VST Vistra Corp. : Bullish and Bearish Analyst Opinions

Vistra is a buy with doubling earnings.

Speaker identifies Vistra as the largest unregulated power producer in the US, operating in liberalized markets (PJM, ERCOT). It owns a large fleet of gas and nuclear assets and has long-term contracts with Amazon and Meta. As an unregulated producer, VST directly benefits from price spikes in electricity caused by AI-driven demand. It has flexibility to manage peak loads and can pass on cost increases. LONG because its business model allows it to fully capitalize on rising and volatile power prices driven by data center demand. Exposure to potential regulatory changes or a sharp economic downturn reducing energy demand.

There is no way for us to deliver on the ambitions that we have as a country for AI and for additional technologies without a growing energy and affordable energy supply. Also in the horizon is a renaissance of nuclear power. AI data centers require massive, uninterrupted baseload power that wind and solar cannot reliably provide alone. This forces a structural reliance on nuclear energy to meet the demands of the computing revolution, directly benefiting unregulated nuclear power producers and uranium miners. Long nuclear utilities and uranium producers as they transition from legacy energy providers to critical infrastructure for the AI revolution. Regulatory hurdles and massive capital costs for nuclear expansion; long lead times for building new reactors; volatility in uranium commodity prices.

"Nuclear energy, modular nuclear energy, those are the things that as people see this kind of crisis... will be looking at even more... innovations around AI... the infrastructure part, the data centers." Extreme volatility in fossil fuel markets is forcing capital allocators to seek stable, baseload power alternatives. Simultaneously, the AI revolution requires massive, uninterrupted electricity for data centers. Nuclear power providers and uranium miners sit at the perfect intersection of these two mega-trends, serving as critical infrastructure for the AI buildout. LONG because nuclear energy is the only zero-carbon baseload power source capable of meeting the exponential energy demands of AI data centers. Regulatory hurdles delaying the deployment of modular nuclear reactors, or a macroeconomic shock that halts AI capital expenditures.

Clearly, the energy efficiency is the biggest threshold. We just don't have enough energy to power these chips... fully pivoting from chips to energy. That is the biggest thing in the world. A single GPU rack uses massive amounts of power, and the current grid cannot support the exponential growth of AI data centers. Because biological computing (which runs on a fraction of the power) is decades away from replacing silicon, the immediate second-order effect of the AI boom is a massive supply-demand imbalance in electricity. Utility companies and independent power producers will command massive premiums to supply baseload power to tech giants. Long the energy and utility sector as the physical bottleneck to AI expansion shifts from silicon chips to raw electricity generation. Regulatory hurdles block new power plant construction, or AI models become drastically more efficient on silicon, reducing projected power demand.

"A lot of the infrastructure needs are currently being driven by some of our technology innovations... whether it's more power that we need for data centers, right, data center construction." "Infrastructure" is now a derivative trade on AI. To support LLMs, you need physical Data Centers (EQIX, DLR) and massive amounts of electricity/power generation (VST, CEG). These "Real Assets" have inflation-linked contracts and secular demand growth. Long Data Center REITs and Power Producers/Utilities. Regulatory pushback on power consumption or a slowdown in AI capex spending.

The White House is meeting with Big Tech to ensure they "pay their way" regarding electricity for AI data centers. Amazon (Holly Sullivan) confirms investing $340 billion in 2025 alone in US infrastructure, largely for data centers and AI. If the government forces Tech to insulate consumers from rate hikes, Tech companies must directly fund power generation and grid upgrades. This capital flows directly from Amazon/Big Tech into Independent Power Producers (VST, CEG) and Grid Equipment manufacturers (ETN) to build dedicated capacity. LONG. The political pressure removes regulatory hurdles for power companies, as Tech giants are now effectively underwriting the grid expansion. strict price caps imposed by the administration on energy contracts.

The President is holding a roundtable with "big tech companies with AI companies... who have pledged to pick up the cost of the electricity tab in these small towns." This confirms the "AI Power Shortage" thesis is now a White House policy priority. Tech giants (MSFT/GOOGL) are bypassing traditional utility ratepayer friction by directly subsidizing infrastructure. This is incredibly bullish for Independent Power Producers (VST/CEG) who have the baseload capacity AI needs, as it removes political regulatory hurdles. LONG. It validates the revenue model for power producers selling directly to hyperscalers. Regulatory pushback on tech companies controlling critical infrastructure.

Arnold notes a "mad scramble" to build data centers. He states the buyers (Big Tech) are the "largest, most profitable companies that have ever existed," are growing free cash flow, and are less concerned about price than they are about speed and reliability. Renewables (solar/wind) are intermittent and transmission is slow to build. To meet immediate, 24/7 AI power demand, tech giants must sign deals with Independent Power Producers (IPPs) that have existing dispatchable generation (nuclear/gas) and grid interconnection rights. LONG. These IPPs hold the scarce asset (reliable electrons) in a seller's market. Regulatory intervention on power prices or a sudden deceleration in AI capex.

Granat focuses on "IPPs" (Independent Power Producers). She notes that while chips are abundant, power is the hard constraint for AI. To turn megawatts into dollars for AI inference, you need grid connection. IPPs own the physical power generation and interconnects. They are signing long-term contracts that extend cash flow duration, repricing them from cyclical utilities to secular growth assets. LONG Independent Power Producers. Regulatory caps on power prices or faster-than-expected SMR (nuclear) deployment disrupting traditional IPPs.

Trump explicitly states regarding AI data centers: "I made it mandatory where they have to build their own electric power plant. You can't take it [from the grid]... They're becoming essentially their own utility." This mandate forces Hyperscalers to spend billions on independent power generation. This is a direct revenue pipeline for Independent Power Producers (Vistra, Constellation) and equipment manufacturers (GE) who will be contracted to build and operate these behind-the-meter nuclear and gas plants. LONG Power Generation & Infrastructure. Regulatory delays in permitting new private power plants.

"We're constrained by watts and wafers, meaning we don't have enough electricity... and as a result, that becomes the limiting factor." If the AI supercycle is bottlenecked by the electrical grid rather than software demand, the value accrues to power producers (VST/CEG) and grid infrastructure providers (ETN). They hold the scarce resource required for AI expansion. LONG. These are the "pick and shovel" plays for the energy crisis described in the transcript. Regulatory caps on power pricing or hyperscalers building off-grid nuclear solutions that bypass public utilities.

"There were a few nuggets a bit around getting tech companies to build power plants... power prices for everybody might go down." The administration is explicitly encouraging Hyperscalers (AMZN, MSFT, GOOGL) to vertically integrate their energy supply to support AI compute needs. This removes regulatory hurdles for Big Tech to partner with or acquire Independent Power Producers (VST, CEG) and nuclear infrastructure, turning energy from a bottleneck into an asset. Long Hyperscalers and Nuclear/IPP utilities as the "AI Energy" trade gets executive backing. Environmental regulation delays or grid interconnection issues.

Secretary Burgum explicitly states the administration is pushing a "Bring Your Own Power" (BYOP) policy, allowing tech companies to build "off the grid" power generation to support data centers and "win the AI arms race." The administration recognizes the grid cannot handle AI power demand. By deregulating or encouraging "behind the meter" power generation, this creates a massive boom for Independent Power Producers (IPPs) like Vistra and Constellation, and equipment manufacturers like GE (turbines). It also removes a bottleneck for AI chip demand (NVDA). LONG. The "Energy Dominance" narrative directly supports the AI infrastructure build-out. Environmental regulations or local opposition to new power plant construction.

Burgum explicitly endorses "behind the meter capability" and "BYOB - bring your own power" for Hyperscalers, stating that data centers should be able to operate "off the grid" to avoid impacting ratepayers. He emphasizes "secure baseload" over intermittent sources. The primary beneficiaries of "behind the meter" policy are Independent Power Producers (IPPs) with nuclear and gas assets (VST, CEG) that can co-locate with data centers. This policy removes the regulatory hurdle of grid interconnection, allowing these companies to sell premium power directly to Big Tech. GE benefits as the supplier of the gas turbines required for these private power plants. LONG. This is a direct regulatory green light for the nuclear/gas co-location trade. Public backlash regarding local environmental regulations or "island mode" reliability failures.

He mentions "demand being driven by the air power" (referring to AI/Compute) and notes that the "global order [is] reorganizing" around this massive demand push. The digital economy is hitting physical limits. The "air power" of AI requires terrestrial power generation and grid modernization. This shifts value from pure software to the physical utilities and infrastructure providers (IPPs like Vistra or Constellation) that supply the electricity required for compute. Long Power and Grid Infrastructure providers. Regulatory caps on energy prices; delays in data center buildouts.

"Data centers are driving the highest electricity demand in 15 years... making the AI race key to oil as well." The AI boom is physically constrained by power availability. As demand hits a 15-year peak, Independent Power Producers (IPPs) and Utilities with capacity (like VST and CEG) possess the scarcity value required to power these data centers. LONG. The structural increase in demand from "new tech" provides a tailwind for power generation assets. Regulatory caps on power pricing or slower-than-expected AI capex deployment.

Burgum argues the US is in an "AI arms race with China" and needs "Energy Addition," criticizing the shutdown of "baseload" in favor of "intermittent" sources. AI data centers require 24/7 uptime (baseload). The administration's policy explicitly favors keeping existing thermal and nuclear plants online to feed this demand. Utilities with existing baseload capacity (Vistra, Constellation) become critical infrastructure assets with pricing power. LONG. Demand (AI) is rising while the administration prevents the supply (Baseload plants) from shrinking. Lower natural gas prices could compress margins for merchant power producers.