The price on Dexter Shoes in 1993 was $433 million, paid in Berkshire Hathaway stock. Buffett later put what that New England shoe company cost Berkshire shareholders at $3.5 billion.
He has spent decades writing losses like that one into shareholder letters and repeating them out loud at annual meetings, which makes his record unusually easy to read in both directions.
Eight Cases and the Figures Buffett Attached to Them
| Investment | When | Figure cited |
|---|---|---|
| Berkshire Hathaway (textiles) | Buying began 1962 | $200 billion estimated cost |
| Dexter Shoes | 1993 | $3.5 billion cost to shareholders |
| ConocoPhillips | 2008 | Several billion dollars |
| Lubrizol | 2011 | $9 billion purchase price |
| Tesco | 2014 letter | $444 million estimated loss |
| Precision Castparts | 2016 | $11 billion loss for the company in 2020 |
| Amazon | Never bought | No figure given |
| March 2020 selloff | 2020 | S&P 500 down about 34% from its high |
Paying With Stock Made Dexter Shoes Far More Expensive Than the Sticker Price
The $433 million Buffett handed over for Dexter Shoes wasn't cash. It was Berkshire Hathaway stock, and that distinction is why the final tally landed at $3.5 billion for shareholders.
When an acquirer pays with its own shares, the real price isn't fixed at closing. It keeps climbing for as long as those shares keep compounding in someone else's hands.
Buy a business that fizzles with currency that appreciates, and the gap between the two widens every year.
His explanation of why the business itself failed was short. Writing to shareholders in 2007, he said the durable competitive advantage he thought he had bought disappeared within a few years.
Precision Castparts and the Price Buffett Says He Overpaid
Buffett added the metal fabrication company Precision Castparts to the Berkshire portfolio in 2016, drawn to its standing in aerospace. Then the pandemic hit the sector in 2020, and the company recorded an $11 billion loss.
His account of it left no room for anyone else. "I paid too much for the company," he wrote to shareholders, adding that no one had misled him in any way and that he had simply been too optimistic about the company's normalized profit potential. That's a mistake about price, not about what the business did.
Buying Near the Oil Peak, Then Sitting Out the March 2020 Low
Two of the errors are pure timing, and they run in opposite directions.
In 2008, with oil trading around $150 a barrel at its mid-year peak, Buffett bought a large position in ConocoPhillips. Energy prices then collapsed. Writing in 2009, he said he had bought when oil and gas prices were near their high and that he in no way anticipated the dramatic drop in the second half of the year. The timing, he said, cost Berkshire several billion dollars.
March 2020 was the mirror image. As COVID-19 closed schools and businesses, the S&P 500 fell about 34% from its previous record. Investors who bought into that drop got their discount; Buffett didn't take it. At the April shareholder meeting, he told the room he had totally missed the opportunity and totally messed up in March of 2020.
The Tesco Loss He Blamed on Dawdling
At its largest, Berkshire's position in the UK grocery chain Tesco ran to 415 million shares. Buffett lost confidence in the company and sold 114 million of them, leaving the rest in place.
In his 2014 letter he wrote that an attentive investor, to his embarrassment, would have sold earlier, and he called dawdling the big mistake with the investment.
Berkshire eventually left the position entirely. Buffett estimated that waiting cost $444 million.
Admiring Amazon for Years Without Buying It
Not every error involves money leaving the building. Buffett passed on Amazon repeatedly, and in 2017 he said he should have bought it long ago because he had admired it long ago.
His reasons for not doing so were that he didn't understand the power of the model as it developed, and that the price always seemed to more than reflect that power at the time. He described it as one he missed big time.
The second reason is the familiar trap for value-minded buyers. A company can look expensive against today's earnings every single year while the earnings themselves keep moving.
Waiting for a better entry point is a coherent discipline, and it can also mean never buying at all.
Lubrizol Closed for $9 Billion, Then the Conflict Surfaced
Berkshire acquired the chemical manufacturer Lubrizol in 2011 for $9 billion. David Sokol, at the time widely seen as a leading candidate to succeed Buffett, bought Lubrizol shares and only afterward pitched the acquisition to him.
Buffett learned of the purchases after the deal was under way. Sokol made money on the transaction and resigned from Berkshire.
At the 2011 shareholders meeting, Buffett called the episode sad for Berkshire, sad for Dave, and still inexplicable in his mind. The business itself stuck: as of 2022, Berkshire still listed Lubrizol among its subsidiaries. What went wrong here was the process that brought the deal to him.
The Textile Company He Called His Dumbest Stock Ever
In 2010, Buffett named Berkshire Hathaway itself as the dumbest stock he ever bought. He started buying in 1962, when the textile operation had already been sliding for years, and kept accumulating shares.
A Berkshire manager later offered to buy him out, and the two settled on a price, but the formal offer that arrived was slightly lower. "He chiseled me," Buffett recalled.
His response was to buy a controlling interest and fire the manager. As he put it, he had now committed a major amount of money to a terrible business, one that spent years carrying the anchor of its textile assets.
He also noted that Berkshire became the base for nearly everything he did afterward. His estimate of what the decision cost: $200 billion.
The Largest Figures Here Are Opportunity-Cost Estimates
A figure like $200 billion, or the $444 million on Tesco, isn't a line on a statement. It's an estimate of what a different choice would have produced, a comparison against a version of history that never happened, built on assumptions about returns, timing and reinvestment.
The same caution applies to any opportunity-cost claim, including the ones investors make about their own portfolios. Someone who sold a position early can calculate a "loss" against the peak price, but the calculation assumes they would have held all the way up and sold at exactly the right moment.
Realized losses and hypothetical ones are different animals, even when both are quoted in dollars.
Prices Agreed To and Moves Delayed
Read across the eight, and the costliest items aren't businesses Buffett failed to analyze. They're prices he agreed to and moves he put off: overpaying for Precision Castparts, buying oil near a peak, holding Tesco too long, waiting out Amazon, and hanging onto a declining textile maker out of spite.
The one case where the analysis of the business held up, Lubrizol, is still a Berkshire subsidiary, and the thing that went wrong there had nothing to do with the chemicals.
