Vistra is one of the largest electricity companies in the United States. It does two things at once: it generates electricity from a fleet of about 44,000 megawatts of power plants, and it sells that electricity directly to roughly 5 million homes and businesses across 18 states and Washington D.C. The company earns money every time a customer uses power, paying Vistra for electricity the way someone pays a phone bill, month after month. Its brands include TXU Energy in Texas, Ambit Energy, and several others. Because Vistra both makes and sells electricity, it can avoid paying outside suppliers for power, which helps keep its costs down. The diagram below traces where the money goes.

Five years of financial data tell a story of a company that went through a rough patch and then pulled itself out. In 2021 and 2022, Vistra was burning through cash. It produced negative free cash flow of $1.2 billion in 2021 and negative $0.8 billion in 2022. Revenue was growing, climbing from $12.1 billion in 2021 to $13.7 billion in 2022, but the company was spending more than it was taking in. Then 2023 changed everything.

In 2023, free cash flow jumped to $3.8 billion and operating cash flow hit $5.5 billion. Revenue kept climbing, reaching $14.8 billion that year. The turnaround was real and measurable. By 2024, revenue had grown further to $17.2 billion, and by 2025 it reached $17.7 billion. But the free cash flow picture got murkier again. It fell from $3.8 billion in 2023 to $2.5 billion in 2024 and then to $1.3 billion in 2025. The reason is not that the business got worse at making money. The reason is that Vistra started spending heavily on expansion.

Vistra has been borrowing heavily to grow. Net debt was $9.2 billion in 2021. It fell to $8.6 billion by 2023, which looked encouraging. But then it reversed sharply. By 2024, net debt had risen to $14.2 billion, and by 2025 it climbed further to $16.9 billion. The company is taking on debt to buy power plants and sign long-term deals. In October 2025, it completed the Lotus acquisition, paying $1.9 billion for seven natural gas plants. In December 2025, it signed agreements to acquire Cogentrix Energy, a deal that adds 10 more natural gas facilities totaling about 5,500 megawatts, with the transaction expected to close in mid-to-late 2026. The risk listed in its own filings makes this plain: Vistra currently carries about $20.7 billion in total debt.

The logic behind all this borrowing is a bet on rising electricity demand. Data centers, electric vehicles, and oil field operations are all consuming more power. Vistra has already started locking in long-term customers. In September 2025, it signed a 20-year power purchase agreement with Amazon Web Services for 1,200 megawatts of nuclear power from its Comanche Peak plant, with delivery expected to begin in late 2027. In January 2026, it announced 20-year agreements with Meta Platforms for 2,609 megawatts from its nuclear plants in PJM, the grid region covering much of the eastern United States. These are not small deals.

Nuclear power now represents 15 percent of Vistra's total generation capacity. It is the part of the fleet that tech companies most want to contract with, because it runs around the clock and produces no carbon emissions. But the nuclear side of the business comes with its own complications. The company is also planning expensive uprates, meaning upgrades that would add 433 megawatts of additional capacity across Perry, Davis-Besse, and Beaver Valley nuclear plants, with full delivery expected by the end of 2034. That is a long construction timeline with regulatory and engineering hurdles along the way.

Vistra's three most serious documented risks all hit the same nerve. First, its power plants sell into wholesale markets where prices swing wildly, and a sustained drop in electricity prices would cut revenues sharply. Second, the fuel it burns, mostly natural gas, is also volatile. A spike in gas prices raises costs at the same time that wholesale power prices may not be rising fast enough to compensate. Third, the company's large debt load means that rising interest rates or difficulty refinancing loans could squeeze the cash available for everything else. These are not hypothetical concerns. They are listed as high-severity risks in the company's own filings.

There is also a fourth risk worth noting separately. A fire broke out at Vistra's Moss Landing 300-megawatt battery storage facility in California in January 2025. The company wrote off approximately $400 million related to the facility and incurred additional impairment charges of approximately $155 million on the adjacent 100-megawatt battery. Cleanup costs under an agreement with the Environmental Protection Agency are estimated at approximately $110 million. The company collected approximately $500 million through February 2026 under its insurance policies, which helped offset the damage, but ongoing litigation and remediation costs remain uncertain.

Meanwhile, new solar and wind plants, often built with government subsidies, are competing directly with Vistra's older coal and gas facilities. The company has already announced plans to retire its Coleto Creek coal plant by 2027 and convert it to natural gas. It is a sign that the competitive pressure from cheaper renewables is real and forcing decisions. Vistra's coal fleet still represents 20 percent of its total capacity, which means this transition is far from finished.

The long-term contracts with Amazon and Meta represent Vistra's attempt to convert its nuclear fleet from a commodity business into something more like a predictable subscription. If those agreements hold and deliver as planned, the company's cash flow base becomes more stable and less exposed to daily wholesale price swings. But delivery on the Amazon contract does not begin until late 2027, and the Meta agreements ramp over several years through 2034. Until then, the company continues to depend heavily on volatile market prices for most of its revenue.