Your stop is at $47. The stock is at $46.80 and the order has not filled yet. You are looking for a reason to move it, and the reason you find is that the whole market is down today, so this one is not about your setup. Ten minutes later the stop is at $44 and the trade is a different trade from the one you sized.
That substitution is the thing this article is about. The psychology behind cutting losses is not mysterious, it is measured, and the measurements say something useful about where to put the decision.
Why holding a losing trade feels reasonable
Losses count for more than gains of the same size
Kahneman and Tversky described the pattern in Prospect Theory (Econometrica, 1979). Their words: "losses loom larger than gains", and the value function they proposed is "generally steeper for losses than for gains". A later paper put a number on the steepness. Fitting the function to the choices of 25 graduate students, Tversky and Kahneman (Journal of Risk and Uncertainty, 1992) reported a median loss-aversion coefficient of 2.25, which they described as pronounced loss aversion.
That figure is the median for one group of subjects in one experiment, not a constant that applies to your account. What survives is the direction: the same amount weighs more as a loss than as a gain, so an exit that books the loss is not a neutral choice from the inside.
Investors act on it, and it is visible in the records
Terrance Odean tested what Shefrin and Statman had named the disposition effect, the tendency to hold losing investments too long and sell winning investments too soon. In Are Investors Reluctant to Realize Their Losses? (Journal of Finance, 1998) he analysed trading records from 1987 to 1993 for 10,000 accounts at a large discount brokerage.
Across the 1,893 accounts where both figures could be computed, the average proportion of gains realised was 0.57 and the average proportion of losses realised was 0.36. Odean also reports that the winning investments those investors sold went on to outperform the losers they kept over subsequent months.
The population is individual investors at one broker in one period, not swing traders as a group. What it establishes is that the preference shows up in transactions, not only in questionnaires.
The money already spent is not part of the decision
The sentence people say to themselves is "I have lost $500 on this, I cannot get out now." The $500 is gone under either choice. The only figure the decision turns on is what happens from here, which is the same question you would ask about a stock you have never owned.
One test converts it: with no position and this chart in front of you, would you buy at this price with a stop where yours is? A no answer and a decision to hold are two positions you cannot occupy at once.
Being wrong costs less than staying wrong
Exiting confirms the trade was wrong, and holding leaves the question open. The open question is the expensive one, because the position keeps trading while you are not deciding about it.
Waiting for it to come back
"It will come back" is a forecast, and it is the one forecast you make without doing any of the work you did for the entry. You did not check the chart, size the trade or set a stop against it. The price moves on what the market does next, which the forecast has no information about.
What holding a losing trade costs
The recovery arithmetic
A loss and the gain that reverses it are not the same percentage, because the gain is earned on the smaller balance that is left. For a loss of L, the gain needed is L divided by (1 minus L).
| Account loss | Gain needed to return to flat |
|---|---|
| 10% | 11.1% |
| 25% | 33.3% |
| 50% | 100.0% |
On $100,000, a 10% loss leaves $90,000, and $10,000 on $90,000 is 11.1%. A 50% loss leaves $50,000, and getting back to $100,000 means doubling it.
One planned loss against four unplanned ones
Take a hypothetical position sized so the planned stop costs $500. That is the 1R loss the trade was built around: one winner of the same size returns you to flat.
Move the stop three times and exit at $2,000 instead, and the trade cost 4R. Now four winners of the original size are needed to get back, on an account that has less to size them from. Nothing about the setup changed. The exit did.
Capital and attention
A held loser occupies money and screen time. Both are finite, and the setups that appear while your capital is tied up are the ones you do not take. That cost never appears on the statement, which is why it is the easiest one to ignore.
Techniques for cutting losses
Decide before you hold the position
Write the stop price before you place the order, and place the stop order at the same time as the entry. The decision made when you have no position is the one made without loss aversion in the room.
Ask the no-position question out loud
Say it as a sentence: "with no position, at this price, with this stop, do I buy?" Saying it converts a mood into a question that takes an answer.
Change what you call the exit
"I am losing $500" and "I am paying $500 to find out this setup did not work here" describe the same transaction. The second one is the version that leaves you able to take the next trade, and it is not a comforting fiction: the $500 bought information about a setup you will see again.
Take yourself out of the execution
If you override stops when you watch them, place the stop order and leave the screen. A resting order does not negotiate. This is the technique that works when the others have already failed.
Judge over the sample, not over the trade
Expectancy is an average across trades. A single loss carries no information about whether the method works, and the trade you are watching is one observation. Judge the method at your review, with the closed trades in front of you, rather than at the moment the stop fills.
Building the habit over three weeks
Week 1: short stops, small size. Take trades where the stop distance is small, so the loss when it fills is small. The point is repetition of the exit, not the result.
Week 2: normal stops. Go back to your usual distance and execute as written. Record how you felt at each fill.
Week 3: read it back. Count the stops you honoured against the stops you moved. Note what was happening in the market each time you moved one. The pattern in that column is your actual rule, whatever the document says.
When to exit a losing trade
Exit when the price reaches your stop. Exit when the reason you took the trade is no longer true, whatever the price is doing. Exit when the position is large enough to keep you awake, because a size you cannot sleep on is a size you will manage badly.
Four situations that put the exit on the table before the stop does:
- The position is bigger than you intended it to be
- You are hoping rather than reading the chart
- You are checking the price every few minutes
- You are explaining to yourself why this one is an exception
What cutting losses does not do
It does not give you fewer losing trades. Honouring stops produces more closed losses, not fewer, because the trades that would have drifted back to flat now close at the stop instead. What changes is the size of the largest ones, and the size of the largest losses is what decides whether a run of them ends your account.
Quick reference for a losing position
| Situation | The question underneath it | What to do |
|---|---|---|
| Price at your stop | Should I give it more room? | Let the order fill |
| Price below your stop | Will it come back? | Exit at the market |
| Reason for the trade is gone | But the stop has not been hit | Exit anyway |
| Position keeps you awake | Can I wait until tomorrow? | Reduce or exit |
| You are listing exceptions | Is this one different? | Exit |
Measure the disposition effect in Swingfolio
The behaviour the research describes has a card of its own. On the Behavioral tab of the analytics page, the "Disposition Effect" card compares how long you held winners against how long you held losers. It leads with one sentence in the form "Holding losers" then the ratio then "longer than winners", flipped when your winners are the ones you sit with. Under it sits a cost figure summed from the trades it flagged. Then two bars, Winners with their average holding time in days and Losers with theirs. Then a trend of Improved, Worsened or Stable against your earlier record. The card needs at least five winning and five losing trades with holding periods before it draws anything.
That is the number to watch while you work on this. If your losers bar is longer than your winners bar, you are holding the trades the research says are held, and you can see it in your own record instead of taking it on faith.
Close your next losing trade at the price you wrote down, then check the two bars after ten more. Start the 30-day trial.