Holding Time and Size: The Disposition Effect in Numbers

[Fairman]

Holding Time and Size: The Disposition Effect in Numbers

A trader's records show winners held for an average of 35 minutes and losers for 95 minutes. The win rate is 55 percent. The average win is 0.8R, and the average loss is 1.2R.

Compute the expectancy. 0.55 times 0.8 is 0.44. 0.45 times 1.2 is 0.54. Expectancy: 0.44 minus 0.54, which is minus 0.10R. The strategy loses money despite winning more than half the time.

Diagnose. Losses average 1.2R, above the planned 1R, which means stops are being widened, removed, or hit with slippage. Wins average 0.8R, which suggests early exits, likely under the pressure of giving back profit. Together with the holding times, this is the classic pattern: cut winners short, let losers run.

Run what-if scenarios. If losses were held to 1.0R: 0.44 minus 0.45 is minus 0.01R, about breakeven. If wins also reached 1.2R, at the same win rate: 0.66 minus 0.45, which is plus 0.21R.

So fixing execution alone, with no change to entry quality, moves expectancy from minus 0.10R to plus 0.21R, a swing of 0.31R per trade. At 20 trades a month, that is over 6R.

Remedies: stops placed as orders and never widened, targets fixed in advance, partial exits to ease the urge to cut, and a rule against closing winners before the plan says so.

Exercise: compute the average hold time and average R of your winners and losers.

[FXS HongKong Scalper]

Very clear. One simple fix that helped people we work with: set the target and stop when placing the order and do not touch either unless a predefined rule says so. It removes most of the room for the disposition effect.

It is also worth comparing holding times for winners and losers every month. If losers are held more than twice as long as winners, that is a strong sign the stop is being moved.

[Fairman]

The holding time difference is a strong behavioural signal. Winners closed after 35 minutes and losers held for 95 minutes is the disposition effect in its classic form: taking profits quickly to lock in the good feeling, and holding losses in the hope they'll recover. The result is small wins and large losses, which produces negative expectancy even with a 55 percent win rate. The pattern is so common that it's been documented widely in studies of retail investors and traders. Recognising it in your own records is the first step, and the numbers make it hard to deny.

[Amanhasnoname]

When I first measured my holding times, my winners averaged around 22 minutes and my losers closer to 50, and I honestly did not believe it until I exported the history and sorted it myself. I had always remembered cutting losers on plan, but the records showed I moved the stop on about one in four of them.

What helped most was removing the decision. I now place the stop and take profit as real orders at entry, and my only allowed change is to move the stop to breakeven after a fixed level is reached. I also write the reason on any trade where I touch the order manually, and after two weeks seeing my own excuses in writing was enough to make me stop.

Another small change was a rule that a winner is closed only by its target, a stop, or a time limit, never because it feels uncomfortable. My average win went from about 0.7R to just over 1R within two months, with nothing changed in entries.

[Fairman]

The what if scenarios show how much execution matters. Holding losses to 1.0R brings expectancy to about breakeven. Letting wins reach 1.2R at the same win rate brings it to plus 0.21R. That's a swing of 0.31R per trade with no change to the entries, purely from following the plan. At 20 trades a month, that's over 6R, which at 1 percent risk is about 6 percent a month. The scenario assumes the win rate stays at 55 percent with larger targets, which may not hold, but even a modest improvement in exits would make a large difference.

[Fairman]

The remedy of placing stops as orders and never widening them addresses the loss side directly. Once the stop is on the server, the decision is made, and holding a loser longer isn't possible without actively moving it. Adding a rule that the stop can only move in the direction of the trade, never against it, removes the temptation. For the win side, fixed targets placed as orders work the same way. Partial exits can ease the urge to grab profits, by banking some gain early while leaving the rest to reach the target. That compromise often helps traders who struggle to hold full positions.

[Fairman]

Computing the average hold time and R for winners and losers, as the exercise suggests, is quick with a journal that records entry and exit times. If the pattern shows losers held much longer than winners, it's a strong signal that the disposition effect is present. Tracking it monthly shows whether new rules are working. Has anyone here measured their own hold times this way? I'd be interested in how large the gap was and which remedy made the biggest difference, whether fixed stops, fixed targets, or partial exits. Shared results would help others choose where to start.