Why Emotion Is the Hardest Part of Trading

Emotion is hard to manage in trading because the decisions that feel natural after a gain or a loss tend to be the ones that cost money: selling winners early, holding losers, trading too often and raising size to win losses back. Researchers cannot measure the feeling itself, but decades of account-level data show these behaviors are common, predictable and expensive. The math of losses then makes each mistake harder to undo than it was to make.

Key takeaways

Experiments put loss aversion at roughly 2: a 2024 meta-analysis of 607 estimates found a mean of 1.955, close to the 2.25 median in Tversky and Kahneman's 1992 study.
In 10,000 US brokerage accounts (1987–1993), a stock showing a gain was about 1.5 times as likely to be sold on a given day as one showing a loss, and the winners sold went on to beat the losers kept by 3.4 percentage points over the next year.
The 20% of households that traded most earned 11.4% a year from 1991 to 1996, against 17.9% for the market; among Brazilian futures day traders who kept at it for at least 300 days, 97% lost money.
A 50% drawdown needs a 100% gain to recover. In a 100,000-path simulation with even odds, doubling the stake after each loss left about 44% of accounts with less than 1% of their starting balance; a fixed 1% risk rule left none.

What the research can measure

No dataset records fear, regret or excitement directly. What researchers have are trading records: what was bought and sold, when, at what price, and how the account did afterward. They then test whether the patterns match predictions from psychology, such as loss aversion or overconfidence. When a pattern is widespread, persistent and costly, emotion is a plausible explanation, but other explanations (taxes, rebalancing, cash needs, plain lack of information) have to be ruled out first, and the better studies try to do that.

So the claim that emotion is the hardest part of trading holds in a specific sense: traders with the same tools and information keep behaving in ways that lower their own results, and that behavior is measurable. Whether the cause is "emotion" in each case is an inference, and for day traders in particular, costs and the lack of an edge also explain much of the losses.

Why losing positions get held

Loss aversion in numbers

Kahneman and Tversky's prospect theory paper (Econometrica, 1979) described how people judge outcomes as gains or losses from a reference point, and how losses loom larger than gains of the same size. They also found people tend to be risk-averse with gains and risk-seeking with losses: they prefer locking in a sure gain but will gamble to avoid a sure loss. Their follow-up, cumulative prospect theory (Journal of Risk and Uncertainty, 1992), estimated a loss-aversion coefficient, usually written λ (lambda), with a median of 2.25 in its experiments.

Later work broadly supports a figure near 2. A meta-analysis by Brown, Imai, Vieider and Camerer (Journal of Economic Literature, 2024) pooled 607 estimates from 150 articles published from 1992 to 2017 and found a mean of 1.955, with a 95% probability interval of 1.820 to 2.102. Individual estimates vary widely by person and setting.

Illustrative example: with λ = 2 and a simple linear value scale, a coin flip that risks $100 only feels acceptable if the win is above $200. At λ = 2.25 the threshold is $225. A trade with a 50% chance of +$150 and 50% chance of −$100 has an expected value of +$25, yet under this model it still feels like a bad bet.

The disposition effect

Shefrin and Statman named the disposition to sell winners too early and ride losers too long (Journal of Finance, 1985), linking it to loss aversion, mental accounting, regret and self-control. Selling a loser turns a paper loss into a realized one, which is the step people avoid.

Terrance Odean tested it in "Are Investors Reluctant to Realize Their Losses?" (Journal of Finance, 1998) using 10,000 accounts at a large US discount broker from 1987 to 1993. He compared two ratios:

PGR = realized gains / (realized gains + paper gains)
PLR = realized losses / (realized losses + paper losses)

PGR (proportion of gains realized) counts how many of the gains available on days a trader sold anything were actually taken; PLR does the same for losses. Over the full year, PGR was 0.148 and PLR 0.098, a ratio of about 1.5. Rebalancing and trading costs did not explain it, and later returns did not justify it: over the following year, the winners sold beat the losers held by 3.4 percentage points. In taxable accounts it also lowered after-tax returns.

Overconfidence and trading costs

Overconfidence, in finance research, means overestimating the precision of one's own information or skill. Models predict that overconfident traders trade too much.

Barber and Odean's "Trading Is Hazardous to Your Wealth" (Journal of Finance, 2000) studied 66,465 households at a large discount broker from 1991 to 1996. The fifth of households that traded most (monthly turnover above 8.8%) earned 11.4% a year net of costs, while the market returned 17.9%. Before costs, their returns were little different from those of households that traded rarely; the trading itself was the drag. The average household earned 16.4% and turned over 75% of its portfolio each year. Their follow-up, "Boys Will Be Boys" (Quarterly Journal of Economics, 2001), used gender as a proxy for overconfidence and found men traded 45% more than women.

What day-trading records show

Two large datasets look at people who trade intraday.

Study	Who and when	Main finding
Chague, De-Losso and Giovannetti, "Day Trading for a Living?" (FGV working paper, 2020)	Everyone who started day trading in the Brazilian equity futures market in 2013–2015 and kept going for at least 300 days	97% lost money; 0.4% earned more than a bank teller; no evidence of learning with experience
Barber, Lee, Liu and Odean, "The Cross-Section of Speculator Skill" (Journal of Financial Markets, 2014)	Day traders in Taiwan, 1992–2006	Less than 1% could predictably and reliably earn positive abnormal returns net of fees

These results describe outcomes, not motives; costs, competition with professional and high-frequency traders, and lack of an edge all play a part. FINRA's investor page on frequent intraday trading makes the same practical point: day traders can lose some or all of what they put in, and more than that when trading on margin.

The fund-flow timing gap

Morningstar's Mind the Gap 2026 study (August 2026) estimates that the average dollar in US mutual funds and ETFs earned 8.7% a year over the 10 years ended December 31, 2025, while the funds returned 9.9%. The 1.2-point gap comes from the timing and size of purchases and sales. It was about 0.4 points a year for the least volatile fifth of funds and more than 2.0 points for the most volatile fifth.

Morningstar itself warns against reading the gap as proof of poor investor behavior: investing part of every paycheck or rebalancing on schedule can also open one. The gap fits buying after rises and selling after falls, but it measures flows, not feelings.

The math that emotion works against

Three formulas explain why emotional decisions compound. Each variable is defined below the block.

Gain needed to recover  = d / (1 − d)
Break-even win rate     = 1 / (1 + R)
Position size (shares)  = (A × r) / (entry price − stop price)

Here d is the drawdown as a decimal (0.25 for a 25% fall), R is the average win divided by the average loss, A is account value and r is the fraction of the account risked on one trade (0.01 for 1%). The break-even formula ignores commissions, spreads and slippage, which all push the required win rate higher. The position-size formula is written for a long position; for a short sale, use stop price minus entry price. It also assumes the exit fills at the stop, and a price gap through the stop makes the actual loss larger than A × r.

Drawdown	Gain needed to get back
10%	11.1%
20%	25.0%
30%	42.9%
50%	100.0%
75%	300.0%

Reward-to-risk (R)	Break-even win rate
0.5	66.7%
1	50.0%
1.5	40.0%
2	33.3%
3	25.0%

Cutting winners short and letting losers run lowers R, so a trader whose average win is half the average loss needs to be right two times out of three. A lower R is not a loss on its own, though, because the same habit also raises the win rate. In a market with no trend, taking profit $5 above entry and cutting losses $15 below it gives R = 1/3 and wins about 75% of the time, which is exactly the break-even rate. The disposition effect costs money when, as in Odean's data, the winners sold keep outperforming the losers held, and when realizing gains early brings the tax on them forward, often at higher short-term rates.

Doubling up after losses: a simulation

Raising size to win back a loss quickly is often called revenge trading; no study cited here measures how common it is, but its effect on risk can be tested. Illustrative example, with these assumptions: a $10,000 account, 200 trades, each with a 50% chance of winning, a win pays exactly the amount risked, no costs, 100,000 simulated paths (Python's random module, seed 42).

Fixed rule: risk 1% of the current balance on every trade.
Doubling rule: risk $100, double the stake after every loss, return to $100 after a win; if the stake exceeds the balance, risk everything left.

Result after 200 trades	Fixed 1% risk	Double after each loss
Median ending balance	$9,900	$11,200
Ended above $10,000	47.2%	51.6%
Ended below $5,000	0%	44.6%
Ended below $100	0%	43.9%
5th / 95th percentile	$7,788 / $12,586	$0 / $20,900

Both rules have the same expected value, $10,000, because the odds are even; doubling does not create an edge. What it changes is the shape of the outcome. Small, frequent wins make the doubling rule look better at the median, while a long losing streak wipes out the account. Seven losses in a row cost $12,700 under the doubling rule, more than the $10,000 starting balance, and the chance of at least one such streak in 200 even-odds trades is about 54%. Not every streak ends the account, because some come after the balance has grown, but about 44% of simulated accounts finished with less than $100. The same seven-loss streak under the fixed 1% rule costs 6.8%, leaving $9,321.

Apps, feeds and the urge to trade

Regulators have looked at how product design and social media may feed these patterns. In August 2021 the SEC published a request for information on digital engagement practices (Release No. 34-92766). It listed examples such as streaks and contests, points and leaderboards, notifications, animations celebrating trades, and screens that change color with portfolio performance, and it asked whether these features affect trading frequency, volume and results. It was a request for comment, not a rule.

FINRA's investor insight "Following the Crowd: Investing and Social Media" (updated April 2026) warns that real-time discussion platforms and buy or sell indicators driven by social sentiment can lead to emotionally driven or impulsive decisions. Neither source quantifies the effect on returns; both treat attention prompts as a risk factor worth recognizing.

Frequently asked questions

Is everyone equally loss-averse?

No. The 2024 meta-analysis reports an average near 2, but the 607 estimates it pooled vary by person, task and study design. A coefficient of 2 is a population average from mostly experimental settings, not a fixed trait that predicts any one trader's behavior.

Does the disposition effect change in December?

In Odean's 1998 data it reversed. In December, PLR rose to 0.128 and PGR fell to 0.108, so losses were realized more often than gains, which fits selling losers to lower that year's tax bill. From January to November, PGR was 0.152 against a PLR of 0.094.

Do men and women trade differently?

In Barber and Odean's 2001 study of more than 35,000 households from 1991 to 1997, men traded 45% more than women. Trading cut men's net returns by 2.65 percentage points a year, compared with 1.72 points for women. The authors used gender as a stand-in for overconfidence, not as a cause in itself.

Is the $25,000 pattern day trader minimum still in effect?

FINRA replaced its day-trading margin requirements, including the pattern day trader designation and the $25,000 minimum equity, with intraday margin standards for margin accounts, effective June 4, 2026, under Regulatory Notice 26-10. Firms that need more time may phase the change in until October 20, 2027, so some brokers may still apply the old rules. Regular maintenance margin requirements did not change.

Does a stop-loss order guarantee the exit price?

No. The SEC's investor bulletin on stop, stop-limit and trailing stop orders explains that a stop order becomes a market order when triggered, so the fill can be far from the stop price, and a short intraday swing can trigger it. A stop-limit order controls the price but may not execute at all if the market moves past the limit.

Rules to write before a trade

None of the studies above tests a routine that removes these biases. The patterns above all arise once a gain or loss is on the screen, so decisions made before a position is open are a reasonable place to start. Points to set in writing:

Risk per trade. A fixed fraction of the account, such as 1%, and the share count from the position-size formula above. The Position Size & Risk calculator on this site computes size from risk per trade, the reward-to-risk ratio, the break-even win rate and drawdown recovery.
Exit conditions on both sides. The price or event that ends the trade at a loss, the one that takes profit, and the reward-to-risk ratio they imply.
Order types. A stop or stop-limit order can carry the exit plan while the trade is not watched, within the limits described in the FAQ above.
A rule for after losses. For example, no size increase after a losing trade, and a set pause or a daily loss limit after which trading stops for the day.
A journal. Entry reason, planned exit, actual exit and whether the plan was followed. Comparing realized winners with realized losers over time shows whether the disposition pattern is present.
Fewer triggers. Turning off price alerts and social feeds that are not part of the written plan reduces the prompts the SEC and FINRA flagged.