Why would Buffett keep such a small position instead of adding to it?

The filing does not say, and guessing at Berkshire's reasoning would be speculation rather than something this data supports. What is published is the outcome: 15.25 years after opening the position, it remains 0.08 percent of the disclosed book despite a 1125.51 percent return, which is consistent with the position simply never having been added to at the scale of Buffett's other larger holdings.

Does a small position with a huge return matter for the overall portfolio?

Not much, by design. Weighted return sizes every position by how much of the book it represents, so a position worth 0.08 percent of the book barely moves the aggregate number no matter how large its own percentage gain. A reader focused only on Buffett's 98.8 percent weighted return would never see MA's 1125.51 percent at all.

Guide · 6 min read

Warren Buffett's best trade returned 1125 percent. It is 0.08 percent of his book

Across Buffett's, Burry's, and Li Lu's disclosed per-trade records, the single largest percentage gain belongs to Buffett's Mastercard position, up 1125.51 percent since 2011. It is also 0.08 percent of his current book, a rounding error next to Apple's 28.4 percent.

Updated 2026-09-16

Warren Buffett's disclosed trading record lists a Mastercard position, ticker MA, opened in the first quarter of 2011, still held, up 1125.51 percent. Across the three tracked investors on this site with a full per-trade breakdown published, Buffett, Michael Burry, and Li Lu, no other single disclosed trade comes close: Li Lu's best is GOOG at 250.36 percent, Burry's is GME2 at 113.96 percent. Buffett's MA position beats both by a wide margin.

It is also 0.08 percent of the book Buffett currently reports. The position's disclosed cost basis, 197.27 million dollars against an average cost of 36.53 dollars a share, is real money by any ordinary measure. Next to the rest of what Buffett discloses holding, it rounds to nothing.

Two Buffett positions, share of disclosed book

Apple is Buffett's largest position by weight. Mastercard is his best position by percentage return. The two lists do not share a name.

Tenachine's guide on 13F concentration already covers Apple's 28.4 percent weight, the largest slice of Buffett's disclosed book. That guide is about what a big position does and does not tell a reader. This is closer to the opposite case: a position that has done better than anything else on the page, sitting small enough that it would be easy to miss entirely in a filing full of far larger numbers. A 13F reports dollar value and share count. It does not sort by how well a position has actually performed, and nothing on the filing itself would point a reader toward MA over any of Buffett's other 139 disclosed trades.

Held 15.25 years, MA's annualized return, its IRR, comes out to 17.86 percent, a strong number in its own right and nowhere near as dramatic as the 1125.51 percent headline. Tenachine's guide on Li Lu's IRR gap covers why a long holding period compresses a large total return into a smaller annualized one, and the same mechanism applies here without needing to be re-derived. The point of this guide is narrower: MA is Buffett's best trade by either measure, return or IRR, and its weight in the book gives no hint of that.

Buffett's win rate across all 140 disclosed trades is 68.6 percent, and his weighted return, sized by how much of the book each position represents, is 98.8 percent. MA contributes almost nothing to that weighted figure precisely because its weight is 0.08 percent. A trade can be the single best-performing line on an entire disclosed record and still barely register in the number most often quoted to summarize that record, because weighted return by design counts a position in proportion to its size, not its percentage gain.

None of this says MA was a smarter buy than Apple, or that a reader should weight small positions more heavily when reading a 13F. It says the opposite of what a filing's own layout suggests: the biggest number on the page, dollar value, is not sorted by the number a reader might actually care about, how well the position performed. Checking a position's return separately from its size is the only way to find a result like this one, and a filing on its own gives no reason to go looking.