The failure rate you have seen quoted for retail traders, whether 70%, 90% or 95%, comes with no citation attached. The numbers that do have citations are narrower, and more useful, because they name the population and the period they measured.
Two of them are worth your time. Neither is a prophecy about you. Both point at habits you can check in your own record this week.
What the research measures
Trading more went with earning less
Brad Barber and Terrance Odean analysed 66,465 households with accounts at a large discount broker from 1991 to 1996. They published the result as "Trading Is Hazardous to Your Wealth" (Journal of Finance, 2000). The households that traded most earned an annual return of 11.4%, against 17.9% for the market over the same period. The average household earned 16.4% and turned over more than 75% of its stock portfolio a year.
The authors attribute the gap to overconfidence rather than to bad stock picking. Gross returns across the turnover quintiles were close together; the difference showed up after trading costs.
Persistent skill exists, in a small group
Barber, Yi-Tsung Lee, Yu-Jane Liu and Odean looked at every day trader in Taiwan from 1992 to 2006. They ranked the traders by one year's returns, then measured how the same traders did the next year ("The cross-section of speculator skill", Journal of Financial Markets, 2014). The 500 top-ranked traders went on to earn 61.3 basis points a day before fees and 37.9 after. The bottom-ranked group went on to lose 11.5 basis points a day before fees and 28.9 after.
Their conclusion, in their words: "Less than 1% of the day trader population is able to predictably and reliably earn positive abnormal returns net of fees."
That is day traders on one exchange over one 15-year window, not swing traders, and not you. What it does establish is that last year's ranking carried information about next year's, in both directions. Whatever separates the two groups is repeatable.
Why traders fail: ten mechanisms
The reasons below are failure modes, not proportions. Each one names something that happens and something you can check.
1. Trading before learning the instrument
Opening an account takes minutes and learning the instrument does not. The gap between those two durations is filled with real money.
What to do instead. Set a length of study you will finish before you fund the account. Write down what finished means: a strategy you can state in rules, a position size you can calculate, and a stop rule you can apply. Paper trade until each of those produces the same answer twice.
2. No written plan
A plan you carry in your head changes shape while you use it. There is no way to tell afterwards whether you followed it, because it has already adapted to what happened.
What to do instead. Write the markets, the setups, the entry criteria, the exit criteria, the sizing rule and the risk limits into one document. Anything you cannot answer with a tick or a cross is not written down yet.
3. Sizing by conviction
When the size of the position moves with how much you like the trade, the one you like most is the one that costs most when it fails. That is the sequence that turns a run of ordinary losses into a hole.
What to do instead. Fix the risk per trade as a share of the account, then derive the share count from the stop distance. Risking 1% of $50,000 is $500; with a $4 stop distance, that is 125 shares. The conviction shows up in whether you take the trade, not in how big it is.
4. Trading the feeling instead of the setup
Revenge after a loss, size-ups after a win, and a stop moved because the market is down today all share a shape: the decision was made after the position existed.
What to do instead. Record the emotion with the trade, then read the emotions back against the results at your review. A tag with a losing total attached to it is a rule you can write for yourself.
5. Expecting a number nobody promised
Return figures circulate without a source, a period or a risk level attached. Any of them can be beaten in a quarter and missed over a decade.
What to do instead. Set the expectation from your own closed trades, not from a figure you read. Once you have enough of them, expectancy per trade multiplied by trades per year is your estimate, with your costs already inside it.
6. Overtrading
Barber and Odean measured this one directly. The most active households in their sample earned 11.4% a year against a market return of 17.9%, and their gross returns were close to everyone else's. Costs account for the rest.
What to do instead. Count your trades per month alongside your win rate and average R-multiple. If a month with twice the trades produced the same result, the extra trades paid the spread and nothing else.
7. No patience for the setup
Chasing an entry, exiting a winner early and forcing a trade on a quiet day are the same impatience pointed at different parts of the trade.
What to do instead. Use price alerts rather than screen time, and log the days you did not trade with the reason. A month of blank days with reasons is a record; a month of blank days with nothing written is a gap.
8. No records
Without records, this month's mistake and last month's mistake are separate events. With them, they are one pattern with a name and a cost.
What to do instead. Log every trade, including the ones you would rather forget, and calculate win rate, average R-multiple, expectancy and profit factor on a fixed schedule.
9. Too little capital for the rules
At a fixed risk percentage, the dollar risk shrinks with the account, and the fixed costs do not. Risk 1% of $2,000 and you have $20 at risk; a $10 round trip is half of it. Risk 1% of $50,000 and the same $10 round trip is 2% of the $500 at risk.
What to do instead. Work out what your broker charges per round trip, then find the account size at which that charge stops dominating your risk budget. That number is yours, and it depends on your market and your broker rather than on a figure in an article.
10. No one to answer to
Trading alone means the only account of what happened is the one you give yourself, written after you know the result.
What to do instead. Send someone your adherence rate and your closed trades on a set day. A number is harder to soften than a story.
How to be different: five practices
Learn before you fund
Complete the study you defined in your plan, then paper trade the strategy until it produces repeatable answers. The threshold is yours to set, and it belongs in writing before you start.
Size every trade the same way
Risk a fixed share of the account, place the stop when you place the entry, and derive the share count from the distance between them. Every trade, including the good one.
Grade the execution, not the result
At the close of each day, score whether you followed your rules on each trade. A winning trade that broke a rule is a failed trade for this purpose, because the rule is what you are testing.
Wait
Cash is a position. The setups you skip cost nothing, and in the Barber and Odean sample the households that skipped more of them kept more of the market return.
Report to someone
A weekly review you send to another person, a monthly statistics pass, and an honest note about what you changed and why.
A 90-day plan to work through
The plan below is an example. Replace the durations and the thresholds with numbers you set.
Days 1 to 30. Complete the study, write the plan, set up the journal, and paper trade the strategy end to end.
Days 31 to 60. Trade small real size. Score rule adherence on every trade and log all of them. The result you are watching is the adherence rate.
Days 61 to 90. Read the record back. Which rule broke most often, and what was happening when it did? Which setup carried the results? Adjust the plan between sessions, never during one, and increase size only if the record supports it.
Track the ten in Swingfolio
Swingfolio records the parts of this that are checkable.
The plan and the trade in one place. A strategy holds your entry and exit rules. When you attach it to a trade, the entry rules appear as a checklist on the trade form and the exit rules on the open position. Each list carries a bar labelled "Adherence" showing the percentage you ticked.
The execution grade. On the Behavioral tab, the "Behavioral Finance Score" card carries a "Rule Compliance" bar. It is scored from the percentage of your last 30 closed trades where every entry rule was ticked. Further down the tab, "Profit Factor by Compliance" puts the profit factor of those trades against the profit factor of the rest. A "Discipline Score" widget you can add to the dashboard prints the same percentage with the prior 30 beside it.
The trading frequency. The Behavioral tab carries a card titled "Trade Pacing", subtitled "Overtrading Detection". It plots your win rate against how many trades you took per week, and marks where you sit on that line. It needs at least 20 closed trades across four or more weeks before it draws. Beside it, the "Disposition Effect" card compares your average holding time on winners against losers.
The emotions. Entry and exit emotion are fields on the trade form, with nine states to pick from. The "Emotion Performance Chart" then draws total P&L for each one.
The numbers. The performance card carries Win %, PF for profit factor, Expect for expectancy, Trades and Avg Hold, and recalculates them as trades close.
Pick one of the ten mechanisms, find the card that measures it, and check where you stand on it. Start the 30-day trial.