If you’ve ever assumed a losing stretch means you need a better signal, a faster feed, or one more indicator, there’s a large and fairly consistent body of research saying that’s probably not where the problem is. The data — collected from actual brokerage accounts and exchange records, not surveys — keeps pointing somewhere less comfortable: behavior, not information.
What the Data on Real Trading Accounts Actually Shows
The foundational study here is Brad Barber and Terrance Odean’s analysis of 66,465 U.S. households’ brokerage accounts, which found that the most active traders underperformed the market by a wide margin — earning an average of 11.4% annually against the market’s 17.9% over the same period. These weren’t uninformed amateurs with no data access; they had the same public information everyone else did. The gap came from what they did with it: trading more, not trading smarter.
More recent work on retail day traders specifically tells an even starker version of the same story. A study of Brazilian futures day traders by Chague, De-Losso, and Giovannetti tracked everyone who started day trading over a two-year window and found that 97% of those who persisted lost money, with only about 1.1% earning more than Brazil’s minimum wage from it. A separate long-running analysis of Taiwanese day traders (Barber, Lee, Liu, and Odean) found aggregate day-trading performance was negative in every single year studied — and, notably, that most unprofitable traders kept trading anyway rather than stopping.
Crypto’s 24/7 markets, high volatility, and low barrier to entry make this dynamic easier to fall into, not harder — there’s no closing bell forcing a pause, and the same information (a chart, an order book, a news feed) is available to literally everyone trading the same pair at the same moment.
The Missing Variable Is Temperament, Not Data
If everyone has roughly the same information, what actually separates the accounts that lose badly from the ones that don’t? A study by Lo, Repin, and Steenbarger that measured the physiological stress responses of day traders while they traded found that traders with more intense emotional reactions to price moves performed worse — not because they understood the market less, but because the reaction itself degraded the decision. As Holdy Lab’s own breakdown of this research puts it, “temperament is the key word” — and importantly, the same research found no single fixed “trader personality type” that predicts success, which is actually the encouraging part: if it were pure temperament, there’d be nothing to do about it. Since it isn’t, behavior is something you can actually observe and change.
What Actually Follows From This
None of the studies above suggest trading is unlearnable or that losses are inevitable — they suggest the lever most people aren’t pulling is the right one. A few concrete, low-cost starting points:
- Record the reasoning behind a trade, not just the result. A win or loss tells you an outcome; a written reason tells you whether the process behind it was sound, which is the thing you actually control.
- Look for patterns across many trades, not one. Any single trade is too small a sample to judge a habit by — the studies above only found their effects by looking at hundreds or thousands of trades per trader.
- Change one behavior at a time. Trying to fix sizing, timing, and stop placement simultaneously makes it impossible to tell which change actually helped.
- Practice the mechanics on a simulator before any of this matters with real capital — our own guide to researching a project before you buy covers the other half of this (what you’re buying); this is about how you behave once you own it.
If you want a deeper, well-sourced walkthrough of the specific studies cited above, Holdy Lab’s article is worth reading directly: Why Most Retail Traders Lose: It’s Behavior, Not Information.
Frequently Asked Questions
Does this mean information and research don’t matter at all?
No — research still matters for picking what to trade and understanding risk. The point is narrower: once two traders have access to the same basic information, the gap between their results is explained far more by what they do under pressure than by who read one more article.
Is day trading crypto just statistically a losing proposition?
The studies cited above (Brazil, Taiwan) are about day trading broadly, not crypto specifically, but there’s no reason to expect crypto’s faster, more emotionally intense price action to produce better behavioral outcomes than slower markets did in that research — if anything, the opposite case is easier to make.
If temperament isn’t fixed, how do you actually change it?
The research doesn’t point to willpower as the fix — it points to structure: rules that don’t depend on your emotional state in the moment (position-size caps, pre-set stop distances, a cooldown after a loss) tend to outperform relying on self-control during exactly the moments self-control is weakest.
Why do unprofitable traders keep trading instead of stopping?
The Taiwan study didn’t fully explain this, but it’s consistent with a well-documented pattern elsewhere in behavioral finance: people chase losses hoping to get back to even rather than cutting their losses, a dynamic closely related to the disposition effect and revenge trading covered elsewhere in our Trading Psychology section.