Strip away everything that is not a reason to sell (the price rose, the price fell, you're bored, a pundit is bearish, you want to "book profits" before year-end, the chart looks toppy) and exactly five legitimate reasons remain. Each is about the business, the value, the alternative, or your portfolio; none is about your cost basis or the direction of the last price tick. Learn them as a closed set, because their power is partly that they are exhaustive: if a proposed sale does not map cleanly onto one of the five, it is probably a bias wearing a costume, and the default is to do nothing (Terry Smith's third instruction, M6.02).
What is the fresh-money test?
The master frame that unifies all five is the fresh-money test from M6.04, applied at the position level: ignoring what I paid and what the position is now worth to me emotionally, if I held cash equal to this position's current market value, would I buy this business, at this price, at this weight, today? The five reasons are simply the five distinct ways the answer can be "no":
- No, because the business is no longer what I thought (the thesis broke). See selling on a broken thesis
- No, because I was wrong about the business from the start (a mistake). See cutting losses
- No, because the price now exceeds the value by too much (margin of danger). See selling above value
- No, because I'd rather own something else with this money (opportunity cost). See switching to a better opportunity
- No, not at this weight; I'd own less (beyond risk tolerance → trim). See trimming a position
Advertisement
Which reasons to sell ignore tax, and which call for a trim?
Two structural distinctions organize the five and you should carry them as the backbone of everything that follows. The first is tax-blindness. Reasons 1 and 2, a broken thesis and a recognized mistake, are non-negotiable and tax-blind: when the reason to own is gone, you sell regardless of the tax bill, because holding a broken or mistaken position to defer tax is letting the tax tail wag the risk dog, and the impairment risk dwarfs the tax. Reasons 3, 4, and 5 (overvaluation, a better idea, and sizing) are discretionary and tax-modulated: they are judgment calls about degree, and the after-tax cost of acting (see tax-aware selling) legitimately raises the bar. The second distinction is action: reasons 1, 2, and often 3 point to a full exit; reason 4 points to a switch; reason 5 points to a trim, not an exit. Keep these straight and most sell decisions resolve themselves.
| # | Sell reason | One-question decision test | Tax | Typical action |
|---|---|---|---|---|
| 1 | Thesis broken | "Has a load-bearing pillar I wrote down factually broken — not merely the price fallen?" | Blind | Exit |
| 2 | Mistake recognized | "Given what I now know, was my original analysis wrong — not just unlucky?" | Blind | Exit; never average down |
| 3 | Price ≫ value | "Does today's price imply a forward return below my hurdle even under the bull case?" | Modulated | Trim/exit (compounder caveat) |
| 4 | Better opportunity | "Is B's risk-adjusted forward return higher than A's by more than tax + costs + the conviction discount?" | Modulated | Switch (only on a wide gap) |
| 5 | Beyond risk tolerance | "Does the position exceed my pre-set maximum weight?" | Modulated | Trim to the cap (not exit) |
The five reasons are the general law, and they hold whatever the company is. Lynch's category triggers, which set a different numeric sell level for a slow grower, a stalwart, a fast grower, a cyclical, a turnaround and an asset play, are the specialisation of reason 3 to a company type, and E11.06 runs them as a drill with ten companies and ten computed triggers.