When to Sell: Almost Never

"I don't want a lot of good investments. I want a few outstanding ones." — Philip Fisher

Fisher's most under-appreciated idea is that selling too early is the single greatest destroyer of long-term compounding. He allowed only three legitimate reasons to sell:

1. You made a mistake — the original reasoning was wrong.
2. The business has genuinely deteriorated — structurally, not just for a quarter or two.
3. You have found a clearly better opportunity and need the capital for it.

Notice what is not on the list: the stock has doubled, it looks "fully valued", the market feels expensive, or you simply want to "lock in gains". These are the reasons most investors sell their best holdings — turning what might have been a 30-fold return into a three-fold one.

The arithmetic is unforgiving. A business compounding at 25% a year multiplies roughly five-fold over five years, but far more over fifteen or twenty. Selling early to "bank a profit" can quietly forfeit the largest part of the eventual return. The decision to sell a true compounder usually costs more than any buying decision ever will.

Illustrative example: the cost of selling early

Consider a sum invested in a business compounding at a high rate. The value after five years looks impressive — but the value after twenty years is many times greater again. An investor who sold at year five to "lock in a triple" gave up the overwhelming majority of the eventual gain. That gap is the real cost of premature selling.