How do you choose traders to copy?
You choose a trader by their risk profile, not their headline return. The single most common way copiers lose is picking the top name on a leaderboard after a hot streak, then watching mean reversion take it back on their own capital. The metrics below are ordered so the first ones can disqualify a leader before you even look at the profit figure.
One rule underneath all seven: judge the process, not the P&L. A 300% return tells you nothing about whether it came from skill, luck, or a martingale that hasn't blown up yet. If you want the mechanics of how copying actually executes first, read how crypto copy trading works. This piece is only about vetting the person you follow.
- 1. Maximum drawdown — how much they've lost peak-to-trough.
- 2. Risk-adjusted return — Sharpe or Sortino, not raw return.
- 3. Trade count — is the sample big enough to trust?
- 4. Track-record length — has it survived a real drawdown?
- 5. Return consistency — many small wins, or one lucky moonshot?
- 6. Leverage — how close to liquidation do they run?
- 7. Position concentration — one giant bet, or a real book?
Metric 1 — Maximum drawdown, the number to check first
Maximum drawdown is the largest peak-to-trough drop in a trader's equity, measured as a percentage. Check it before anything else, because it tells you the worst pain you'd have inherited if you'd been copying at the wrong moment — and drawdowns transfer straight onto your capital.
The recovery math is brutal and non-linear. A 20% drawdown needs a 25% gain to get back to even. A 50% drawdown needs 100%. An 80% drawdown needs 400%. So a leader who has touched a 60%+ drawdown isn't just volatile; they've shown they can put you in a hole that takes years to climb out of. For the full framing on where the line sits, see how deep a drawdown is too deep.
Practical thresholds: under 20% max drawdown is disciplined, 20–35% is normal for an active crypto strategy, and anything past ~50% needs a very good explanation. Beware leaders showing 0% drawdown — that usually means a short track record or hidden averaging-down, not genius.
Metric 2 — Risk-adjusted return, not raw return
A raw return figure is meaningless without knowing how much risk bought it. Two metrics fix that. The Sharpe ratio divides excess return by the standard deviation of returns — reward per unit of total volatility. The Sortino ratio does the same but only counts downside volatility, which fits trading better because nobody complains about upside swings.
As a rough read: an annualized Sharpe near 1 is respectable, above 2 is strong, and much above 3 for a retail crypto trader is a flag to distrust, not celebrate — it usually signals a short sample or a strategy selling hidden tail risk. Most leaderboards don't publish Sharpe at all. If you can only see raw return and drawdown, a crude stand-in is return divided by max drawdown; you want that ratio comfortably above 1.
Metrics 3 and 4 — Sample size and track-record length
These two decide whether the numbers above mean anything. A stat computed on 15 trades is noise. A coin-flipper can post a 70% win rate over 20 trades roughly one time in fifty — put fifty lucky flippers on a leaderboard and the top of it is pure survivorship.
Trade count (sample size). You want enough round-trips to separate skill from luck. A few hundred trades is a reasonable floor for an active strategy; a swing trader with 40 trades over a year simply hasn't shown you enough. More trades also make the drawdown figure trustworthy — one you actually lived through, not one you got lucky enough to avoid.
Track-record length. Three months inside a bull run tells you a leader can be long in an up market. That's it. You want a record that spans at least one meaningful drawdown — ideally 6–12 months covering both a rally and a real drop. A track record that has never faced a down move is an untested one — the same trap as an overfit backtest that looks perfect until live conditions change.
Metric 5 — Return consistency
Consistency asks: did the equity curve grow from many modest wins, or from one lucky trade that flatters every other number? Pull the monthly return breakdown if the platform shows it. A steady curve of +2% to +6% months with the occasional small loss is worth far more than a flat line with a single +180% spike doing all the work.
This is also where you catch the strategy that's about to kill you. A suspiciously smooth curve with tiny, relentless gains and no losing days is the signature of a martingale — averaging down into losers with no stop. It produces a beautiful line right up until one trade takes the whole account to zero. If a leader never seems to have a red day, that's not skill; that's an unrealized loss they're refusing to book. Style drift matters too: a trader who was steady for six months and suddenly tripled their size is telling you the old track record no longer describes them.
Metrics 6 and 7 — Leverage and concentration
These are the two fastest ways a copied account goes to zero, and both are things you inherit whether you like them or not.
Leverage. When you copy a trader, you copy their leverage. A leader running 10x can lose your position to a liquidation on roughly a 10% adverse move before fees; 20x cuts that to about 5%. Check the maximum leverage they've used, not the average — one 25x swing-for-the-fences trade is enough to end it. There's nothing wrong with modest leverage used deliberately, but a leader who routinely runs 15–20x is showing you how they'll eventually blow up. The full trade-off is in how much leverage is too much.
Position concentration. Look at how much equity sits in a single position or a single asset. A leader who regularly puts 60%+ of the book into one BTC long isn't running a strategy, they're running a conviction bet you'll be along for. Diversified sizing — no single position dominating, uncorrelated names — is what lets a record survive being wrong. Concentration is also why you should never copy just one leader; spreading across several uncorrelated traders is the point of building a copy-trading portfolio.
The 7-metric scorecard, and how to use it
Run every candidate through the same table before you allocate. If a leader trips two or more red flags, skip them — there's always another leader, but there isn't always another account.
| Metric | Green flag | Red flag |
|---|---|---|
| Max drawdown | Under ~25%, explained | Over ~50%, or a suspicious 0% |
| Risk-adjusted return | Sharpe ~1-2, or return > max drawdown | Sharpe over ~3 on a short sample |
| Trade count | Hundreds of round-trips | Dozens or fewer |
| Track-record length | 6-12 months, includes a drawdown | Under 3 months, bull run only |
| Consistency | Many small wins, some small losses | One spike, or zero losing days |
| Leverage | Modest, deliberate | Routinely 15-20x+ |
| Concentration | Diversified, no single dominant bet | 60%+ of equity in one position |
Three habits make the scorecard actually work. First, weight the risk metrics — drawdown, leverage, concentration — above the return metrics; a leader you can survive beats a leader who looks brilliant. Second, remember your net result will lag their advertised figure once slippage, taker fees (often 2–5.5 bps a side), and funding are counted, so build in margin. Third, size any single leader small enough that their full blowup is survivable, then diversify across a handful. Even with perfect vetting, most copy traders still underperform the underlying asset after costs — going in with that expectation is what separates the survivors from the leaderboard chasers.
Frequently asked questions
What is the most important metric when choosing a trader to copy?
Maximum drawdown. It shows the worst peak-to-trough loss you would have inherited, and unlike raw return it's hard to fake with a lucky streak. Because recovering from a deep drawdown is non-linear — a 50% loss needs a 100% gain to break even — a shallow, well-explained drawdown protects your capital more than any high return figure.
How long should a trader's track record be before I copy them?
Prefer at least 6 to 12 months, and insist it spans a real drawdown, not just a bull run. Three months of gains in a rising market only proves someone can be long when everything goes up. A record that has survived a down move tells you far more about how the trader handles being wrong, which is what actually costs you money.
Is a higher return always better when picking a copy trading leader?
No. Raw return is meaningless without the risk that produced it. A 300% return built on 20x leverage and one concentrated bet is far more dangerous than a 40% return with a 15% drawdown. Judge return against drawdown and, where available, the Sharpe or Sortino ratio, and treat suspiciously high numbers as a warning rather than a win.
How many trades should a leader have before I trust their stats?
Ideally hundreds of round-trips. A win rate or drawdown computed on a dozen or two trades is statistical noise — a pure coin-flipper can post a 70% win rate over 20 trades often enough to top a leaderboard. A large trade count makes every other metric, especially the drawdown, something the trader actually lived through rather than got lucky enough to dodge.
How much leverage is safe to copy?
You inherit the leader's leverage, so the safe answer is modest and deliberate. At 10x, roughly a 10% adverse move liquidates the position before fees; at 20x it's about 5%. A leader who routinely runs 15–20x or higher is showing you their eventual blowup. Check the maximum leverage they've used, not the average, since a single oversized trade is enough to end the account.