The Repricing of Reliable Generation Assets — Can VST Turn AI Power Contracts into Cash Per Share?

Vistra Corp. (NYSE: VST) In-Depth Equity Research Report

Analysis Date: 2026-05-21 · Data Through: 2026-05-21 / Q1 2026 disclosures

Chapter 1: Put VST in the Right Place First: It Is Not an Ordinary Utility

When researching VST, the easiest mistake is to make one of two opposite errors. One is to treat it as an ordinary utility and look only at dividends, defensive attributes, and regulated returns; the other is to treat it as an AI concept stock and look only at data center power demand growth. Both readings miss what truly matters: VST owns a portfolio of already-built nuclear and gas assets that can provide reliable output when the power system is tight.

The company is closer to a market-based generation platform. It sells power in electricity markets, sells reliable capacity available for future peak periods, and signs long-term power purchase agreements with large customers. The tighter the power system becomes, the more valuable reliable generation assets become; but more valuable assets do not automatically mean the stock is cheap. Investors must also assess whether that value can pass through contract prices, the capacity market, hedging, capital expenditures, debt, and buybacks, and ultimately become cash flow per share.

VST's main line can first be compressed into one cash pathway:

Repricing of reliable generation assets → higher capacity prices and long-term contracts lift EBITDA → operating cash flow less capital expenditures, margin, and interest → residual cash used for buybacks and debt reduction → higher cash flow per share.

The following chapters all revolve around this pathway. Chapter 3 first explains what business VST actually is; Chapter 5 examines whether the Q1 2026 results validate profit improvement; Chapter 6 explains why PJM (which can be simply understood as a large electricity market and grid dispatch system in the eastern and midwestern United States; it is not a company, but a power market rule set and dispatch platform, and a large portion of the value of VST's East segment is tied to PJM capacity prices) capacity prices and the installed-capacity base could lift East segment profits to a new level; Chapter 9 returns to ordinary free cash flow; and Chapter 10 then judges whether the current price leaves sufficient return potential.

To avoid being blocked by terminology at the outset, first clarify several terms:

After reading this report, one only needs to return to four questions:

1.1 Report Information and Data Basis

Data basis note: company-disclosed data primarily uses Vistra earnings releases and SEC 10-Q filings; PJM capacity prices and cleared volumes use PJM official documents and Vistra investor materials; ordinary free cash flow, capacity revenue, and price ranges in the report are research estimates, with formulas and limitations noted in the corresponding chapters.

---

Chapter 2: What Judgment Can Be Made Now: VST Deserves Research, but the Purchase Price Matters

The most important disagreement around VST today is not "whether AI data centers need more electricity," but whether "the uplift in the value of reliable generation assets can continue to translate into ordinary free cash flow and cash flow per share in 2027-2028." This company is no longer a traditional low-growth utility, nor is it simply an AI concept stock. It is more like a cash-flow portfolio composed of already-built nuclear, gas, coal, retail electricity customers, a hedge book, and long-term power purchase agreements.

The market's valuation of VST has already acknowledged part of the change: PJM capacity prices have risen sharply, and East segment profit has improved materially; Meta and AWS have pulled some nuclear assets out of short-term wholesale power price volatility and turned them into 20-year contracts supported by customer credit; if Cogentrix closes, it will add about 5.5GW of modern gas assets; the company also continues to repurchase shares, with share count down about 30% since 2021. But equally important, much key information remains undisclosed: Meta/AWS contract prices, VST's hedge prices by region, actual interest and maintenance capital expenditures after Cogentrix consolidation, and 2028/2029 PJM capacity rules. These information gaps do not prevent an investment judgment, but they limit the current share price's margin of safety.

The one-sentence conclusion can be compressed as follows:

VST's assets have become more valuable, but whether the stock is still cheap must be proven by 2027-2028 ordinary free cash flow, net debt, and buyback quality.

---

Chapter 3: How VST Actually Makes Money: It Is Not a Traditional Utility That Collects Electricity Bills

VST is easily placed in the "utilities" sector, but viewing it through the lens of a traditional electric utility creates major misjudgments. Traditional regulated utilities make money through regulated asset base, allowed return on equity, and rate approvals, with a growth path more like "capital expenditures enter the regulated asset base and are then recovered through electricity rates." VST's core is different. It mainly owns generation assets in competitive power markets, while also having retail electricity customers and a hedge book, using contracts, the capacity market, wholesale power prices, and retail load to turn power assets into cash flow.

This distinction means VST's valuation cannot be based only on P/E, nor only on dividend yield. VST's value is closer to the following chain:

Already-built reliable generation assets → capacity prices, energy margins, long-term power purchase agreements, and retail hedging → adjusted EBITDA → operating cash flow → ordinary free cash flow → buybacks and debt reduction → cash flow per share.

If any part of this chain breaks, the share price cannot be supported solely by "AI power demand." For example, if capacity prices are high but regulators claw back value, the profit ceiling falls; if long-term power purchase agreements exist but prices are low, cash flow quality improves but upside is limited; if EBITDA grows but maintenance capital expenditures, margin, and interest consume the cash, per-share value will not increase in tandem.

VST should also be separated from CEG, NRG, and TLN. CEG has higher nuclear purity and stronger quality, but its valuation is more expensive; NRG is cheaper, but retail, M&A, and cash-flow friction are more complicated; TLN has high sensitivity, but it is smaller and has higher financing and project concentration. VST sits in the middle: its quality is not as pure as CEG's, its sensitivity is not as intense as TLN's, and its price is not as cheap as NRG's, but its asset portfolio, buybacks, hedging, and project uplift make it easier to establish a reproducible cash-flow pathway.

Therefore, VST's key question is not "whether it is a utility stock," but "whether the reliable power assets it controls can, in the right regions and at the right time, turn into cash flow per share through legal and sustainable pricing mechanisms."

3.1 VST's Sources of Value Need to Be Split into Five Layers

VST's business value cannot be explained with a single label. It is not a "nuclear stock," not a "gas stock," and not simply an "AI power demand stock." Its cash flow comes from five places, and each place has different sustainability, volatility, and valuation multiples.

The first layer is already-built reliable generation assets. The value of nuclear, gas, and some coal assets lies not in nominal installed capacity, but in whether they can actually connect to the grid during peak periods, pass system reliability tests, and dispatch output when the market needs it. AI data centers need continuous, stable, predictable electricity, and ordinary renewable power sources cannot solve this problem alone. VST's nuclear and gas assets are therefore more likely than in the past to be repriced by long-term customers, the capacity market, and grid reliability mechanisms.

The second layer is capacity revenue. The PJM capacity market brings VST not one-time transaction revenue, but payment by the system for future reliability. In the three delivery years 2025/2026, 2026/2027, and 2027/2028, VST's cleared PJM capacity has all exceeded 10GW, and prices are far above the old cycle. It has turned part of "power tightness" from news into a revenue item, which is the key difference between VST and pure concept companies.

The third layer is energy margins. VST remains subject to power prices, natural gas prices, heat rates, basis, and unit availability. Rising natural gas prices are not necessarily positive, and rising power prices do not necessarily all stay with shareholders. What truly affects profit is the spread between power prices and fuel costs, as well as the company's ability to hedge and lock in that spread. The Texas segment is representative of this type of margin.

The fourth layer is retail customers and the hedge book. VST does not simply sell power to the market, nor does it sell all its power to long-term customers. The retail business provides customer load and also allows the company to reduce part of market volatility through the natural matching between generation and retail. But Q1 2026 has already shown that retail can drag profit in mild weather and cannot be valued as a high-quality growth business.

The fifth layer is long-term power purchase agreements and M&A. Meta, AWS, and Cogentrix represent three different types of value: Meta improves the cash flow quality of PJM nuclear power, AWS supports the long-term value of Comanche Peak, and if Cogentrix closes, it will add dispatchable gas assets. They all strengthen VST's long-term narrative, but the speed and certainty with which they enter cash flow per share are not the same.

Combining these five positions is VST's true form. For investors, the two most dangerous misreadings are: first, treating VST as a regulated utility and looking only at dividends and defensive attributes; second, treating VST as an AI theme stock and looking only at data center load. The more reliable reading is that every profit source must land in revenue, profit, cash flow, and share count.

3.2 Differences Versus CEG, NRG, TLN, and NEE Determine Valuation Boundaries

VST and CEG are often placed in the same basket because both companies own nuclear assets and both have signed long-term power agreements with large technology customers. But CEG has higher nuclear purity, cleaner assets, and lower leverage, so the market is willing to assign it a higher multiple. VST's advantage is not being "purer," but being "somewhat cheaper, with more diversified assets and more visible cash buybacks." If investors only want to buy the highest-quality nuclear platform, CEG is cleaner; if they want to find a balance between quality and price, VST deserves more research.

The difference between VST and NRG lies in cash-flow readability. NRG is cheaper, and its retail and generation portfolio also has value, but its M&A, retail business, free cash flow seasonality, and financial structure are more complicated. VST's current advantage is that its core projects, segment profit, and 2026-2028 cash-flow path are easier to disaggregate. It may not be cheaper than NRG, but it is easier to track with a clear set of cash-flow disciplines.

The difference between VST and TLN lies in scale and event concentration. TLN has stronger sensitivity to PJM nuclear power and data center contracts, and a single event may bring greater share price sensitivity, but financing, project concentration, and regulatory events are also more sensitive. VST's assets are more widely distributed, and the impact of a single contract on total company value is not as extreme as for TLN, making it more suitable for a steadier segment and cash-flow model.

The difference between VST and regulated utilities such as NEE, DUK, and SO is even greater. Regulated utilities earn allowed returns by putting capital expenditures into regulated asset base and then recovering them through rates. Their growth depends on regulated asset growth, allowed return on equity, cost of capital, and electricity bill affordability. VST does not have this stable regulated-return protection; its upside is larger and its downside is also more direct. The market should not give VST the stable multiple of a traditional defensive utility, nor should it demand that VST have the same low volatility as a regulated utility.

This means VST's reasonable valuation should not mechanically apply a single peer multiple. It is more suitable to cross-check across four dimensions: dual cash-flow yield, segment EBITDA, contract quality, and leverage. Looking only at EV/EBITDA can easily ignore maintenance capital expenditures, margin, and buyback prices; looking only at P/E can be misleading because of derivative revaluations; looking only at the AI narrative can easily capitalize contract prices and delivery timing too early.

3.3 VST's Real Business Question: Who Pays for Reliability

When the power system is loose, all generation assets are easily viewed as commodities; when it is tight, reliability itself becomes a price. VST's opportunity comes from this change. Data centers, industrial load, electrification, extreme weather, and coal retirements will all make the system need more dispatchable power. But the system needing more power does not mean every generator makes money. Only companies that can provide available capacity in tight regions and lock in prices through contracts, the capacity market, or a retail portfolio may turn scarcity into shareholder cash flow.

This is why VST research must track the "payment mechanism." If payment comes from the capacity market, investors need to look at clearing prices, cleared volumes, rules, and regulatory response; if payment comes from long-term power purchase agreements, they need to look at price, term, customer credit, minimum purchase volume, and delivery cadence; if payment comes from the energy market, they need to look at spot prices, fuel costs, hedge coverage, and margin; if payment comes from retail, they need to look at customer load, gross spread, and weather. Different payment mechanisms correspond to completely different valuation multiples.

The most valuable change in VST now is that it is beginning to move from "market price exposure" toward a hybrid form of "market price + long-term contracts + capacity revenue + buybacks." This will improve cash flow quality, but it will also reduce some spot upside sensitivity. Investors cannot simultaneously demand that it be as high-quality as CEG, as high-beta as TLN, and as cheap as NRG. VST's appeal lies in its compromise position: scarce enough, reproducible enough, and its valuation has not yet fully become the price of a quality asset.

3.4 Three Charts to Build the Verification Map

Following the more visual reference page, the public section now uses charts instead of extra tables. The goal is to show VST's core verification path at a glance: the PJM capacity price curve, the ordinary free cash flow range, and the timing of capacity additions and major projects.

These three charts are not new price targets. They are the reading map for the rest of the report: first the price curve, then the cash range, and finally whether projects land on schedule.

---