He went all-in on streaming. The filing that proved it landed after he had gone.

The Action

Eighty-four days later he sold the entire position. His own letter of that day put the damage at four percentage points of the funds' year-to-date return. The only 13F that ever reported the holding was filed on May 16, 2022, twenty-six days after the exit, listing 3,109,965 Netflix shares worth 1,164,962 thousand dollars as of March 31.

The Gap

The reasons given for buying were structural: moat, scale, pricing power, culture. The reason given for selling, twelve weeks later, was a single quarter and a plan to add advertising.

The Read

Say the honest part first, because it is the part most coverage skipped. Nobody made him announce either move. He published the purchase the week he made it, published the exit the day he made it, and put a number on his own loss in his own letter rather than waiting for anyone to work it out. He also named the rule he was following: act promptly when new information contradicts the original thesis. Measured against that rule, he did exactly what he said he does.

The tension is between the two letters, and both halves are his own words. January listed seven reasons to own the company, and every one of them was structural: recurring revenue, a best-in-class management culture, economies of scale, pricing power, margin expansion, an improving cash profile, a widening competitive moat. None of those changed by April. What changed was one quarter of subscriber numbers and an announcement that the company would add advertising and go after password sharing.

There is a sharper echo inside it. The January letter praised Netflix for its "remarkable pivot from DVD rental by mail, to video streaming", holding up the company's willingness to remake its own business model as evidence of the management quality worth paying for. In April, the company remaking its business model was the stated reason the future had become too hard to predict.

Then the paperwork, which is the reason this one belongs on the record at all. The purchase happened in January, so it missed the year-end filing. The exit happened on April 20, three weeks before the quarterly filing deadline. The result is that the only 13F in which Netflix has ever appeared under this manager was filed on May 16, reporting a position as it stood on March 31, which he no longer held on the day anyone could read it. Nothing improper occurred. That is simply what a filing is: a photograph of one day, published six weeks later. Anyone reading that document in May as a picture of what he owned was reading something that had stopped being true twenty-six days earlier.

He is back. Netflix is absent from every quarterly filing from March 2025 through March 2026, then appears in the filing for June 30, 2026 at 13,081,465 shares. Coverage put that position at about 3.15 million shares, which is close to the 2022 figure and is not what the document says. Netflix split its stock ten for one in November 2025, so the 2022 holding is worth 31,099,650 shares in today's terms. The new position is a little over two fifths the size of the one he called all-in.