"The best businesses in the world are not the ones that chase opportunity. They are the ones that become unavoidable."

The Question Nobody Is Asking — But Everyone Should Be

Let us begin not with a stock price, but with a simple, almost brutal, question.

When you switch on a light, you do not think about the wire inside the wall. When you drink clean water, you do not think about the pipe beneath the street. And when ChatGPT answers your question in three seconds flat — when Google's AI summarises an entire research paper before you have finished your morning chai — you do not think about the electricity that made it possible.

But someone, somewhere, is thinking about it very hard indeed. And that someone is the Chief Executive Officer of Microsoft, Google, Amazon, and Meta — four of the wealthiest corporations in human history — all of whom are currently in a quiet, dignified panic about the same problem: they cannot get enough electricity.

This, dear reader, is the most important infrastructure crisis of our generation. Not the most discussed. Not the most photographed. But the most consequential — the one that will silently determine which companies win and which companies wait, which economies accelerate and which ones stall, which investors compound their Wealth and which ones merely watch.

And Vistra Corp — a company most people have never heard of, whose name does not appear in technology columns or innovation roundtables — sits at its very epicentre.

The River and the Dam

Saurabh Mukherjea often speaks of businesses that sit on the river of money — entities so structurally positioned that the money must flow through them, whether the economy is booming or stumbling. Vistra is not merely sitting on that river. It is the dam.

Consider the arithmetic of artificial intelligence for a moment. Every query you fire at an AI model — every image generated, every email summarised, every code debugged — consumes electricity. Not a whisper of it. A roar. The International Energy Agency tells us that global data centre power consumption will more than double between 2022 and 2026. Goldman Sachs estimates that AI data centres alone could consume 8% of all American electricity by 2030 — up from roughly 3% just a few years ago.

Now here is the delicious complication.

America's electricity grid — largely assembled in the era of typewriters and transistor radios, designed for a world of factories and air conditioners, not for a world of trillion-parameter neural networks — cannot keep pace. New power plants take a decade to permit and build. Grid interconnection queues in some American states now stretch beyond ten years, according to Lawrence Berkeley National Laboratory's most recent interconnection data. You cannot download a new power plant. You cannot disrupt your way around the laws of physics. You cannot move electrons faster than copper allows.

The world's most valuable technology companies have unlimited Capital, unlimited ambition, and some of the finest engineering minds alive — and they are being told, politely but firmly, to wait in line.

That queue is Vistra's competitive moat. And it grows longer every year.

The Crown Jewel That Cannot Be Counterfeited

This is where Vistra's genius becomes apparent — not the genius of invention, but the rarer, more durable genius of irreplaceable possession.

Vistra Corp (NYSE: VST) is America's largest independent power producer — 41,000 megawatts of electricity-generating capacity spread across Natural Gas, nuclear, solar, and battery storage Assets, operating primarily in the deregulated markets of Texas (ERCOT) and the Mid-Atlantic and Midwest (PJM). Unlike a traditional regulated Utility that earns a fixed, commission-approved return regardless of market conditions, Vistra sells power at wholesale market prices — prices that rise, structurally and meaningfully, when Demand surges and Supply cannot follow.

But the asset that separates Vistra from every other company in this conversation is its nuclear fleet — the second largest in the United States, forged through the landmark Acquisition of Energy Harbor in October 2023, which added the Beaver Valley, Davis-Besse, and Perry nuclear plants across Ohio and Pennsylvania to the already formidable Comanche Peak Facility in Texas.

Why does nuclear matter so profoundly? Because it delivers what every hyperscaler is quietly, desperately searching for: electricity that runs twenty-four hours a day, seven days a week, three hundred and sixty-five days a year — regardless of whether the sun is shining, the wind is blowing, or the weather has turned. Carbon-free. Predictable. Uninterruptible. In the language of Mukherjea, it is the perfect moat product — something that solves a real, structural, multi-decade problem that nobody else can replicate in any reasonable timeframe.

Building a new nuclear plant in America today costs upward of thirty billion dollars and takes fifteen years — a timeline confirmed by the painful, expensive experience of Georgia's Vogtle Units 3 and 4, the only new nuclear plants completed on American soil in a generation. No venture capitalist can fund that timeline. No hyperscaler can disrupt their way around it. The plants that exist today — the ones Vistra owns and operates — are, quite literally, national infrastructure that cannot be rebuilt from scratch.

When Microsoft signed a twenty-year power purchase agreement to restart Unit 1 of Three Mile Island specifically to power its AI data centres — a transaction announced in September 2024 and valued at over one billion dollars — it was not making a quirky green energy gesture. It was not a Public Relations exercise. It was an act of civilisational desperation dressed in polite corporate language — a confession, written in ink and signed by lawyers, that there is not enough power, that the queue for new power is too long, and that whoever owns existing power owns the future.

Vistra owns existing power. A great deal of it.

The Compound Machine, Quietly Humming

Mukherjea's most celebrated framework is the identification of Consistent Compounders — businesses with clean balance sheets, rising return on capital, and Earnings that grow with the quiet, relentless certainty of a well-tended garden in good soil. These are not businesses that dazzle. They are businesses that endure.

Vistra, astonishingly for a Business that emerged from Bankruptcy less than a decade ago bearing the scars of the old Texas utility TXU, is beginning to exhibit precisely these characteristics.

The company guides for nearly five billion dollars in adjusted EBITDA for 2025 — a figure that would have seemed fantastical to any analyst covering the legacy merchant power sector in 2016. Its nuclear fleet receives a direct government Subsidy of fifteen dollars per megawatt-hour under the Inflation Reduction Act's nuclear Production Tax Credit — translating to approximately six hundred million dollars a year flowing directly to the Bottom Line simply for keeping clean, reliable electrons moving through American wires. The Balance Sheet is deleveraging with disciplined speed. The Dividend is growing. Share Buybacks are ongoing. The stock delivered returns exceeding 230% in 2024 alone — making it, remarkably, the best-performing stock in the entire S&P 500 for that calendar year.

And yet, this is not a speculative technology bet. There are no hockey-stick Revenue projections anchored to assumptions that may or may not materialise. This is a cash-generative industrial fortress — one that happens to be positioned at the most critical choke point of the artificial intelligence economy, collecting a toll from the most powerful corporations on earth for the simple, irreplaceable service of keeping their servers alive.

Morningstar's Equity research, covering Vistra under its utilities and energy transition analysis, has noted that the structural repricing of American wholesale power — driven by data centre demand — represents a multi-year earnings catalyst for competitive generators with dispatchable, zero-emission assets. This is not a seasonal tailwind. It is a tectonic shift.

The Elegant Simplicity of the Thesis

Mukherjea has always insisted that the finest Investment theses can be stated in a single, clear sentence — free of jargon, free of complexity, free of the fog that lesser analysts use to disguise thin thinking. If you cannot explain why a business will be worth more in ten years in one sentence, you do not yet understand the business.

Here is Vistra's thesis in one sentence:

Artificial intelligence runs on electricity; America does not have enough electricity; Vistra owns electricity assets that nobody can build again; therefore, Vistra's pricing power will compound for many years to come.

That is it. No financial engineering. No narrative sleight of hand. No moonshot assumptions about markets that do not yet exist. Simply a business that owns something the world urgently, structurally, physically needs — and cannot easily get elsewhere.

A Word of Civilised Caution

Intellectual honesty demands we note that no fortress is entirely without a gate.

Power markets remain subject to regulatory intervention — state public utility commissions and federal bodies retain the authority to reshape the rules of competitive electricity markets in ways that could compress margins. Wholesale electricity prices, currently elevated by the surge in demand, could in theory soften if new generation capacity were to arrive faster than the current permitting timelines suggest is possible. Nuclear plants, magnificent as they are, carry the ever-present possibility of unplanned operational disruption, as any student of energy history will attest.

These risks are real. They deserve acknowledgement. They deserve to sit in your investment framework alongside the opportunity.

They are also, in this Writer's considered view, substantially outweighed by the structural demand surge of a magnitude that the American electricity system has never encountered in its hundred-and-forty-year history — a demand surge that is not cyclical, not speculative, and not dependent on a single company's execution. It is the aggregate consequence of every AI model trained, every inference served, every digital assistant answering every question, everywhere, simultaneously, indefinitely.

The Final Word

In Mukherjea's philosophy, the greatest businesses are not found by chasing what is fashionable. They are not found in the pages of technology magazines or the ticker tapes of momentum screens. They are found by asking a single, deceptively simple question: What does the world need so badly that it will pay almost any price to get it?

In 1850, that answer was steel. In 1950, it was oil. In 2025, the answer — hiding in plain sight, beneath the server racks and cooling towers of ten thousand humming data centres — is clean, reliable, dispatchable electricity.

Vistra Corp does not write code. It does not design chips. It does not train models or craft algorithms. It simply does what only it can do, at a scale that only it possesses, in markets where the laws of physics, the timelines of permitting, and the irreversibility of decades-old Capital Investment have conspired to make genuine competition nearly impossible.

In the quiet, unflashy language of enduring wealth creation, that is called a wonderful business.

And wonderful businesses, patiently held, with the conviction that comes only from truly understanding what you own — have a way of becoming extraordinary investments.

The electrons will keep flowing. The question is only whether you own a piece of the wire.