They could sell a loser. Once a year.
You know which position it is. The one your eye skips past when you open the platform. Not quite bad enough to do anything about, and you've changed your reason for keeping it twice now.
Somebody already measured how common that is, in real accounts, with real money.
10,000 accounts, seven years, one very strange month
In 1998 Terrance Odean published a study in The Journal of Finance using the trading records of 10,000 accounts at a large discount brokerage, every trade from 1987 through 1993. Real people, real money, no idea they were being watched.
He counted what proportion of winning positions those people actually sold, and what proportion of losing positions they actually sold.
Same people, same accounts. Winners went out half again as often as losers.
Across the whole period they sold 14.8% of their winners and 9.8% of their losers.
That's the finding everyone quotes. Here's the one almost nobody does.
In December it reverses. For that single month the same accounts sold 12.8% of their losers and only 10.8% of their winners. Eleven months of the year they can barely close a losing position. Then a tax deadline appears on the calendar and they can do it without blinking.
It was never that they couldn't sell a loser. They needed a reason that came from somewhere other than the trade itself.
The paper is free to read and every number above is in it: Are Investors Reluctant to Realize Their Losses?
You've already done this, and not in December
You've closed a losing position at a month end "to start clean". Before a holiday, so you weren't watching it from an airport. Because the account touched a round number and it felt like a tidy place to stop.
Every one of those is a December. An outside reason that let you do the thing you couldn't do on a normal Tuesday on the trade's own merits.
The reason had to come from outside because you never wrote one on the inside. No exit condition means closing the position is a fresh judgement call every day, and a judgement call that requires admitting you were wrong is one nobody volunteers for. So it waits for a date.
"But the loser might come back"
Everybody holding a losing position believes some version of this, so Odean tested it.
He took the winners those accounts sold and the losers they kept, and measured what each group did afterwards against the market. Over the following year the winners they sold beat the losers they kept by 3.4 percentage points.
Not patience then. A preference for one feeling over another, with a price on it.
The bit that cost me something this week
I did the winner half on Wednesday, in public, three days before writing this.
I was short a currency pair with a rate decision coming that afternoon. It was in profit and the exit price had been published in advance where anyone could see it. Held to that published exit the trade was worth just over two units of risk.
I closed it before the announcement for about one and three quarters.
Both losers went the full distance. The winner didn't.
A third of a unit of risk, handed back, for a reason that existed nowhere in the plan. Ten of those a year is three units, which on my book is a month. And it doesn't feel like anything at the time, because I booked a profit and profits don't feel like errors.
Here's the part that makes it useful rather than just embarrassing. Both of my losing positions that same week went to the exact price I'd said I'd get out at. Every point of both. I didn't cut either one early and I didn't widen either one.
That isn't willpower. The exit on a loser was a number on a screen before I had money on it, so closing it was never a decision I had to make while it was happening. The exit on the winner was a number on a screen too. I just overrode that one, because nothing stops you overriding a target and everything stops you overriding a stop.
Odean's investors got both halves wrong. I got one half wrong, and the half I got right was the half I'd written down first.
Four things that actually help
1. Write the exit before the entry. Not a target, an invalidation: the price at which the idea is wrong. Written down before there's money on it, closing the trade stops being an admission and becomes the plan finishing.
2. Give every position a maximum age. "It hasn't gone against me" isn't the same as "it's working".
3. Log the exit you planned next to the exit you took. Two columns, every trade. The gap between the totals is what your exits cost you, and it's usually bigger than anything a better entry would have won you.
4. Review open and closed positions in the same sitting. The whole mechanism works by keeping your mistakes off the closed list, so looking at both at once is the cheapest defence there is.
None of this needs more discipline than you've already got. Those investors had plenty of discipline every December. What they didn't have was a reason written down in advance.
Where the rest of this lives
I publish the setups before they trigger, with the entry, the stop and the target written down in advance, and every trade that closes goes on a public record either way. Winners and losers, updated every Sunday, open to anybody: tradedesk.rbtrading.site/track-record
The full letter goes out three times a week here: Inside the Trade
And if point 3 above landed, the journal is the thing that does it for you. It carries the planned R and the realised R on every position, so the gap shows up on its own instead of hiding inside a column of green numbers. 20% off at rbtrading.site/?ref=RBT20.
Reply and tell me one thing: the last winner you closed early, was there anything in your written rules that told you to, or did it just feel like the right moment ?
I read every one.