When to sell: the decision nobody teaches

Every course teaches you what to buy. Almost none teach you when to sell, which is strange, because selling is where most of the emotional difficulty lives and where a great deal of the return is decided.

The principle everything follows from

The market has no memory of what you paid. Your purchase price is a fact about your history, not about the company.

The only question that matters is: knowing what I know today, at today's price, would I buy this? If the answer is no, the reason you own it is not investment. It is reluctance to acknowledge a decision.

Three good reasons to sell

1. The thesis broke

You bought for specific reasons. One of them has stopped being true. The competitive advantage eroded, the debt became unmanageable, management changed and the new team is doing something different, the regulatory environment shifted.

This is the cleanest reason to sell, and it is only available to people who wrote down the thesis at purchase. Without that note, "the thesis broke" is indistinguishable from "the price fell and I am uncomfortable".

2. You found something clearly better

Capital is finite. If a materially better opportunity exists and you have no cash, funding it by selling your weakest holding is rational.

The test is that the new idea must be clearly better, not marginally more exciting. Frequent switching between roughly equivalent ideas is a reliable way to convert a decent portfolio into transaction costs.

3. The position outgrew its place

A holding that has risen until it dominates the portfolio should be trimmed back toward its target weight. Not because you have lost faith, but because concentration risk does not care how well the position has done so far.

Four bad reasons to sell

1. It went up

Selling purely because a holding has risen means systematically disposing of the things that are working. Long-term returns in most portfolios come from a small number of holdings that did very well over a long period, and every one of them looked like it had "run too far" at some point.

2. It went down

A falling price is information about sentiment. It may or may not be information about the business. Check which before acting, and check it against your written thesis rather than against how you feel.

3. To get back to what you paid

Waiting to "break even" before selling is the purest form of letting your purchase price make the decision. The company does not know what you paid, and the price at which you become willing to sell says nothing about what the business is worth.

4. Someone told you to

If a tip was not a good enough reason to buy, it is not a good enough reason to sell.

Decide before you own it

The practical solution is to write the sell conditions at the moment you buy, when you have no money at stake and can think clearly.

Something as simple as: "I will sell if debt-to-equity exceeds X, if the dividend is cut, if the segment that drives profit stops growing for two consecutive years, or if this exceeds 15% of the portfolio."

Then, when the moment arrives, you are following a decision you already made rather than making one under pressure. That is the entire technique, and it is worth more than any amount of additional analysis.

On taxes and costs

Selling has costs: commission, and potentially capital gains tax. Factor them in, but do not let them hold you in a broken position. Paying tax on a gain is a better outcome than watching the gain disappear to avoid it.