Vistra Still Leads the AI Power Trade-But at 19x Earnings, Is the Moat Fully Priced?

VST remains a leading AI power stock, but the setup is less forgiving

At roughly 19x forward earnings, Vistra's AI-power advantage still looks real, but the margin for error is shrinking.

The stock is in an awkward spot: it is still more than 300% up in the past year, but it has also fallen 22% from its September peak. That is why the debate is active again.

Why Vistra still has a real moat in AI power

This moat is not a slogan. It comes from the fit between what Vistra owns and what AI infrastructure now needs.

U.S. data-center demand is getting much larger

U.S. data-center demand is now on pace for ~750 TWh by 2030, with capacity projected to reach 95 gigawatts. The spending behind that demand is still accelerating, not fading: the Magnificent Seven are expected to deploy $527 billion in fiscal 2026 AI and data-center capex. And the customers are no longer buying only marginal megawatts. They are leasing power for campuses drawing 100+ megawatts each, with some approaching 1 gigawatt. That shifts the competitive question from cheap electricity to reliable, scalable power at scale.

Scale and firmness are the advantage

Vistra is positioned for that shift because it already has the assets. It operates roughly 44,000 MW of generation capacity across gas, nuclear, coal, solar, and storage. More important, it has already converted part of that fleet into contracted, long-duration supply. It has a 20-year PPA for up to 1,200 MW from Comanche Peak, and 20-year agreements for more than 2,600 MW from three nuclear plants.

Bears can fairly argue that those deals cover only part of the fleet and that execution risk remains around uprates, relicensing, and customer siting. But the strategic point stands: Vistra has both the scale to serve hyperscalers and the nuclear base to offer lower-carbon, firm power that data-center buyers increasingly want.

Why that matters more over time

When demand grows this large, the scarce asset is not just electrons. It is deliverable power, siting flexibility, and multi-year cash-flow visibility. Vistra's asset mix gives it a better chance than most pure gas or pure renewable power names to turn AI's power appetite into contracted earnings. The main boundary condition is execution: if uprates, relicensing, or customer siting slip badly, the moat can still exist while the monetization timeline stretches.

The earnings story is stronger, but the stock now needs discipline

The question is no longer whether Vistra is a good business. It is whether the stock still offers enough margin of safety for the risks attached.

Vistra now has a clearer bridge from thesis to numbers

What changed is that Vistra now has a sturdier bridge from thesis to numbers. Management has pointed to 2026 ongoing operations adjusted EBITDA guidance of $6.8 billion to $7.6 billion, along with a 2027 Ongoing Operations Adjusted EBITDA midpoint of $7.4 billion to $7.8 billion. That is the kind of cash-generation path that can support a re-rating if contract execution keeps improving. It also helps explain why the $4.0 billion senior note offering mattered: this was not a distress raise. It was balance-sheet support for a company trying to fund growth without weakening its credit posture.

Valuation leaves less room for mistakes

The current price no longer leaves much room for delay or execution wobbles. That is fine if success keeps arriving on schedule. The risk is that the market has already capitalized much of the near-term contract upside.

The clearest example is accounting noise versus operating reality. Vistra reported an $808 million unrealized pre-tax hedge loss in 2025. Bears will say that is a warning about earnings volatility and hedging discipline. Bulls will say it is a non-cash item that obscures the cash engine underneath. I think the bulls are closer, but only if management keeps execution clean. If hedge accounting and settlement patterns keep creating large swings, investors may continue to punish the stock even when the underlying business is holding up.

Against Constellation, Vistra looks more offensive

Against Constellation, Vistra looks more like the higher-beta version of the same AI power trade. CEG is often favored for income stability and capital preservation, with a tripled dividend and 10% annual growth target. Vistra, by contrast, leans more on buybacks over yield and has signaled confidence in cash flows through financing and capital return actions. That makes Vistra more offensive, not more bulletproof.

So the practical setup matters. I would view the $140 to $160 range as a more disciplined accumulation zone if the earnings bridge holds. Buying after insider sales would be a different decision than buying during patient accumulation. Right now, the business still looks stronger than most peers; the stock just needs a price that matches the execution risk.

What would confirm the thesis-and what would challenge it

Valuation is the filter, but execution is the test.

What would confirm the bull case

• More contracts like the Meta and AWS deals. Those 20-year PPAs for more than 2,600 MW and the 20-year AWS agreement for up to 1,200 MW matter because they show hyperscalers are willing to lock in long-duration power, not just spot megawatts. Additional signed demand would show the moat is widening, not just being admired.
• A cleaner bridge from earnings power to cash. Vistra has already pointed to a 2027 Ongoing Operations Adjusted EBITDA midpoint of $7.4 billion to $7.8 billion. If management keeps that bridge intact while funding growth through the $4.0 billion senior note offering and maintaining investment-grade ratings, the market can keep treating it as a growth asset rather than a cyclical utility.

What would challenge the thesis

• Uprates and relicensing could slip. Management has said uprates are likely to come after 2028. That gives timing flexibility, but it could disappoint investors who wanted sooner proof that contracted nuclear output could expand.
• Policy or market friction could rise. Regulators proposed measures to curb power costs, which hurt sentiment around the IPP model. If that trend spreads, contract economics and timing could come under pressure.
• The stock still pays for execution. Even after the pullback, Vistra no longer has much room for messy delivery.

How the VST vs. CEG choice looks now

Constellation still has the better income and capital preservation story, with a tripled dividend and a 10% annual growth target. Vistra remains the more offensive vehicle, with buybacks over yield and a bigger balance-sheet expansion profile. Choose CEG for durability. Choose VST if you want more upside from contract execution and can tolerate a harsher penalty if that execution wobbles.