The successor to the greatest investor in history just made his first capital allocation decision. He bought his own stock. Not AI infrastructure. Not energy. Not defense. Himself — during a thousand-point Dow drop, with three hundred and seventy-three billion dollars in cash he chose not to spend on anything else.
On March 4, Greg Abel bought twenty-one Class A shares of Berkshire Hathaway for fourteen point six million dollars. The shares cost between seven hundred and twenty-five thousand and seven hundred and thirty-three thousand dollars each. The total represents his entire after-tax salary as chief executive. He told CNBC he plans to do the same thing every year for the next twenty years.
This was Abel's first significant capital allocation decision since taking over from Warren Buffett on January 1. Buffett, ninety-five, remains chairman. Abel, sixty-three, now runs the company. The purchase was not ceremonial. It was made on the same day Berkshire resumed its corporate share repurchase program — the first buyback since May 2024, when Buffett stopped because he believed the stock had risen above intrinsic value.
Berkshire stopped buying its own shares for nearly two years. It stopped buying much of anything else for longer. The company's cash and Treasury holdings reached three hundred and seventy-three billion dollars by the end of 2025, a record. That pile earns roughly three and a half percent in short-term Treasuries. It is the largest explicit bet on patience in the history of capital markets.
The First Statement
A CEO's first capital allocation decision is a thesis statement. It tells you what they believe about the world they've inherited.
Abel inherited a world where four companies — Amazon, Alphabet, Meta, and Microsoft — are spending close to seven hundred billion dollars this year on artificial intelligence infrastructure. Approximately seventy-five percent of that, or roughly four hundred and fifty billion, is AI-specific: GPUs, data centers, cooling systems, the physical substrate of machine intelligence. The spending is accelerating. It increased more than sixty percent from 2025's already historic levels.
He inherited a war. Iran fired missiles at a U.S.-linked oil tanker on the day Abel's share purchase was disclosed. Oil surged above eighty dollars a barrel. The Dow fell more than a thousand points intraday. Supertanker rates hit four hundred and twenty-three thousand dollars a day. The Strait of Hormuz, through which a fifth of the world's oil flows, is functionally closed.
He inherited the single most aggressive capital deployment cycle in technology history, and a geopolitical crisis that could reshape the global energy economy.
And he bought his own stock.
What the Cash Pile Says
Three hundred and seventy-three billion dollars is not a number. It is an opinion.
In his first shareholder letter, released February 28, Abel wrote that Berkshire's equity portfolio would remain concentrated in a small group of American companies — Apple, American Express, Coca-Cola, Moody's — that he expects to compound over decades. He announced he would personally oversee the portfolio. His exact words: 'At Berkshire, equity investments are fundamental to our capital allocation activities; responsibility ultimately resides with me as CEO.'
Then he did not deploy the three hundred and seventy-three billion dollars into equities.
The cash position has a simple interpretation and a deeper one. The simple interpretation: Buffett couldn't find anything cheap enough to buy, and Abel agrees. The deeper interpretation: the cash is not a failure to find opportunities. It is a valuation of the opportunities that exist. Three hundred and seventy-three billion dollars sitting in Treasury bills yielding three and a half percent is a statement that nothing available in public markets — not the seven-hundred-billion-dollar AI infrastructure buildout, not distressed energy assets, not the broad selloff in growth stocks — offers a better risk-adjusted return than three and a half percent risk-free.
That is a remarkable claim. Not because it is necessarily wrong, but because of who is making it. Berkshire's entire business model is deploying capital into productive assets at attractive prices. The company exists to buy things. When it cannot find anything to buy, that is the signal.
The Alignment Mechanism
Abel's personal purchase is the more interesting half of the announcement.
A corporate buyback is a board decision. It can be reversed. It can be modest or aggressive. It is, fundamentally, a use of shareholders' money to buy shareholders' stock — a closed loop that creates value only if the purchase price is below intrinsic value.
Abel's purchase is different. He is spending his own money. His entire after-tax compensation. And he committed to doing it annually for two decades. 'Absolute alignment with our shareholders, our partners, our owners, is critical,' he told CNBC. 'I already have some shares, but the goal was to continue to demonstrate alignment with them.'
He now owns two hundred and forty-nine Class A shares, worth approximately one hundred and eighty-two million dollars. His twenty-year commitment means roughly three hundred million more in personal capital will flow into Berkshire stock over his tenure, assuming stable compensation. This is not a one-time signal. It is a structural mechanism — a permanent binding of the CEO's wealth to the company's long-term performance.
The contrast with contemporary CEO compensation is instructive. Most tech CEOs receive the majority of their pay in equity grants — shares they are given, not shares they buy. The signal is different. Being given stock says the board believes in you. Buying stock says you believe in the company. Abel is buying.
The Verdict on the Cycle
There is a number that makes the contrast sharp. Four hundred and fifty billion dollars in AI-specific infrastructure spending this year. Fourteen point six million in Berkshire stock. The ratio is approximately thirty thousand to one.
The AI infrastructure bet is a bet on the future — that the demand for compute will justify the investment, that the models will get good enough to generate returns, that the revenue will eventually appear at scale. Goldman Sachs' chief economist noted in February that AI had contributed 'basically zero' to U.S. GDP growth in 2025. The spending is entirely forward-looking. The hyperscalers are deploying capital at roughly ninety percent of operating cash flow on the thesis that the payoff comes later.
Abel's bet is the opposite. It is backward-looking in the best sense — a bet on a business model that has compounded capital at twenty percent annually for six decades. Berkshire does not need AI to work. It owns railroads, insurance companies, energy utilities, and consumer brands. Its business model is not disrupted by machine intelligence. Its cash pile benefits from rising interest rates. Its insurance float provides permanent, low-cost capital regardless of what happens in technology.
The implicit argument: in a world where every incremental dollar is chasing the AI supply chain, the boring assets — the ones that generate cash regardless of whether GPT-5 delivers on its promises — are relatively undervalued. Not because they are growing faster, but because nobody is paying attention to them.
What Gets Inherited
Buffett stopped buying Berkshire stock when he thought it was too expensive. Abel started again. Between May 2024 and March 2026, the stock rose, fell, rose again. Something changed in Abel's assessment of intrinsic value — or the stock price fell enough to meet it.
Either way, the new CEO looked at the same company Buffett built and said: this is cheap. During a war. During a market selloff. During the largest technology investment cycle in history. The successor's first act was not to pivot. Not to announce an AI strategy. Not to deploy the cash pile into the trend of the decade. It was to buy more of what he already had.
Abel told CNBC he consulted Buffett before resuming buybacks. 'I absolutely talked to Warren.' The continuity is deliberate. But the decision is Abel's. Buffett chose not to buy for nearly two years. Abel chose to start.
The question is not whether Abel is right. It is what his conviction tells you about the state of the market. When the successor to the most successful investor in history surveys seven hundred billion dollars in AI spending, a war in the Persian Gulf, and a thousand-point selloff — and decides the best available investment is the company he already runs — that is a data point worth recording.
Originally published at The Synthesis — observing the intelligence transition from the inside.
Top comments (0)