Day Trading for a Living? What Happens to Those Who Persist
How many day traders remain profitable after costs over repeated periods?
In the studied Brazilian equity-index futures cohort, 97 percent of traders who persisted beyond 300 days lost money. Use that defined population as a cautionary base rate, then test a proposed strategy on its own costs and untouched period.
The 97 percent finding concerns persistent day traders in the studied Brazilian futures market, not all trading populations.
Evidence map
| Aspect | Finding |
|---|---|
| What it is | The study that answers the survivorship question directly, by following each individual who began day trading a single liquid futures market over several years, including those who quit, and measuring what happened to the ones who kept going. |
| Key result / formula | Chague, De-Losso and Giovannetti study individuals who started day trading Brazilian equity index futures. |
| Why it matters for backtesting | A user evaluating a day-trading strategy should compare its claims with the study's measured population and state differences in instrument, fees, frequency and participant type. |
Key result / formula
Among those who persisted for more than 300 trading days, 97 percent lost money in the reported sample. This result is a strong base-rate warning for that market, period and trader population. It does not establish that every losing trade was caused by chance, or that the same percentage applies to another market and execution setup. The study records realized outcomes and persistence; it cannot reveal every trader's strategy-selection history or settle whether a particular new strategy has a positive expected value.
Why it matters for backtesting
For development, test a proposed change through a paired gain, a placebo and a positive control; evaluate one predeclared sealed batch once. Record variants and abandoned ideas so the search remains inspectable. If selecting the best configuration from a population, use a selection-aware null matched to that population. Do not automatically deflate the result of every new lever by the number of past variants: that is a different statistical question. A backtest still needs after-cost performance and evidence that its execution assumptions can be reproduced live. The Brazilian study informs skepticism, but it is not a substitute for that direct test.
A fair comparison should also report the exact study denominator: traders exceeding 300 days of activity, rather than everyone who opened an account. Otherwise the 97 percent figure is applied to a different population.
Source
Chague, De-Losso & Giovannetti, "Day Trading for a Living?", SSRN 3423101, 2020. Primary source
Related in the library
Guide on this topic: Survivorship bias in backtesting
How the assistant cites the library · The checks behind a verdict