Why 90% of Your Profit Comes From 10% of Your Trades

Last week this site argued that you should take your profit too soon. The arithmetic was sound and the conclusion held.

This week it argues the opposite, and the arithmetic is the same code.

Both are true. Which one applies to you is the most consequential unanswered question in your trading, and it takes an evening with your journal to settle.

Sort your winners

Take a trend system. Thirty percent of trades win. The winners are not all alike: most are modest, some are good, and a very few are extraordinary. Model that with a fat-tailed distribution averaging 3R.

Now line up three hundred thousand winners, largest first, and ask where the money actually is.

Thirty-three percent of gross profit from a tenth of the winners. Striking, but gross profit is the wrong denominator, because it ignores the seventy percent of trades that lost.

Subtract those. Now compute the same tenth as a share of what actually reaches your account.

Sit with that. Ninety percent of the trades you win are, collectively, a slightly losing business. You are not paid for winning. You are paid for the handful of trades where you were still holding when the move became absurd.

Three consequences, none of them optional

1. You cannot cap the winner

Capping at 1R turns +0.202R into −0.444R. Capping at 3R, which is the average winner, still produces −0.135R. Even a 5R cap barely clears zero.

This is the property that breaks intuition. In a fat tail the mean sits far to the right of the typical outcome, dragged there by rare events. A cap set at the mean therefore removes most of the profit and none of the losses.

The trader who cuts at 3R feels prudent. He has taken a winner three times the size of his risk. He has also, quietly, turned a good system into a bad one.

2. You cannot skip a trade

Here is the part nobody says out loud. The 148% is not spread across your good judgement. It arrives in a handful of trades, and you cannot identify them in advance. If you could, you would take only those, and the distribution would not be fat-tailed, and the edge would not exist.

Every skipped signal is a lottery ticket you declined to buy in a lottery where a tenth of the tickets pay for everything.

Which means the trade you talk yourself out of, on the day you are tired, in the week you are already down, is drawn from the same distribution as the one that would have made your year. There is no version of you that skips only the bad ones.

3. You must survive the flat periods, because that is where you live

A system whose profit is concentrated in a tenth of its winners spends most of its time going sideways or down. This is not a bug in trend following. It is what trend following is, viewed from inside the equity curve.

The flat period is not the strategy failing. It is the strategy waiting, and the waiting is the fee. The trader who abandons it during month seven has correctly identified that nothing is happening and incorrectly concluded that something is wrong.

Why this contradicts last week, and why both survive

The companion article demonstrated that a mean-reversion trader who holds past target loses money once more than 35.7% of his winners reverse. That number is real and it was computed from this same model.

The difference is the shape of the winner.

Neither man was wrong. Each was describing the only distribution he had ever traded, and neither told you that the advice was conditional, because to them it did not feel conditional. It felt like the truth.

Measure your own tail. Tonight.

Two computations, both trivial, both decisive.

Compute the concentration. Sort every winning trade you have taken by size in R. Sum the top decile. Divide by your net profit for the period. If the answer is above one, you are running a fat tail, and every instinct you have about “locking in gains” is costing you the year.

Compute the reversal rate. Of the winners that reached your target, how many would have retraced through your entry had you held? Above roughly thirty-six percent, you are thin-tailed, and the companion article is your instruction manual, not this one.

Most traders have never computed either. Most traders therefore take their exit doctrine from whichever book they read first, and roughly half of them are running the wrong one.

The uncomfortable symmetry

There is a version of this article that is popular and wrong, and it says: be patient, hold your winners, the big trade is coming.

That is not the claim. The claim is narrower and colder.

If your winner distribution is fat-tailed, then your profit is concentrated in outcomes you cannot foresee, cannot select, and cannot cap. It follows that your job is not to find the big trade. Your job is to be mechanically present, at full size, in every trade that meets your rules, for long enough that the tail has room to arrive. Nothing about that is patience. It is the removal of choice.

And if your distribution is not fat-tailed, none of the above applies, and holding out for a tail that does not exist is the most expensive form of hope in the market.

Find out which one you are. It is one evening’s work, and it is worth more than every entry technique you will study this year.