How Many Crypto Traders Are Profitable? Data From 1,076 Hyperliquid Traders

Jay Sharma · Founder, Botsfolio

Published October 19, 2026

"90% of traders lose money" is one of the most repeated lines in trading. It is rarely sourced, and it usually comes from forex brokers or old stock studies. Crypto has a better source: on Hyperliquid, a large crypto exchange, every trade sits on a public blockchain, so anyone can check who made money and who didn't.

So we checked. Here is what 1,076 real human traders did with real money, how the winners differed from the rest, and what the top of the leaderboard really looks like.

The short answer: about 3 in 10

31.4% of the human traders in our sample made money over their own active period, from their first trade to their last, between May 2024 and October 2026. 68.6% lost money.

That figure comes from accounts that traded at least $50,000 in total, so these are active traders, not people who tried it once. The typical one held about $61,000 at their peak and traded for about a year.

How we counted

• The pool. Hyperliquid's public leaderboard lists 47,391 accounts. 32,839 of them had traded at least $50,000, after removing market makers.
• The sample. We drew 2,500 of those at random. 1,844 could be measured; the rest moved money in and out too often between data points to measure fairly.
• People vs bots. We checked each account's trades over its last 14 active days. Accounts trading 150 or more times a day, or around the clock, count as bots. That gave 1,076 humans, 340 bots and 428 accounts with no recent trades to classify.
• Profit. We used Hyperliquid's own trading profit and loss: closed trades, fees and funding. Deposits and withdrawals don't count. We rebuilt the profit from actual trades for 59 accounts, and "made money or lost money" matched for 58 of them.
• Returns. Return is the result as a share of the most money the trader ever held, which makes losses look smaller, not bigger.

The numbers

Two numbers stand out. Half of these traders lost 90% or more of their largest balance at some point, which is what leverage does when a trade goes the wrong way. And two thirds of them would have done better by simply buying Bitcoin and Ethereum and doing nothing, over the very same dates.

Bigger accounts did better, but most still lost

Small accounts had it worst: only about 1 in 5 traders with $1,000 to $10,000 made money, and nearly 3 in 4 lost 90% or more at some point. Even among traders with $1 million or more, most lost money.

The winners fell less

The clearest difference between traders who made money and those who didn't was not how often they traded or how long they stayed. It was how far they let their account fall:

Traders who never let a fall pass 25% made money 71% of the time. Traders who lost 90% or more made money 8.6% of the time. The other habits we checked mattered far less:

• How often they traded. Traders making under 5 trades a day made money 28.2% of the time; 5 to 20 a day, 29.3%; 20 to 50 a day, 28.5%; 50 to 150 a day, 37.5%.
• How long they traded. By time active, the share that made money ranged from 27.7% to 37.9%, with no steady pattern.

Even the winners had deep falls

Among the 338 traders who made money, the typical one made 34.9% on their largest balance. But 40.2% of them fell 50% or more along the way, and 13.6% lost 90% or more at some point before recovering. Making money as a trader usually still means sitting through falls most people can't.

The top of the leaderboard is mostly bots

Hyperliquid's leaderboard looks like a list of star traders. We sampled 150 accounts from the top 1% by profit and checked how they traded. Of the 78 we could classify, 53 were bots and 25 were humans. The bots made a median of about 900 trades a day; the humans, about 23.

Bots did better in the main sample too, but not by much: 42.9% of the 340 bot accounts made money, against 31.4% of humans, and 40% of the bots lost 90% or more at some point. The leaderboard's own percentage returns are also unreliable. They divide by whatever balance happened to be in the account when the period started, so one account showing a 1,731% "24 hour return" had made $1,731 on a $219,000 account. We never used them.

How this compares with other studies

Our 31.4% sits in the same range as other research on retail traders. The methods differ a lot, so read these as context, not like-for-like numbers:

Sources: ENVY Protocol, The Coin Republic, BIS Working Paper 1049, ESMA, Barber et al., Chague et al..

The short-window Hyperliquid snapshots find fewer winners, 13% to 17%, partly because one month catches many traders in the middle of a losing stretch. Measured over a trader's whole active period, more of them come back above water, but still fewer than a third. Across every study, the pattern is the same: most retail traders lose, and very few win reliably.

Why most traders lose

The data points to the same few causes again and again:

• Leverage. Half of the traders lost 90% of their largest balance at some point, and 23% were liquidated at least once in just their last 14 active days. Across all exchanges on October 10, 2025, about 1.6 million traders were liquidated in 24 hours, according to CoinGlass data reported by Decrypt.
• Big falls. Once a trader let a drop pass 60%, the odds of ending in profit fell below 40%.
• Chasing prices. A study of 200,000 eToro accounts found the same people who buy dips in stocks chase rising prices in crypto, and JPMorgan Chase Institute data shows most US crypto users made their first purchases during price spikes.
• Emotion. Chasing losses, cutting winners early and moving stops. We cover the three habits that show up most in losing records in why crypto traders lose money.

What would have happened with the same money held, or traded by an AI?

Simply holding half Bitcoin and half Ethereum over each trader's exact dates did better than 66.9% of them.

We also gave Botsfolio AI, running as the BTC & ETH Cycle Trader, the same money on each trader's first day and ran it to their last. Tested on those exact dates with spot coins and no leverage, it made money in 88.6% of the windows. The traders used real money, mostly with leverage, so this is a test of the AI on the same dates, not a trade-by-trade race. Some traders did beat it: 2.4% on both return and worst drop, and about 3 in 4 of the profitable traders on return alone, usually with much deeper falls. The full head to head is in AI vs 1,076 human traders.

What this means if you trade

• Expect to be in the 69% unless you have a reason not to be. Most active traders, including big accounts, lost money over this period.
• Your worst drop matters more than your best trade. The traders who kept falls small were the ones who made money. Our guide to drawdown explains why.
• Compare yourself with holding. If your trading doesn't beat holding the same coins over a year, it isn't adding value. See how to read a buy and hold comparison.
• Test before you risk money. Our guides to how long to paper trade and how to become a profitable crypto trader cover the process.
• Or hand the trading to rules. You can watch Botsfolio AI trade live, paper trade your own AI portfolio for free, or read do AI trading bots actually work first.

Method and limits

How we measured Botsfolio AI's side is on the portfolio methodology page.

FAQ

Educational analysis, not financial advice. Trader results come from Hyperliquid's public data. Botsfolio AI's results are tested on past prices. Trading crypto can lose money. Past performance does not predict future results.