Copy trading leaderboards: what they rank, and what they miss
Copy trading lets you follow another trader's positions automatically, and the leaderboard is how most people choose whom to follow. What a leaderboard sorts by decides who reaches the top of it, and the most common sort, the highest recent return, is the one most easily won by luck.
Key takeaways
- Copy-trading leaderboards such as Bybit's rank traders by return over short windows, such as 7 or 30 days.
- Sorting many traders by a short-window return tends to put the luckiest at the top, and leverage multiplies it.
- None of the boards we looked at compares a trader with simply holding Bitcoin.
- A record worth trusting is long enough, steady enough, ahead of its benchmark, and repeats.
What the leaderboards rank
Bybit publishes its rules in its help centre. Its "Top by ROI" board first filters traders, for example on recent activity, a maximum drawdown of 40% over 30 days and a positive return over 7 and 90 days, and then sorts them by return over the last 7 days. Its "Top Balanced Traders" board sorts by a weighted score that includes 30-day return, maximum drawdown, volatility and Sharpe and Sortino ratios, which is a step towards measuring risk. Its "Top New Talents" board is for traders with fewer than 30 trading days.
Each of these measures a trader against cash. A trader who did well because the market rose looks the same as one who did well by trading it, and a week or a month is the window.
Why short windows reward luck
Crypto returns swing hard from day to day, and over a few days luck can outweigh any skill. We measured how much on Sharpe itself: in simulations built from the day shapes of real records, a trader with no edge over Bitcoin reached Diamond or above in a single 90-day window about one time in four. A week is far noisier than 90 days.
Sorting thousands of traders by last week's return then picks out the ones whose week went best. Some of them are skilled, and many are simply the luckiest of a large crowd. Leverage makes it worse. On perpetual futures a trader running ten times leverage turns a 3% move into a 30% return, and the same move against them into a 30% loss, so the highest short-window returns tend to come from the largest bets.
What a record needs to show
A record says more the longer it is, the more steadily it was earned, and the more clearly it beat the obvious alternative. Four things are worth checking before reading a trader's return as skill.
- How long the record is. A week proves almost nothing. Ninety days is the least that says much, and two quarters in a row say a good deal more.
- Whether the lead was steady. A return made over many days is stronger evidence than the same return made on two or three, and the drawdown says how hard the trader fell along the way.
- Whether it beat the market the trader was in. In a quarter when Bitcoin rose 15%, a crypto return of 10% lost to doing nothing.
- Whether it repeats. Luck rarely holds for two windows running. In the same simulations, a trader with no edge reached Master in two windows running 2.9% of the time, against 25.2% for a strong trader.
How Sharpe ranks traders
Sharpe ranks traders on all four. Every record is measured over 90 days against holding Bitcoin, day by day, and discounted by half its own uncertainty, so a short or lucky record counts for less. Beside every rank sits Confidence, the probability that the lead over Bitcoin is real, and a Diamond or Master record that held its tier for two windows is marked as such. Where one trader runs several wallets together, only the strongest of them can take a Challenger seat.
Sharpe reads spot trades and on-chain wallets, and does not read perpetual futures, so it is not a ranking of copy-trading lead traders. It is a way to see how any Solana or Ethereum wallet's record measures, including your own: paste one into Sharpe or browse the leaderboard. How the ranking works is set out in full in the methodology.