What Percentage of Day Traders Lose Money? The Studies

Most day traders lose money: studies that follow individual day traders find 80% to 97% of them losing after costs, and the losing share rises the longer people keep trading. Day trading means opening and closing a position within one session, paying costs on every round trip (an opening trade plus its closing trade). Broader retail figures that do not isolate day traders still show a majority losing, from about 60% in US futures to 74-89% in EU contracts for difference (CFDs, contracts paying an asset's price change without ownership), while the claim that 90% of traders lose 90% of their money in 90 days has no traceable source.

Key takeaway

Of Brazilians who day traded index futures on more than 300 days, 97% lost money after fees, and in Taiwan fewer than 1% of day traders showed skill that predicted future profits. Costs turn a market that is zero-sum before costs into a losing one for the average trader, and every extra round trip without an edge makes a net loss more certain.

What percentage of day traders lose money, study by study?

Every credible study finds most traders losing; figures run from about 60% to 97% because each source counts a different population and defines a loss differently. Share losing is the number of losing traders divided by the number measured, so the choice of who gets measured largely sets the answer.

Study	Market	Period	Sample	What "lose" meant	Share losing
Chague, De-Losso and Giovannetti (2020)	Brazil mini-Ibovespa futures	2013-2017	1,551 who day traded on more than 300 days (of 19,646 starters)	No net profit after exchange and brokerage fees	97% (1,504 of 1,551)
Barber, Lee, Liu and Odean (2014)	Taiwan stocks	1992-2006	About 277,000 a year day trading over NT$600,000	No positive abnormal return after fees in a year	About 80% a year; over 99% not predictably profitable
Barber, Lee, Liu, Odean and Zhang (2020)	Taiwan stocks	1992-2006	All day traders	Likely to lose money in future day trading	97% (forecast)
ESMA, from national regulators (2018)	EU retail CFDs	Studies before 2018	Retail CFD accounts	Account lost money	74-89%
Ferko, Mixon and Onur, CFTC (2024)	US CME futures, one broker	Feb 2021 to Nov 2022	36,538 retail accounts, end-of-day positions only	Estimated P&L per trading event at or below zero, no fee deduction described	About 60%
tastyfx disclosure (2026)	US retail spot forex	April to June 2026	8,801 active accounts	Not profitable in the quarter	65.5%

The spread is mostly a measurement effect. The low end counts short windows without fees; the high end follows the same people for years after every cost, and that is the number that describes day trading as a practice.

Three measurement choices drive most of the spread. The persistence filter matters most: the Brazil study counts only people who day traded on more than 300 days, while a one-quarter broker snapshot includes traders who were lucky for three months. Cost treatment comes next: the CFTC estimate describes no fee deduction, and the Brazil figures leave out income tax and the cost of trading platforms and courses, which the authors say makes their results overestimate day trading profits. The time window does the rest, because luck averages out over long records and not over short ones.

Do 97% of day traders lose money?

Among futures day traders who persist, yes: Chague, De-Losso and Giovannetti (2020) found that 97% of Brazilians who day traded mini-Ibovespa futures on more than 300 days lost money after exchange and brokerage fees. The authors used records from Brazil's securities regulator (CVM) on everyone who started day trading the mini-Ibovespa, a small contract on Brazil's main stock index traded on the B3 exchange, from 2013 to 2015, and followed them to 2017.

Of 19,646 starters, 1,551 (7.9%) traded on more than 300 days. Of those 1,551, 47 (3.0%) ended with a net profit, 17 (1.1%) earned more than Brazil's minimum wage of about US$16 a day, and 8 (0.5%) earned more than a bank teller's starting pay of about US$54 a day. Measured against all 19,646 starters, only 1 in 418 both persisted past 300 days and finished with a net profit, 1 in 1,156 did so while beating the minimum wage, and 1 in 2,456 while beating the bank teller.

Source: Chague, De-Losso and Giovannetti (2020), using Brazilian securities regulator records. Profit figures are after exchange and brokerage fees but before income tax, platforms and courses, so they flatter the traders.

The share of Brazilian traders with a net profit fell as days traded rose, from roughly 30% of those who day traded once to 3.0% of those past 300 days, and the authors report "no evidence of learning by day trading." Their February 2020 abstract on EconPapers concludes that "it is virtually impossible for individuals to compete with HFTs and day trade for a living, contrary to what course providers claim" (HFTs are high-frequency traders). Versions of this working paper differ slightly: that abstract says 0.4% beat a bank teller, while the June 2020 revision posted on SSRN gives the 1.1% and 0.5% used here.

What percentage of day traders are profitable?

Fewer than 1% of day traders are profitable in a way that repeats, according to Barber, Lee, Liu and Odean (2014) in the Journal of Financial Markets, who studied Taiwan's day traders from 1992 to 2006. Their abstract states: "Less than 1% of the day trader population is able to predictably and reliably earn positive abnormal returns net of fees." An abnormal return is a return above what the market's own movement would explain.

In an average year about 277,000 Taiwanese individuals day traded more than NT$600,000 (about US$20,000), and about 20% of them beat the market benchmark after fees that year. Most of those winners were lucky rather than skilled: the traders whose past results reliably predicted further net profits numbered about 4,000 of roughly 450,000 day traders a year, and as few as 1,000 if commissions sat at the 14.25 basis-point maximum.

Taiwan's costs (a 0.3% tax on every sale plus commissions of 5 to 14.25 basis points, or hundredths of a percent, each way) are stock-market costs, so the Brazil study is the closer match for futures traders.

What percentage of CFD and forex traders lose money?

Between 74% and 89% of retail CFD accounts lose money, according to the national regulators' studies that the European Securities and Markets Authority (ESMA) cited when it restricted CFDs across the EU in 2018. The ESMA announcement of March 2018 says those analyses show "74-89% of retail accounts typically lose money on their investments". The measures that followed required a "standardised risk warning, including the percentage of losses on a CFD provider's retail investor accounts", the origin of the loss percentage that EU CFD providers display.

US retail forex dealers publish a comparable figure under CFTC rule 17 CFR 5.5, which requires the number of non-discretionary retail forex accounts and the percentages that were profitable and not profitable in each of the most recent four calendar quarters. In its public disclosures, tastyfx reported 65.5% of its 8,801 active accounts unprofitable in the quarter to June 30, 2026, up from 59.16%, 62.07% and 64.81% in the three prior quarters. One quarter at one dealer is a short window, which may help explain why the figure sits below ESMA's range, and neither source isolates day traders; futures vs CFD prop firms covers the structural differences.

What percentage of US futures traders lose money?

About 60% of US retail futures traders lose money or break even on a typical trading event (a run of consecutive days holding a position), according to a 2024 CFTC staff paper by Ferko, Mixon and Onur, which states: "In general, we find that retail traders lose money in futures markets." The authors tracked 36,538 retail accounts at one large futures commission merchant (FCM, a futures broker) in the 50 most active CME contracts from February 2021 to November 2022; the median trader's estimated loss was about $100 to $200 per trading event, and the 60th percentile of the distribution sat at break-even.

The CFTC paper is not a day-trading study, and this article found no public dataset that measures US futures day traders after costs. The authors saw end-of-day positions only, so a trade opened and closed within one session is invisible; profit and loss is estimated from settlement prices with no fee deduction described; and the views are the authors', not the Commission's. The sample's median required margin was $3,840, a reference point for how much money you need to day trade futures.

Is the 90-90-90 rule true?

No primary source supports the claim that 90% of traders lose 90% of their money in 90 days: this article's search for its origin found only blogs, forums and trading-education pages repeating it. The verified data miss the claim in both directions. Among persistent Brazilian futures day traders the losing share (97%) is higher than 90%. Loss size in the CFTC sample depends on how long people keep trading: the typical one-time trader lost an estimated 7% of initial margin (the deposit required to hold a position, not the whole account), while the typical trader in the most active fifth, with about 25 trading events, lost an estimated $5,350, or 1.69 times initial margin. None of the sources measures how fast accounts are lost, so the 90-day part is untested.

Why do most day traders lose money?

Most day traders lose money because short-term trading is close to zero-sum before costs and negative-sum after them. In futures every long position has a matching short, so gross gains and losses cancel across all traders (the reason futures trading is zero-sum), and commissions and exchange fees and slippage are then subtracted from that zero total.

In an illustrative model (inputs chosen for teaching), a trader with no edge whose round trips are independent and normally distributed around $0 before costs, with a standard deviation of $100, and who pays $4 per round trip, is down after n round trips with probability Φ(c√n ÷ σ) = Φ(4√n ÷ 100), where Φ is the standard normal distribution function, c the cost per round trip and σ one trade's standard deviation.

Round trips	Expected net, $4 cost	Standard deviation	Share losing, $4 cost	Share losing, $10 cost
1	-$4	$100	51.6%	54.0%
10	-$40	$316	55.0%	62.4%
100	-$400	$1,000	65.5%	84.1%
300	-$1,200	$1,732	75.6%	95.8%
1,000	-$4,000	$3,162	89.7%	99.9%
3,000	-$12,000	$5,477	98.6%	over 99.9%

At $4 a round trip, the zero-edge trader reaches the Brazil study's 97% after about 2,211 round trips, roughly 7 a day for 300 days; at $10 it takes about 354. The model needs no negative skill, only costs and repetition, and it mirrors the Brazil pattern of profit shares falling as days traded rise.

Illustrative model: each round trip is a fair $100 coin toss minus cost, so the only thing working against the trader is the fee. Repetition alone reaches the Brazil figure, with no bad decisions required.

Losing traders also keep trading. Barber, Lee, Liu, Odean and Zhang (2020) found in the Taiwan data that "74% of day trading volume is generated by traders with a history of losses" and predicted that 97% of day traders are likely to lose money in future day trading. The authors call the pattern consistent with overconfidence and biased learning rather than rational learning, the behavioural side covered in trading psychology for funded accounts.

What do profitable day traders have in common?

The few profitable day traders in the evidence share one trait above all: they were already profitable. Barber and colleagues (2014) ranked Taiwan's day traders by one year's risk-adjusted returns, and in the following year the top 500 earned abnormal returns of 61.3 basis points a day before fees and 37.9 after, while the bottom-ranked earned -11.5 before fees and -28.9 after. Costs took 23.4 points a day from the best group and 17.4 from the worst, so past results predicted future ones, and only a large gross edge survived the fees. Beyond past performance, concentrating on a few stocks and short selling predicted success only modestly.

Past results did predict future ones, which is the evidence that skill exists. It also shows the price of admission: only a large gross edge survived fees that took a third of it.

Skill also arrives with heavy noise. The best of the 1,551 persistent Brazilian traders earned US$310 a day with a daily standard deviation of US$2,560, more than eight times the daily average, and far less than Taiwan's top group, a gap the Brazil authors suggest may reflect competition from high-frequency traders. Assuming independent days, even that trader needed about 273 trading days before the average sat two standard errors above zero, since (2 × 2,560 ÷ 310)² ≈ 273. Judging a record therefore means measuring results net of costs in risk units, which R-multiples make possible, over a horizon discussed in how long it takes to become a profitable trader.

Does a funded prop account change the odds?

A funded prop account caps a trader's direct loss at the evaluation and account fees, but it does not change the trader's edge, because the same skill and the same costs decide every trade. In this article's illustrative zero-edge model ($100 swings, $4 per round trip), 300 round trips end in a net loss 75.6% of the time in either setting; a firm's drawdown limit only stops some losing runs earlier, and the worst case differs, the account balance in one setup and the fee in the other. Repeated evaluation fees add up the way round-trip costs do (see the common reasons funded traders fail and what funded traders actually earn).

A funded account caps what a losing run can cost; it does not change the probability of the losing run. Repeated evaluation fees behave like round-trip costs and accumulate the same way.

A trade copier does not change the odds either. Thor, this blog's own product, copies one master account's trades to many accounts, so it multiplies whatever edge the master has, including a negative one: a losing strategy copied to five evaluations multiplies the losses and the fees by five. A copier earns its cost only after a trader's own net record, over hundreds or thousands of trades, shows an edge worth scaling.

Go deeper

Frequently asked questions

Do day traders get better with experience?

Rarely, on the available evidence. Barber, Lee, Liu, Odean and Zhang (2020) found in Taiwan that "unprofitable day traders are more likely than profitable traders to quit", so a group of survivors can look better over time because losers leave, not because anyone learned.

Can a skilled trader still end up losing money?

Yes, often, over short samples. With this article's illustrative $100 trade swings, a trader with a genuine $2 net edge per round trip still finishes 300 round trips in the red about 36% of the time, and 1,000 round trips about 26% of the time.

How many trades does it take to know whether you have an edge?

Usually thousands, not dozens. With illustrative $100 trade swings, a true $2 net edge per round trip needs about 6,800 round trips before its expected total reaches 1.645 standard errors, the usual one-sided 95% confidence bar. The count scales with (1.645 × swing ÷ edge) squared, so a larger edge relative to the swing needs far fewer trades.

How much does a losing retail trader lose on average?

No single average applies: ESMA reported average losses per retail CFD client ranging from €1,600 to €29,000 across the national studies it cited in 2018. The spread reflects different countries, periods and account sizes, and none of those averages isolates day traders.

Are these studies too old to apply today?

No, newer data point the same way. The Taiwan sample ends in 2006, but the Brazil study runs to 2017, the CFTC paper to November 2022 and the tastyfx disclosure to June 2026, and every one shows a majority losing.

Do ordinary investors lose money by trading too?

As a group, yes. Barber, Lee, Liu and Odean (2009, Review of Financial Studies) found in Taiwan data from 1995 to 1999 that the aggregate portfolio of individual investors suffered a performance penalty of 3.8 percentage points a year, losses equal to 2.2% of Taiwan's GDP, while institutions gained 1.5 points a year. The authors traced virtually all of those individual losses to aggressive orders.

Are the odds better in futures than in CFDs or forex?

No like-for-like comparison exists, because the futures, CFD and forex figures use different populations, windows and loss definitions. The reliable difference is structural: futures trade on a central exchange with published fees, while a CFD or retail forex dealer is often the counterparty to the client's trade.

Is day trading just gambling?

For most participants the results behave like a negative-expectation game, because costs are paid on every round trip whatever the outcome. The difference from a casino is that a small minority show repeatable skill after fees, which no roulette strategy can produce.

What percentage of prop firm traders are profitable?

No independent, industry-wide figure could be verified for this article. Individual firms sometimes publish their own pass or payout numbers, but the definitions (passing an evaluation, receiving one payout, staying funded) differ and none is comparable with the net-profit measures in the academic studies.