The Time Warren Buffett Sold 50,000 Puts on Coca-Cola

Buffett called derivatives financial weapons of mass destruction. He wrote it in a shareholder letter and the line stuck, which makes what happened in the spring of 1993 worth sitting with.

Coca-Cola was trading around $39 a share. Berkshire already owned a large position and wanted more, but not at $39. Buffett had decided $35 was the right price. So rather than place a limit order and wait, he sold 50,000 put contracts struck at $35, expiring in December, collecting $1.50 a share.

Fifty thousand contracts covers five million shares. At $1.50 each that is $7.5 million, paid up front, his to keep no matter what happened next.

Coca-Cola never went below $35. The contracts expired. Berkshire bought nothing, and kept the money.

A note on where these numbers come from

The trade is recounted in Andrew Kilpatrick's Of Permanent Value: The Story of Warren Buffett, chapter 34, and it has been repeated widely since. It does not appear as a line item in a Berkshire filing you can pull up and check. That is a weaker chain of evidence than a 10-K and you should know which one you are reading. The figures are consistent across every retelling, but they all trace back to the same place.

What the trade actually was

Strip out the scale and the mechanics are ordinary.

Selling a put obliges you to buy the stock at the strike if the buyer exercises. They only exercise if the stock falls below it. So there were two ways this could end. Coca-Cola drops under $35 and Berkshire buys five million shares at a price it had already decided was fair. Or Coca-Cola holds up, the contracts lapse, and Berkshire keeps $7.5 million for a promise nobody called in.

The second thing happened.

Why the boring ending is the interesting part

Most retellings get this backwards. They frame it as Buffett cleverly buying the dip. He did not buy anything. The story is that he was paid seven and a half million dollars for a commitment that was never called.

Put selling has a shape that people coming from directional trading find hard to accept. The overwhelming majority of contracts sold at a sensible distance from the money expire worthless, and every single one of those is the strategy doing its job. Nothing happens. The premium lands, the obligation lapses, you do it again.

If you are used to judging a position by how far it moved, this feels like failure. It is not. A long run of quiet, uneventful expiries is the mechanism, not a sign you were too timid.

The condition that makes it work

There is one requirement holding all of this up, and it is the one people skip.

Berkshire sold puts on Coca-Cola because Berkshire wanted more Coca-Cola at $35. The strike was not selected to squeeze out the fattest premium available that week. It was a price at which owning the business made sense on its own terms, and the premium was what the market happened to pay for a promise to buy there.

Reverse the order and the logic collapses. Start from the premium, work backwards to a strike, accept whatever underlying is attached to it. Now the trade only survives if you are never assigned, and you have no answer for the case where you are.

That case is a loss. Worth saying plainly, because a lot of writing on this subject goes soft here. When a put you sold expires in the money, you lost. Not a discount. Not a strategic entry. Not an improved cost basis. A loss. Any framework that describes assignment as a good outcome has quietly stopped keeping honest score.

What does not transfer

Berkshire's position is not yours and pretending otherwise would be silly.

Fifty thousand contracts is a commitment to buy five million shares, roughly $175 million at that strike. Berkshire could absorb that without blinking. It also had decades of familiarity with the business, a research operation, and a balance sheet that made the obligation genuinely comfortable rather than theoretically comfortable.

What does transfer is the order of the decision. Establish first that this is a company you would accept at this price. Then ask whether the premium is worth the promise. Premium first is how people end up owning things they never looked at.

Why this one is worth knowing

Plenty of trades make money. This one is worth knowing because the most famous long-term investor of the last century used a structure that gets dismissed as a retail income gimmick, used it exactly the way its own logic prescribes, and had it pay off in the least dramatic way available.

Nothing happened. He got paid anyway.