Most traders know the feeling. A position shows a profit and you want to lock it in before it disappears. A position shows a loss and you give it a few more days to come back. Closing winners more readily than losers is called the disposition effect, a name Hersh Shefrin and Meir Statman gave it in 1985.
A book can be right half the time and still lose money. Below are the findings from one large study, a worked example of ten trades, and a way to check how often your own trades followed the rules you set.
What one study of 10,000 accounts found
Terrance Odean studied 10,000 randomly selected accounts at a nationwide discount broker, covering trades from January 1987 to December 1993 ("Are Investors Reluctant to Realize Their Losses?", Journal of Finance, 1998).
On each day an investor sold something, he counted every position in the account as a realised or an unrealised gain or loss. He then compared how often gains were realised with how often losses were realised, pooled across all accounts and dates.
| Measure, all months | Result |
|---|---|
| Gains realised, as a share of all gain positions (count, not dollars) | 0.148 |
| Losses realised, as a share of all loss positions (count, not dollars) | 0.098 |
| Ratio | a little over 1.5 |
In his words, a stock that is up in value was "more than 50 percent more likely to be sold from day to day than a stock that is down."
The study also tracked what happened next. Over the following year, the winners investors sold went on to beat the losers they kept by 3.4 percentage points, with both groups measured against the market.
Three limits apply to these figures:
- One broker, one period. The sample is one broker's clients from 1987 to 1993, not every investor.
- December is different. In December the pattern reversed, which is consistent with investors closing losing positions for tax reasons. The all-months figure includes December.
- The 3.4 points is a gap between two groups. It is not a return anyone earned, and it is not a forecast.
Worked example: right half the time, still down
This example is hypothetical.
| Trades | Average hold | Result each | Total | |
|---|---|---|---|---|
| Winners | 5 | about 5 days | +$400 | +$2,000 |
| Losers | 5 | about 15 days | -$1,000 | -$5,000 |
| All ten | 10 | -$3,000 |
Half the trades made money. The book is still down $3,000, because the losses were bigger than the gains. The losers also stayed open three times as long as the winners.
The holding times do not prove why the losses were bigger, and the example has no planned exits to compare against, so it cannot measure what different exits would have produced. It shows how a book can be right half the time and still lose money.
Where your exit rule comes in
Breaking a plan is not part of the disposition effect's definition, and a profitable trade can break a rule while a losing trade follows one. But if your plan has an exit rule, such as a stop you set before entering the trade, your journal can record whether you followed it. That gives you a second question to ask alongside the result: how many of your trades followed the rules you set, and how did those trades do against the rest?
Checking rule adherence in Swingfolio
Swingfolio's strategy page has a Rule adherence tab. For each strategy, it takes the closed trades in the selected period, portfolios and markets, and lets you choose which checks count: Entry and exit, Entry only or Exit only. It then works out the share of recorded checks marked as followed, rounded to a whole percentage, and groups the trades:
- High adherence: a rounded score of 80% or more
- Low adherence: a rounded score under 80%
- Unrecorded: trades with no checks recorded for the selected view
For each group it shows the average R, net P&L, average net trade and win rate, where the inputs exist. A High minus low row compares average R, average net trade and win rate when both groups have trades; otherwise it says there are not enough trades to compare. If no checks are recorded, the tab points you to the trades that need them.
The tab states its own limit: the groups show observed differences between sets of trades, not proof that following the rules caused the difference. It is a way to see the pattern in your own records.
Open a strategy in Swingfolio, choose the Rule adherence tab, and compare the two groups before your next trade.
Frequently asked questions
Is taking profits early always a mistake?
No. A planned exit at a profit target is a rule being followed. The disposition effect describes a tendency across many trades to close winners more readily than losers, whatever the plan said.
Why do losing trades get held longer?
One possible explanation is that selling a loser makes the loss final, so some investors put off that moment. The study measures the pattern; it does not measure each investor's reasons.
Does the study apply to swing trading today?
It studied individual investors at one broker between 1987 and 1993, so it cannot say how common the pattern is among swing traders today. Your own trades are the most direct evidence for you.
Source
Terrance Odean, "Are Investors Reluctant to Realize Their Losses?", The Journal of Finance 53(5), 1998, pp. 1775 to 1798: author's PDF.
The trades in the example are hypothetical. General information only. Not financial advice.