Growth investors are paying for a decade of moat protection. AI's real threat was never to earnings. It's to the calendar.

🧮 What a moat is actually worth: Chamath runs the maths. No growth, and you are paying 12 times cash flow. The S&P trades near 20.

🏰 The bull's steel-man: One author's Microsoft case for buying the duration, not the discount.

🔎 Where the premium hides: Screen for the companies carrying the biggest terminal value bets.

Most of a growth stock's value is not this year's earnings. It is cash flows a decade out, what Mauboussin at Morgan Stanley calls the present value of growth opportunities, routinely 60 to 70% of the whole thing. That number is not really a bet on the business. It is a bet on how long the business stays hard to copy.

Which is why AI is a stranger threat than the bear case makes out. It does not need to break your company's earnings. It only needs to shorten the distance between a good business and a commoditized one, and the multiple falls apart on its own. The gap between 12 times and 20 times is the market quietly assuming that distance holds for ten years.

Here is the part worth arguing. That assumption is not wrong everywhere, it is just applied evenly, which cannot be right. AI is compressing duration for some businesses and extending it for others, and the market is charging both the same rent. Some of the gap has already closed this year too, with earnings outgrowing price rather than moats breaking.

How many years of moat is your largest holding priced for, and do you actually believe it?