In 1992, a retail trader in Taiwan paid real money for a broker, a phone line and a gut feeling. In 2026, you get commission-free execution, TradingView on three monitors and a chatbot that explains order flow. Tools got dramatically better.
The results didn't.
The numbers haven't moved
Barber, Lee, Liu and Odean studied every day trader on the Taiwan Stock Exchange over about fifteen years. Fewer than 1% were reliably profitable year after year. The rest paid for the privilege.
Then Brazil. Researchers followed more than 19,000 people who tried day trading futures between 2013 and 2015. Of those who kept going for at least 300 days, 97% lost money. Only about 1% earned more than a minimum-wage salary.
Different country, different decade, different asset. Same outcome. Crypto looks no kinder to me, and I've watched plenty of people prove it with their own accounts.
What that tells you
Here's the thing. Fees collapsed. Data became free. Charting went from expensive software to a browser tab. Education is everywhere, and half of it is decent. If the failure rate depended on any of those, it would have dropped.
It didn't, so the problem isn't access. It isn't information either.
The problem sits in the one component nobody upgraded: the person pressing the button. Same loss aversion as 1992. Same urge to close a green trade early and hold a red one. Same itch to do something when the screen is moving.
Better tools make that person faster. Faster isn't better. A trader who panics at 2am now panics with one tap instead of a phone call.
There's a second layer. Costs fell, but so did the friction that used to slow people down. When a trade took a phone call and a $50 commission, you thought twice. Now a bad idea costs nothing to try forty times a day, and the spread and slippage quietly collect on every one.
Why most people get this wrong
Most traders respond to failure by shopping for another input. A new indicator. A new course. A faster platform. I did this for about two years. Every purchase felt like progress, and none of it touched the thing that was actually costing me money.
It's an understandable mistake. A tool is easy to buy and easy to measure. Your own behavior is harder to look at, especially when the evidence is sitting in your trade history.
Ask yourself a blunt question. If you took your last 100 trades and replaced your entries and exits with a fixed set of rules, would the result have been better or worse? If you honestly don't know, you haven't measured your edge yet. You've measured your mood.
Don't read this as a lecture. I'm not saying you're undisciplined. I'm saying discipline is a bad thing to rely on, because it runs out at exactly the moment you need it, usually on the third red day in a row.
What to do with this
Stop upgrading inputs and start constraining behavior. Write down your entry rule, your exit rule and your position size before the session starts. Then count how many trades you took that matched the written rules, and how many didn't.
In my experience, the gap between those two groups is the whole story. The rule-following trades are boring and roughly fine. The improvised ones are where the account goes to die.
That's also the honest case for automation. Not that a bot is smarter than you. It's that a bot is the only participant in this 30-year dataset who doesn't get tired, bored or scared.
If you want to see what a fully rule-based approach looks like over thousands of trades, including the long ugly stretches, the full data and methodology are at v33systematic.com.