Trading Biases That Cost Money: Loss Aversion, the Disposition Effect and 6 More

Most traders don’t lose money because they don’t know what to do. They lose because they can’t do what they know while money is moving. The brain has decision shortcuts that work well in daily life, but in markets they steer us wrong in the same way again and again. Researchers in behavioural finance have studied these habits for decades. This article covers the 8 most common ones, each with a trading example and a fix you can start using right away.

What a bias is, and why knowing it isn’t enough

A cognitive bias is a decision error that repeats in the same direction, not a random one. You don’t close winners sometimes early and sometimes late in equal measure. You close them early almost every time.

The bad news is that knowing a bias’s name doesn’t make it go away. The good news is that you don’t have to beat it while the chart is moving. You only have to decide in advance, while you’re calm, and let the plan do the work. That’s why almost every fix in this article is the same idea: move the decision to before you open the trade.

1. Loss aversion: a loss hurts more than an equal gain

In 1979 Daniel Kahneman and Amos Tversky proposed prospect theory. It says people don’t judge outcomes by their total wealth, but by gains and losses against a reference point, such as their entry price. It has three key features.

• Losses weigh more than gains. Losing 100 feels worse than winning 100 feels good. This is loss aversion
• The gain side curves down. The first bit of profit feels great, and each extra bit adds less. So people want to lock in the profit they already have
• The loss side curves too. Going from −50 to −100 hurts less than going from 0 to −50. So people will gamble on holding a losing trade, hoping it comes back to break-even

In trading this shows up as moving the Stop loss further away as price gets close, or deleting it because “it will come back”. Meanwhile trades in profit get closed early for fear of losing the gain.

The fix is to put the SL and TP in the order when you open the trade, not in your head, and to risk little enough per trade that a full SL loss doesn’t hurt so much you have to run from it, for example around 1% of the account. See Risk 1% per trade.

2. The disposition effect: selling winners early, holding losers long

Hersh Shefrin and Meir Statman named this the disposition effect in 1985: the tendency to sell what’s in profit too early and hold what’s at a loss too long. They explained it partly with the prospect theory above (locking in gains, gambling on losses).

In 1998 Terrance Odean studied the trading records of a large number of retail stock accounts. Investors sold stocks that were up clearly more readily than stocks that were down. Worse, the winners they sold went on to do better afterwards than the losers they kept. So the habit didn’t even help with picking which stock to keep.

In trading you open a trade planning a TP at +2R (R is one unit of risk, the distance to your SL). As soon as it’s up half an R, you close it out of fear. The losing trade gets its SL pushed out until it loses twice what you planned.

Worked example: what cutting winners and holding losers does to expectancy

Expectancy is the average result per trade, in R.

Expectancy = win rate × average win − loss rate × average loss

Assume the system wins 40% in every case. Only the exit changes.

1. As planned, win +2R, lose −1R: 0.4 × 2 − 0.6 × 1 = 0.8 − 0.6 = +0.2R per trade
2. Cut winners only, win +0.5R, lose −1R: 0.4 × 0.5 − 0.6 × 1 = 0.2 − 0.6 = −0.4R
3. Hold losers only, win +2R, lose −2R: 0.8 − 0.6 × 2 = 0.8 − 1.2 = −0.4R
4. Both, win +0.5R, lose −2R: 0.2 − 1.2 = −1.0R per trade

Exit	Expectancy per trade	100 trades	Break-even win rate
As planned +2R / −1R	+0.2R	+20R	33.3%
Cut winners at +0.5R	−0.4R	−40R	66.7%
Hold losers to −2R	−0.4R	−40R	50.0%
Both	−1.0R	−100R	80.0%

Break-even win rate = average loss ÷ (average win + average loss). For the last row, 2 ÷ (0.5 + 2) = 80%. If 1R = 10 dollars, the plan makes +200 dollars over 100 trades, while doing both loses 1,000 dollars, from the very same entries.

Check: the plan over 100 trades wins 40 × 2 = 80R and loses 60 × 1 = 60R, net +20R, matching 0.2 × 100. Doing both wins 40 × 0.5 = 20R and loses 60 × 2 = 120R, net −100R.

In real life the win rate won’t stay exactly the same. Closing winners early makes you win a bit more often, and some of the losers you hold do come back. But look at the break-even column: doing both needs a win rate above 80%. A system that used to win 40% almost never gets there, and a single loser that never comes back can run far beyond −2R. For more on break-even win rates, read A high win rate doesn’t mean profit.

The fix is to let the TP and SL orders you set at the start close the trade. If you want to trail the SL or close part of the position, write the rule in advance and apply it to every trade. Note in your journal which trades you closed before the plan, and each week work out how many R you would have made by following it.

In this example the entries are the same in every row. What turns +0.2R into −1.0R per trade is purely how the trades are closed.

3. Overconfidence and overtrading

Overconfidence is rating your own knowledge and skill higher than it really is. Brad Barber and Terrance Odean studied tens of thousands of retail accounts. In their 2000 study, the households that traded most earned clearly lower net returns than those that traded least, mainly because of the costs paid on every trade. In their 2001 study, men traded more often than women and earned less after costs, which fits psychology research finding that men tend to be more overconfident about finance.

In trading, after a run of wins you feel you can read the market, so you raise your lot size, take trades outside your setup and trade every move. Every trade pays the spread, so a trade with no edge doesn’t come out at zero, it comes out negative by the cost. See Spread cost per trade.

The fix is to set a maximum number of trades per day and a fixed risk size in advance, without adjusting them by mood after wins, and to let your journal, not your feelings, decide whether your skill has really improved.

4. Gambler’s fallacy and the hot hand

The gambler’s fallacy is believing that after several losses the next trade is “due” to win. The hot hand is the flip side: believing you’re on a roll and wins will keep coming. A classic 1985 study by Gilovich, Vallone and Tversky looked at this in basketball (later research argued that some hot hand may exist in sport). But when you trade the same rules, the last trade’s result says almost nothing about the next one.

In trading, you lose 5 in a row and double the lot because “this one has to win”, or win 5 in a row and raise risk because you’re “hot”. Either way your biggest risk lands at a moment with no reason behind it.

The fix is to work out in advance how long a losing streak your system normally has. See Losing streaks are normal. Keep risk per trade fixed and set a daily loss limit. See Revenge trading and daily loss limits.

5. Recency bias: giving recent events too much weight

Recency bias is judging by what just happened rather than the whole picture. Things you remember clearly seem to happen more often than they do.

In trading, you drop a system tested over hundreds of trades after its last 4 losses, or you see the market run one way for three days and assume it will keep doing so. It also includes picking a signal or a trader to copy based on last month’s results. See Evaluating copy trading and signals.

The fix is to judge a system on a large number of trades in your journal, not the last few, and decide in advance when you’ll review it: after a set number of trades, or when a losing streak goes beyond the range you calculated. See Trading journal template. If you suspect the market really has changed character, read Market regime change.

6. Confirmation bias: looking only for information that agrees with you

Confirmation bias is seeking out, favouring and remembering information that fits what you already believe, while overlooking what contradicts it.

In trading, you’re holding a Buy, so you only read analysis saying price will rise, bad news gets dismissed as “already priced in”, or you flip through timeframes until you find a chart that still looks bullish. It works hand in hand with looking back at charts and seeing only the times a line was spot on. See Hindsight bias and lines that always look right.

The fix: before opening a trade, write one sentence on “what would prove me wrong”, and put the SL there. Use a checklist that must be fully ticked before you enter, and count the results of every signal, not just the ones that worked.

7. Anchoring: sticking to the first number you saw

Tversky and Kahneman described anchoring in 1974: when people estimate a value, they start from the first number they see and don’t adjust far enough from it, even when that number is irrelevant.

In trading, the most common anchor is your entry price: “I’ll wait for it to get back to break-even, then close”, even though the market has no idea where you got in. Another is the previous high, “it used to be up there, so it’s cheap now”, or a daily profit target you’ve fixed in your head.

The fix is to ask yourself: “If I didn’t have this trade, would I open it now at this price?” If the answer is no, the only reason you’re still holding is usually your entry price.

8. Sunk cost: not stopping because you’ve already lost

The sunk cost fallacy is carrying on because of what you’ve already put in, not because what happens next is worth it. Hal Arkes and Catherine Blumer showed this in experiments in 1985: people stuck with what they had already paid for, even when they knew another option was better.

In trading, you add to a heavily losing trade to average the price, because you don’t want what you’ve lost to go to waste. See Grid, Martingale and averaging down. Or you keep running a system you spent time and money on even though its real results don’t pass your criteria.

The fix is to write stop conditions before you start, for example you’ll stop using a system once drawdown exceeds a set level, or if results after a set number of trades fall below your threshold. The money and time already spent don’t come back whichever way you decide. Only the results from here on are still yours to choose.

Every bias in this article is hard to beat while the chart is moving, but easy to prevent if you decide before opening the trade.

All the fixes in one place

Bias	How it shows up	Fix
Loss aversion	Moving the SL away, refusing to lose	SL in the order from the start, small risk per trade
Disposition effect	Closing winners early, holding losers	Pre-set TP/SL, log trades closed before plan
Overconfidence	Overtrading, bigger lots after wins	Daily trade cap, fixed risk
Gambler’s fallacy / hot hand	Doubling after losses or after wins	Fixed risk, daily loss limit
Recency bias	Dropping a system after a few losses	Judge on many trades in the journal
Confirmation bias	Reading only what agrees with you	Write what proves you wrong, use a checklist
Anchoring	Waiting to get back to break-even	Ask if you’d open this trade now
Sunk cost	Averaging down, keeping a failed system	Stop conditions written before you start

The core tools are just three: SL/TP set in advance, a pre-entry checklist, and a trading journal that records both results and reasons. To check what win rate your planned RR needs, use the RR calculator at the end of this article.

Common mistakes

• Thinking that knowing means you won’t do it. Biases act while money moves, not while you read. You need rules that act for you
• Setting an SL, then cancelling it as price gets close. An SL that moves with your mood is the same as no SL
• Being proud of a high win rate that comes from cutting winners early. Look at expectancy and the break-even win rate too, not the win rate alone
• Logging only winners, or only numbers. A journal without the reasons for entry and exit can’t catch a bias
• Switching systems after every losing streak. You never get enough trades to know which system actually works