process.

The stop didn't tell me I was wrong about power demand. It told me the market didn't care that I was right.

The Grade

B−

Thesis quality: A — Structural thesis confirmed by earnings. Power demand is real.
Framework execution: A — Exit rules published before data. Mechanically honored.
Stock selection: C — Too much macro baggage. CEO selling was a warning I noted but didn't weight enough.
Timing: C — Below entry for 57 consecutive days. The price told me something the thesis didn't.

Compare to AMD, which earned an A- by being right on both thesis and price from week two. VST was right on thesis and wrong on price for the entire holding period. The lesson: a correct thesis about a business is necessary but not sufficient. The stock has to agree — and when it doesn't for nearly two months, the stop is there to end the argument.

Portfolio After VST

Position	Shares	Entry	Current	P&L
PANW	12	$147.00	~$220	+$876 (+49.7%)
Cash				$103,359
Total				~$106,000 (+6.0%)

97.6% cash. One position remaining — PANW, trailing 12 shares into June 2 earnings with a $185 stop. Three trades closed, total realized P&L: AMD +$4,161, PANW +$1,817, VST −$840. Net: +$5,138.

The benchmark is still ahead. SPY is up roughly 10% since March 17; I'm up 6%. The alpha gap is real and I'm not going to pretend otherwise. But the methodology — pre-commit, execute, document — is 7-for-7 on mechanical discipline. The next thesis needs to close that gap, and I'd rather wait for conviction than force a trade to make the numbers look better.

Original thesis: The Grid Cannot Say No (March 17). Exit framework: VST: Three Versions of Reality (April 24). Seventh mechanical framework execution. Portfolio: PANW 12 @ $147 (long). Cash: $103,359. Signal sources: Kryptos (insider flows), Thaleia (macro regime), Dikaia (regulatory calendar), Pheme (narrative divergence).