These days, Vistra is surrounded by an odd tension that is almost palpable in the numbers. Even though the stock closed at $147.72 last Friday, down roughly 4% for the day, discussions surrounding it consistently center on the notion that this could be one of the most significant utility brands in the nation at the moment. It’s because of its placement rather than anything particularly ostentatious. Nuclear-powered, gas-fueled, Texas-based, and becoming more and more entangled in the development of AI infrastructure that everyone is attempting to understand.
Nobody was reassured by the first-quarter report. Adjusted EBITDA reached $1.494 billion, which management proudly referred to as a record for any calendar first quarter, and revenue increased by about 43% year over year to $5.64 billion. These are the kinds of figures that ought to boost a stock. However, VST has been weak, strangely so, declining in a market that otherwise continues to grind. This kind of discrepancy raises questions about what investors are truly pricing in.
Margin is probably part of the answer. Over the preceding year, net margin decreased from 14.3% to 4.2%. It is the kind of thing that is quoted in bearish notes for months, and that is a steep decline. Naturally, the bulls have an answer: over the previous five years, trailing earnings grew at a rate of about 53.6% annually, and analysts continue to predict growth of about 24.8% annually. One of the reasons Vistra is so difficult to read is that both stories are located in the same dataset.
The company’s developing relationship with the hyperscalers is more difficult to overlook. For about 2,600 megawatts at its PJM nuclear sites, Vistra and Meta entered into long-term power purchase agreements. The guidance for 2026 and 2027 does not yet include any of that. Neither is the impending purchase of Cogentrix, a 5,500 megawatt natural gas portfolio that was revealed back in January. Wall Street isn’t quite sure how to model these, but management talks about them as if they’ve already been cooked into the future. There’s a feeling that the conservative outlook still conceals the true earnings power.
As you browse the trade press, you begin to see the same names—Constellation Energy, GE Vernova, NVIDIA—appearing next to Vistra. The fact that this group differs from the typical utility comparison set is noteworthy. AI narratives are not typically traded by utilities. They trade on changes in interest rates and regulated returns. The 69.3x P/E, which is significantly higher than the industry average of about 17x, indicates that Vistra is doing something different. The market might be compensating for growth that is still primarily theoretical.
Then there’s the debt issue, which is overlooked in the bullish articles. Interest coverage is uncomfortable, and quarterly earnings have been erratic, ranging from a $317 million loss in Q1 2025 to a $604 million profit in Q3. That volatility is important for a business that has actual leverage. Although those models assume all contracts and acquisitions land cleanly, TIKR’s model places fair value close to $190, suggesting about 29% upside. Usually, they don’t—at least not precisely on time.
It’s difficult to ignore the similarities to Tesla’s trading a few years ago, when the market was unsure whether it was a software company or an automobile manufacturer. Vistra is experiencing a similar identity crisis. Is it an AI infrastructure play dressed like a utility, or is it actually a utility? The response will likely have a greater impact on the stock over the next few years than any one quarter.