Why Do Most Day Traders Fail? The Statistics and the Mechanism

The headline numbers are not controversial. European regulators require CFD brokers to publish the share of retail accounts that lose money, and those disclosures have consistently sat around 70 to 80 percent. Academic work is harsher on persistence: a widely cited study of Brazilian futures day traders found that among individuals who persisted for hundreds of sessions, some 97 percent lost money, and the tiny remainder mostly earned less than minimum wage for the hours involved. Exact figures vary by market and period. The shape does not.

The interesting question is not the rate. It is the mechanism, because the mechanism is what an individual can actually do something about.

Mechanism one: costs eat small edges

Day trading means many trades, and every trade pays a toll. A strategy with a genuinely positive gross edge fails anyway when the per-trade cost exceeds it, and the failure is disguised as bad trading. We audited a book where exactly this happened: 11,888 trades, positive gross P&L, and costs at 4.3 times the per-trade edge, netting minus $72,983. The equity curve declined in a nearly straight line, which is the signature of constant drag rather than bad decisions. The decomposition is public in our case study.

The frequency of day trading multiplies whatever cost inefficiency exists. The same edge, expressed in fewer, larger trades, can be viable. Most failing day traders have never computed their gross versus net, and so cannot see which problem they have.

Mechanism two: the edge was never measured

Most day traders cannot state their expectancy, their sample size per setup, or the conditions under which their results were earned. Without those numbers, every decision is vibes with a brokerage account. Losing streaks trigger strategy changes at precisely the moments a measured trader would recognise as ordinary variance, and winning streaks trigger size increases on evidence that would not survive a sample-size check. The strategy churn itself then guarantees no setup ever accumulates enough trades to be judged.

Mechanism three: variance reads as skill

Short-horizon trading produces streaks. A zero-edge process generates runs of wins long enough to build total conviction, and the trader who scales up mid-streak converts a harmless coin flip into an expensive one. The Brazilian data's most sobering detail is persistence: the failure rate barely improved with experience, because experience of noise teaches noise.

What the survivors share

Not a secret setup. Measured expectancy net of costs, on samples large enough to mean something, in conditions they can name, tested against data the conclusions were not fitted to. Every element of that sentence is unglamorous and checkable, which is precisely why it is rare.