Why 90% of Crypto Traders Lose Money: 3 Mistakes We Actually Measured

Jay Sharma · Founder, Botsfolio

Published August 18, 2026

The "90% of traders lose money" statistic is one of the most repeated lines in trading content. It is also almost always followed by generic advice: "use stop losses," "risk management is key," "control your emotions."

None of that answers the actual question. Which mistakes? How much do they cost? How would you see them in your own trades?

We can answer that. When we look at losing traders' records (CSV exports, exchange history, connected accounts), the same three behavioral patterns show up over and over. Not chart analysis mistakes. Not entry timing mistakes. Behaviors that happen AFTER the trade is on, that quietly convert edge into losses.

Here they are, with what they look like on the PnL and how to catch them.

Mistake 1: Position size spikes after a loss (revenge sizing)

The pattern: a trader takes a loss. On the very next trade, their position size is 30-80% larger than their average.

You can measure this in one line of Python. For every closed loss, take the notional (or risk) of the immediately-next trade. Average them. Compare to the trader's overall average risk. In losing accounts, the post-loss ratio is consistently 1.3-1.8. In profitable accounts, it sits at 0.95-1.05 — essentially unchanged.

What it looks like on the PnL: your equity curve has a normal downward drift, then a sharp cliff. That cliff is almost always a 2-3x-sized position that took a stop right after a losing trade. Losing traders do not blow up from bad analysis. They blow up from a single oversized trade taken in an emotional state.

Why the brain does it: loss aversion. Neurologically, a $1000 loss feels roughly twice as painful as a $1000 gain feels good. The brain wants that pain gone as fast as possible. Sizing up on the next trade is the shortest path to "making it back" — or so it feels. In reality, it is the shortest path to a much bigger loss.

What profitable traders do: they follow a hard rule. Post-loss position size is the same as pre-loss. Some go further — they cut size in half for the next three trades after any loss, to force emotional cool-down. Either works. What matters is that the decision was made in advance, not in the moment.

Mistake 2: Winners get cut, losers get held (the disposition effect)

The pattern: average hold time on winning trades is shorter than average hold time on losing trades.

Measurable: split closed trades into wins and losses. Compute the mean hold duration in each bucket. Divide winners by losers. Losing accounts consistently sit below 0.9 — meaning they hold their losers longer than their winners. Profitable accounts sit above 1.5.

What it looks like on the PnL: your R multiples on wins are 0.6-0.8R, and your R multiples on losses are 1.0-1.5R. You have positive win rate but negative expectancy. Every trade is a small win or a large loss, and the math slowly bleeds you.

Why the brain does it: the same loss aversion, seen from the other side. When a trade is winning, the brain fears losing what it has. So it takes profit early ("lock it in!"). When a trade is losing, the brain fears realizing the loss. So it stays in ("give it room to recover"). Both feelings are correct given the pain math. Both destroy your expectancy.

This is the single most-studied bias in trading academia, going back to Shefrin and Statman's 1985 paper. Every trader knows about it. Almost none of them fix it, because knowing about a bias and overriding it in the moment are different things.

What profitable traders do: they set the target and stop before entry, and they do not renegotiate mid-trade. If the trade hits target, exit. If it hits stop, exit. If it does neither, hold. The decision is mechanical, not emotional. This is boring and effective.

Mistake 3: The stop moves mid-trade (stop drift)

The pattern: a trade is entered with a stop at level X. Then the trader watches the price approach X. Then they move the stop to level Y (further out). Then price hits Y instead.

Measurable: harder to compute from a CSV, because most exports do not show the original stop. But if you connect an exchange with API history or use a tracking tool, this shows up as multiple stop-loss modifications per trade.

What it looks like on the PnL: your average loss size drifts upward over time. Trades that "should have" been 1R losses become 1.5R, 2R, 3R losses. Your win rate holds. Your expectancy collapses.

Why the brain does it: sunk cost. Once you have invested time, attention, and pain into a trade, closing it at the original stop feels like admitting the whole idea was wrong. Moving the stop feels like "giving the trade room to breathe." It is almost never the trade's issue. It is your relationship with being wrong.

What profitable traders do: they treat the original stop as inviolable. Once entered, the only allowed stop modification is to move it in your favor (trailing to breakeven, or trailing profit). Never further out. This is enforced by rule, not willpower. Some traders write it on their monitor. Some set the stop as a hard exchange order so they cannot move it easily.

Why these three, specifically

There are dozens of trading mistakes in the literature. Chasing tops. Ignoring higher timeframe context. Trading during low-liquidity hours. All real, all worth fixing.

But these three are the ones that show up most consistently in losing records. And they share a common structural feature: they all happen in response to emotion, not analysis. Revenge sizing is a response to the pain of a loss. Cutting winners is a response to the fear of losing gains. Moving stops is a response to the discomfort of being wrong.

That is why they are so hard to fix by reading about them. You cannot out-analyze an emotional response with more knowledge. The fix is procedural: rules set in advance, enforced by mechanism, reviewed after the fact.

What profitable traders actually do

Not motivational. Specific.

1. Position size rule: fixed percent of account per trade. Same before and after any loss. If a review shows this was violated, the trader logs the violation and reviews weekly.

2. Target and stop pre-defined: written down before entry. Alerts set at both levels. Once triggered, exit. No mid-trade renegotiation, ever.

3. Stop modification rule: only movable toward breakeven or into profit. Never further out. Some traders set the stop as a hard exchange order to enforce this.

4. Review cadence: weekly review of the last 7 days of trades. For each: was the entry within setup criteria? Was the exit within the plan? Was the size correct? Was the post-loss behavior clean? Any "no" gets logged. Patterns get addressed.

None of this requires any software. All of it can be done with a spreadsheet and a set of hard rules. What software helps with is measurement — computing the numbers so you know if the rules are actually being followed, or if you are drifting.

How to see them in your own history

Export your last 100 closed trades from your exchange. Then:

1. Post-loss size check: for every loss, note the notional (or risk) of the next trade. Average them. Compare to your overall average. Ratio above 1.2 means you have this problem.

2. Winner vs loser hold time: split into wins and losses. Compute average hold. Winners divided by losers below 1.0 means you have this problem.

3. Stop drift: harder, requires stop-modification logs. If you use an exchange API, pull the position modification events. Any pattern of stops moving further away from entry mid-trade means you have this problem.

If you connect your exchange to the companion or upload a CSV of your last 100 trades, all three checks are automated and returned as a scorecard in about a minute. Or do it manually. The tool matters less than the audit.

FAQ

Educational analysis, not financial advice. Past performance does not predict future results.