Is Day Trading Worth It? What the Research Found, and How Long It Takes to Know
Whether day trading is worth it depends on having an edge that survives costs, and the published evidence says most people do not. In complete Taiwan exchange records, day traders as a group lost money in every year studied after costs, and most stopped within two years.
Is day trading worth it? The question has two parts that get mixed together. One is whether day traders as a group come out ahead, and researchers with complete exchange records have answered that. The other is whether one particular person has an edge, and that can only be answered by trading long enough to measure it. This page takes the published evidence first, then works out how long the second answer takes.
How it works: what “worth it” has to beat
It has to beat doing nothing. Money that is not day traded can sit in an index fund, so the honest comparison is against holding the market, not against zero. A strategy that makes a small profit but trails a plain index was still a cost.
It has to beat its own costs. Commissions, spreads, fees and taxes come out of every round trip. A trader can pick direction slightly better than chance and still lose once those are paid, which is the pattern the Taiwan data below shows.
It has to be worth the hours. Day trading takes the market session itself. The time spent is part of the price, whatever the account shows.
And the rules around it changed in 2026. The US pattern day trader rule and its $25,000 minimum were eliminated by FINRA effective 4 June 2026, with brokers allowed to phase the change in until 20 October 2027. The pattern day trader page covers what replaced it. A lower entry bar changes who can start; none of the research below depends on it.
What the research found
Taiwan, every day trader for 15 years. Brad Barber, Yi-Tsung Lee, Yu-Jane Liu, Terrance Odean and Ke Zhang studied day traders on the Taiwan Stock Exchange from 1992 to 2006. In the December 2018 version of their paper, day traders on average lost 7 basis points a day on their day trading before costs, and costs more than tripled that to 23.9 basis points a day. A basis point is 0.01%. Net performance was reliably negative in every year, and gross performance in every year but 1992. The paper was later published in the Review of Asset Pricing Studies in 2020, and its published abstract does not restate these figures.
Most of them stopped. Of those who day traded on at least 10 days, only 2.5% dropped out within a month, but 44% were still day trading after one year, 24% after two and 15% after three. The authors’ abstract adds that unprofitable traders were more likely to quit, yet a great many persisted despite persistent losses.
Heavy trading hurt ordinary investors too. Barber and Odean’s 2000 study of 66,465 households at a large US discount broker, 1991 to 1996, found that the households trading most earned 11.4% a year while the market returned 17.9%. The average household earned 16.4% and turned over 75% of its portfolio a year. These were active investors, not day traders, but the direction matches.
And across a whole market. A 2009 study by Barber, Lee, Liu and Odean used every trade in Taiwan and found the aggregate portfolio of individual investors suffered an annual performance penalty of 3.8 percentage points. Their losses equaled 2.2% of Taiwan’s gross domestic product, while institutions enjoyed an annual performance boost of 1.5 percentage points.
Brazil points the same way. Fernando Chague, Rodrigo De-Losso and Bruno Giovannetti followed everyone who began day trading Brazil’s equity futures between 2013 and 2015. Their abstract says 97% of those who kept at it for at least 300 days lost money, and calls it “virtually impossible” for individuals to compete with high-frequency trading firms and day trade for a living. The full counts are on the how much day traders make page.
There is a dissent. A paper posted to arXiv argues that the evidence does not show day trading as an occupation to be economically unsustainable, nor that traders cannot improve. Its abstract describes it as a counterpoint to a 2019 version of the Brazil study by Chague and Giovannetti.
How long it takes to know
An edge has to be told apart from luck. A common yardstick is two standard errors: the average daily result has to sit at least two of its own standard errors above zero. The number of sessions that takes is (2 x daily swing / daily edge) squared.
Measured on SPY. Over the ten years from 26 September 2016 to 25 September 2026, SPY’s open-to-close return, one session with nothing held overnight, had a standard deviation of 0.856%, or 85.6 basis points (85.58 before rounding, the figure used below). Taking that as the day-to-day swing of a full-size position held through the session:
- A 2 basis point daily edge needs 7,324 sessions, about 29.1 years.
- A 5 basis point edge needs 1,172 sessions, about 4.7 years.
- A 10 basis point edge needs 293 sessions, about 1.2 years.
- A 20 basis point edge needs 73 sessions, about 0.3 years.
Even the market’s own drift is slow to prove. SPY’s daily total return from 1 February 1993 averaged 0.0478% with a standard deviation of 1.167%. By the same test that takes 2,389 sessions, about 9.5 years of trading days, for a rise the market has shown for decades.
A worked example
A $10,000 position and a 5 basis point edge. Suppose, hypothetically, a trader’s method makes 0.05% a day on average on a $10,000 position held from open to close.
- The edge in dollars. $10,000 x 0.0005 = $5.00 a day, on average.
- The swing in dollars. At SPY’s measured 0.856% open-to-close standard deviation, a typical day moves the position by about $85.58 either way.
- Sessions to see it. (2 x 85.58 / 5.00) squared = about 1,172 sessions, which is 4.7 years at 252 sessions a year.
- What a single year looks like. Over 252 sessions the edge adds up to about $1,260, while the year’s result has a standard deviation of about $1,359. Under a normal approximation, about 17.7% of such years would still end in a loss, with the edge perfectly real.
Costs come off the $5, not the $86. A round trip that costs a few basis points consumes most of a 5 basis point edge, and the Taiwan traders’ costs were larger than their gross losses.
The original data
Two figures for the same question. The first chart is the survival pattern from the Taiwan study; the published figures used on this page are in the research figures file, each with the paper’s exact wording.
Reading the survival figures. The steepest drop came in the first year: 56 of every 100 who reached 10 days of day trading had quit by the one-year mark. Between the first and second years another 20 in 100 stopped, and between the second and third, 9 more. Quitting here means a full 12 months with no day trades, so a trader who paused for six months still counts as active.
How the question is asked online. In the site’s study of 24,971 unique finance videos, 114 titles contain a form of “quit”, a job or a 9-to-5, and a trading word (“trad” or “forex”). They come from 90 channels, at a median of 8,783 views. Only 20 titles pair “quit” with a trading word and no job, at a median of 5,521, and several of those are about moving from one kind of trading to another rather than stopping. The Taiwan survival figures show how often stopping happens in practice.
The second is the SPY arithmetic. The measured standard deviation and the sessions needed for each edge size are in the edge detection file, recomputed from 8,472 SPY daily bars.
When it fails: reading the evidence wrongly
Treating the average as a verdict on one person. The studies describe groups. Some Taiwanese day traders did earn reliable profits, and the how much day traders make page gives their numbers. The group result says the odds are poor, not that no one succeeds.
Treating a good year as proof. At the swings measured above, one profitable year says little about a small edge. A strong first year and a real edge are different things.
Leaving out the time. A comparison that counts only money ignores the session hours, which full-time trading treats as part of the arithmetic.
Assuming the old markets are today’s. The studies cover Taiwan from 1992 to 2006, US households from 1991 to 1996 and Brazilians who started between 2013 and 2015. Costs, rules and competition have changed since, in both directions.
Using SPY’s swing for every trader. The 85.6 basis point figure fits one full-size SPY position held through the day. Smaller positions, several trades a day or a more volatile stock change the swing, and the sessions needed change with its square.
Related
How much day traders make reports the Brazil and Taiwan income figures in full. Full-time trading works through what a monthly withdrawal does to an account. Day trading explains what the style is and what each round trip costs. And expectancy shows how to turn a trading record into the average per trade that this page’s arithmetic starts from.
Before deciding it is worth it, write down the edge you think you have in basis points a day, and work out how many sessions it would take to see it through the noise. If the answer is years, trade small while you find out.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.