So I don't think it's enough to evaluate Coke solely on its nominal or CAGR return over the past 15 years. Part of its value has been functioning as a tax-efficient reservoir of capital. The money that otherwise would have gone to Uncle Sam stayed inside Berkshire's compounding engine, continuing to earn returns for shareholders.

Though true, I wouldn't overstate the case. One merely has to compare the after tax proceeds of selling out on any given day to other current opportunities available at scale.

I presume a very significant reason that a sale hasn't been considered is not that it's a great "forever business", since it isn't any more, and hasn't been for a very long time. They might still arguably count on the "forever" front, but there are people old enough to drive who were born after the last time shares rose in real intrinsic value. I'm sure head office is well aware of all that. The shares have been falling in real intrinsic value for ages, and the dividend is tasked with the double job of making up for that slide and providing a modest positive return on top.

Rather, I assume head office has left it alone in large part because there wasn't a decent opportunity available at the required size, longevity, and price. And, OK, they're not going anywhere. No need to sell unless/until you have a better idea for the money--it's not as if head office is short of cash.

As an aside, don't get carried away with the "buybacks are great for us" trope in this instance. Not only have Coke been destroying capital by overpaying for a flatlined business when they do so, but also the share count is, if I'm not mistaken, actually higher than it was about a decade ago. It depends on the precise baseline date you choose, but it certainly hasn't shrunk materially.

Jim