The Buffett
Rotation.

This is the headline of the entire filing. Buffett didn't merely add to an existing Google position — he more than tripled it, and simultaneously opened a brand new stake in the Class C shares. Two separate, deliberate filings pointing in exactly the same direction.

The conventional narrative heading into 2026 was that AI challengers would erode Google's search dominance and the advertising revenues underpinning the entire business. Buffett just bet the other way — loudly, with tens of billions of dollars. His read appears to be that Google is the AI infrastructure, not the target of disruption. Search, YouTube, Google Cloud, and DeepMind are compounding assets, not legacy liabilities.

When the most celebrated long-term investor alive concentrates into a single name at this scale, the institutional money flow effect is significant. Other allocators watch Berkshire's 13F filings as a directional signal. This move will be felt well beyond Berkshire's own books.

In May 2020, Buffett stood at the Berkshire annual meeting and publicly admitted he was wrong about airlines — sold every position at a loss, called the industry's future fundamentally uncertain. One of his most candid admissions of error on record.

Five years later, he is back. Not tentatively — 39.8 million shares is a committed, meaningful position. He chose Delta specifically: the strongest balance sheet in U.S. aviation, the highest mix of premium passengers, and the most resilient revenue model among the legacy carriers. This is simultaneously a sector rehabilitation call and a structural consumer behaviour thesis. Post-pandemic travel demand has permanently reset — spending on experiences over goods is not a sugar-rush rebound but a durable shift in how people allocate discretionary income.

Smaller in dollar terms, but the near-200% addition is unmistakably deliberate. The Times has quietly built one of the most defensible digital subscription businesses in media — over 12 million paying subscribers across news, games, cooking, and Wirecutter. In an era of AI-generated content at scale, editorial brands with earned trust command a genuine pricing premium. This is a classic Buffett durable-brand play — not a bet on journalism, but a bet on a subscription business that happens to produce journalism.

Adding to Lennar reinforces a thesis Berkshire has been building across multiple quarters: America's housing supply deficit is not a rate story, it is a structural reality. The undersupply of homes — particularly in entry-level and mid-market segments — provides builders with a demand floor that persists regardless of short-term rate noise. Not a trade. A view on the built environment of the United States over the next decade.

The most surprising buy of the quarter. Macy's is deeply out-of-favour — a legacy department store in a consumer spending slowdown. But Buffett has a long history of buying hated brands when the market prices the business and ignores the assets beneath it. Macy's owns significant real estate in premium urban locations — Herald Square in Manhattan being the most visible. This reads less like a bet on department stores, and more like a classic asset value at a discount position.

Selling either would be notable. Selling both in the same quarter is a statement. These are two of the most dominant toll-booth businesses on the planet — companies Buffett loved precisely because they clip a fee on every transaction without bearing the credit risk. He held them for years for exactly that reason.

Three forces appear to be converging. First, regulatory risk: interchange fee legislation has been advancing steadily. Second, structural disruption: real-time payment rails (FedNow), the emergence of stablecoins, and account-to-account payments are beginning to route around the Visa/Mastercard network in ways that weren't credible five years ago. Third, valuation: both stocks are priced for perfection at a moment when the moat is becoming harder to see clearly. The institutional money flow signal here is significant — capital that was parked defensively in payment names is now moving.

UNH was getting hammered when Buffett sold. The natural instinct for a long-term value investor is to hold through sector weakness, or buy more if the thesis is intact. He did neither. The thesis has changed, not just the price.

Managed healthcare faces a genuine structural inflection: sustained political scrutiny over claim denial rates, bipartisan pressure to reform Medicare Advantage reimbursement, and rising medical costs increasingly difficult to pass through to payers. These are not cyclical headwinds — they are the kind of regulatory forces that permanently compress an industry's earnings multiple. Buffett watched what happened to newspapers when the structural environment shifted. He appears to be reading the same signals in managed care, and moving before the re-rating is complete.

Selling Amazon while dramatically adding to Google is not a rotation out of tech. It is a quality-of-moat distinction within tech. Both are mega-cap platforms — but Buffett appears to be saying Google's advertising and search dominance is more deeply entrenched, more compounding, and more defensible than Amazon's hybrid retail-plus-cloud model. In a period of deliberate concentration, he is choosing the cleaner, deeper moat.

Domino's is a consumer discretionary name in an environment where middle-income spending is compressing — credit card delinquencies rising, savings rates low. In a tighter allocation environment, lower-conviction positions get cleared.

Charter is a more structural exit. Cable is a slowly declining infrastructure business — cord-cutting is permanent, fixed wireless is accelerating, subscriber loss is irreversible. Buffett learned the hard lesson about legacy media with newspapers. He is not repeating it with cable, regardless of what the valuation looks like.