Buffett told us to sell Berkshire — he just didn’t put it that way

To grow 10% a year, Berkshire Hathaway now has to add about US$100 billion of value. Every year.

That’s the arithmetic of being a US$1 trillion company, and it’s the main reason Global Masters Fund, the listed investment company I manage, has started selling some of its Berkshire. It has been our largest position since we started. For the past few years we’ve been moving the rest of the portfolio towards the kind of company Berkshire was before it got big. Trimming Berkshire itself has made us think hard about when to sell a great business, and the answer applies well beyond Berkshire.

A holding can get too big in two ways. It can outgrow its own market, or it can outgrow your portfolio. Berkshire has done both. Most investors have at least one holding that has done the second.

Too big to grow quickly

Buffett saw this coming. In his February 2024 letter he wrote that Berkshire’s size left “no possibility of eye-popping performance”. Only a handful of US companies were large enough to make a difference to it. He expected Berkshire to do a bit better than the average American company, and anything more, he said, was wishful thinking.

The record shows why. From 1965 to 2024 Berkshire compounded at 19.9% a year, close to double the S&P 500’s 10.4%. Most of that came when it was much smaller. Over the past 20 years the B shares have returned about 11% a year, much closer to the market.

Go back to that US$100 billion. A business worth US$3 billion needs to add US$300 million to grow at the same rate. One target is far easier to hit than the other.

None of this makes Berkshire a poor business. Greg Abel has started well: buybacks have resumed and the handover from Buffett went as planned. It means returns from here are likely to look more like the market’s than like Berkshire’s past. That’s a reasonable outcome. It may just not be the one you bought it for.

Too big for your portfolio

Most Firstlinks readers don’t own Berkshire, but plenty own the local version: a great Australian company bought years ago that has quietly grown into a big slice of the portfolio. Nobody decided that. It just happened.

CBA is around 10% of the ASX 200 on its own, so an index investor already has a tenth of their money in one bank. CSL shows the other side. For years it was the kind of quality compounder investors were comfortable holding in size. Then its share price fell about 70% in the two years from July 2024.

I’m not making a call on either. CSL has bounced since its August result, and CBA has kept rewarding its holders. The point is what concentration does to your result when a great business has a bad stretch. Concentration that was earned feels safer than it is.

Buffett has done it himself

If trimming a great business feels like disloyalty, look at what Berkshire did with Apple.

At the end of 2023 Berkshire’s Apple stake was worth about US$174 billion, against a cost of around US$31 billion. Within a year it had sold roughly two-thirds of it. At the 2024 annual meeting Buffett called Apple an even better business than Coca-Cola or American Express, and said it would likely remain Berkshire’s largest investment. He pointed to tax as one reason for selling. Whatever the full reasoning, admiring a business and holding all of it are two separate decisions.

It cuts both ways, and Buffett has been honest about that too. He trimmed Apple in 2020 and later said that was probably a mistake. Trimming isn’t always right. It just shouldn’t be off the table because a holding has been good to you.

Four questions we asked

We didn’t want to sell on instinct or hold out of loyalty, so we worked through a few plain questions. They apply to any long-held winner.

1. What can this business realistically return from here, given its size? Not what it returned in the past. A great business can be a mediocre investment from today if it has run out of room.
2. Is my position size a decision or an accident? If you wouldn’t buy it at this weight today with fresh money, the weight is probably an accident.
3. Where would the next dollar work harder, under the same discipline? Selling only makes sense if there’s somewhere better to go. For us, the principles that made Berkshire great when it was smaller still apply: pricing power, high returns on capital, reliable cash flow and a long holding period. They work best where size isn’t in the way, so we’re moving gradually towards quality mid-sized companies.
4. Can I do it gradually, and what will the tax cost? Selling in stages lowers the risk of getting the timing badly wrong. Tax is the harder part. The new CGT rules from 1 July 2027 will make selling a long-held winner more expensive for many investors, as Shani Jayamanne’s recent Firstlinks piece shows, and the natural reaction is to sit tight. That’s a costly mistake of a different kind: you avoid a tax bill but keep a risk you no longer want. Tax should shape how you trim. It shouldn’t decide whether you do.

Not a call on Berkshire

We still own a lot of Berkshire and expect to for some time. We’re not selling because we think it’s overvalued. There’s an obvious irony in selling while Berkshire buys back its own shares, but we don’t see a contradiction. Berkshire is judging value. We’re judging fit. Our shareholders expect returns that a US$1 trillion company is unlikely to deliver on its own, and having that much of our money in one business, however good, was a risk we no longer wanted to carry.

Often the right call with a great holding is to leave it alone. But it’s worth asking, honestly and regularly, whether it’s still the best place for the next dollar. For us, after a long time, the answer changed. Not because Berkshire got worse. Because it got big.