How to Copy Trade on Polymarket
Updated
Copy trading on Polymarket means mirroring another trader's positions in your own wallet, automatically. You pick traders from a catalog, decide how much of your money each copy may use, set the losses you are prepared to absorb, and a bot places the same orders you would have placed by hand — seconds after the trader does, on the same market. This guide covers how to copy trade on Polymarket end to end: what a prediction market is, what happens mechanically when a trade is copied, how to read a trader's numbers without fooling yourself, and which limits to set first. Replix is one way to run it; the mechanics below apply whichever tool you use.
What copy trading on Polymarket actually is
Polymarket is not an exchange in the usual sense, and that changes what copying a trader means.
On a spot or futures venue you buy an asset that has no fixed end state. Bitcoin at $60,000 can go to $90,000 or $30,000; the position lives until you close it. On Polymarket you buy shares in a binary outcome — will this happen, yes or no — and each share pays exactly $1 if the outcome resolves your way and $0 if it doesn't. Nothing in between, ever.
Because the payoff is capped at a dollar, the price is capped too. A share trading at 63¢ is the market saying the event has roughly a 63% chance of happening. Price is probability. That single fact reshapes everything about copying someone:
- There is no leverage and no liquidation. Your maximum loss on a position is what you paid for it.
- Every position has an expiry. Markets resolve. You cannot hold a losing prediction until it recovers the way you can a losing stock — the event either happens or it doesn't.
- Entry price is most of the edge. If a trader buys at 40¢ and you fill at 46¢, you didn't take a 6% worse entry. You took the same bet with a payoff ratio cut by roughly a third, on a maximum move of 60¢. On a leveraged instrument a slightly worse fill is noise. Here it is the difference between a good trade and a mediocre one.
- The skill you're copying is forecasting, not chart-reading. Consistently profitable Polymarket traders are people who price events better than the crowd — who read polling, court calendars, sports injury reports or macro releases faster and more accurately. That kind of edge tends to be category-specific.
So copy trading on Polymarket is closer to following a forecaster than a momentum trader, and two questions decide your result: did I get in near their price, and did I size the position so a wrong forecast doesn't hurt.
How copying works mechanically
This is the part most descriptions skip. Here is what actually happens between a trader clicking buy and the same position appearing in your wallet.
Where the signal comes from
Polymarket trades settle on Polygon and the order book is public. Every fill is visible — market, outcome token, size, price, wallet. The bot watches the wallets you subscribed to through the live exchange feed and on-chain events, so it needs no permission from the trader, and the trader never learns they're being copied.
Where your money sits
Replix creates a dedicated trading wallet for you and holds its signing keys. During the closed beta this is a custodial arrangement, and it is what lets the bot execute in seconds without you confirming every order. It trades only inside the limits you set, and withdrawals go to the address you specify. Treat the wallet the way you would treat an exchange account: fund it with the capital you have allocated to copy trading and keep the rest elsewhere. Collateral on Polymarket is denominated in a dollar token — you deposit USDC and it converts on arrival. A smart wallet where ownership stays with you and Replix holds trading-only permissions is planned for public launch.
How the size of your copy is decided
Two modes cover almost every case:
- Fixed size — every copy is the same dollar amount, whatever the trader risked. Predictable, and the right default at the start.
- Proportional — your copy scales with the trader's position relative to your capital, so a trade they clearly believe in becomes a bigger trade for you too.
Your hard cap sits above both. If the trader's position is larger than your per-trade maximum, the copy is trimmed to your cap, not skipped — same trade, smaller size. Nothing exceeds your limit, and nothing is silently dropped.
Speed, and why it's the real constraint
The bot reacts in seconds, not minutes — which matters because probabilities reprice fast when news breaks, exactly when good traders act. Two protections cover the gap. A copy is placed as an order with a short lifetime rather than an open-ended one, so if the price has already run away it expires instead of filling at a level you never agreed to. And markets with an empty book are filtered out, because a copy that has to cross a 10¢ spread has lost before it started.
Exits, not just entries
Copying only entries is how people lose money with copy trading. When a trader sells out — taking profit, cutting a loss, reacting to news — the exit is mirrored the same way the entry was, in proportion to what you hold.
What it costs
Three separate costs, worth keeping straight:
| Cost | Who charges it | Notes |
|---|---|---|
| Taker fee | Polymarket | Charged on market orders in most categories and scaled by price; resting limit orders are not charged, and some categories stay fee-free. Check the current schedule before you size up — it has changed more than once. |
| Gas | Polygon | Cents. Matters for wallet setup, negligible per trade. |
| Replix fee | Replix | None during the closed beta. The pricing model is being finalized and published before public launch. |
How to choose a trader
The catalog gives you PnL, win rate and volume. None of the three means what it looks like on its own.
Win rate is the most misleading number on the page. A trader who only buys shares at 90¢ can be right 88% of the time and still lose money: they win 10¢ and lose 90¢. A trader who buys long shots at 15¢ can be wrong two times out of three and print. Always read win rate next to average entry price. If the two aren't shown together, treat the win rate as decoration.
PnL needs a denominator. $40,000 of profit is excellent on $50,000 of capital and unremarkable on $2 million. Look for return relative to the capital deployed, and for whether the profit came from many trades or one enormous lucky call on a single election night.
Volume tells you whether you can follow them at all. A trader taking $200,000 positions in deep political markets moves prices you'll be filling into. If their typical size is large relative to the book, your copy arrives after theirs has already moved the price.
Sample size and period. Ninety days is a floor, and more useful is whether the record spans more than one kind of environment — an election cycle and a quiet month, not just the week everything went their way. A brilliant 14-day record is a coin flip you're seeing the good side of.
Signs the record is manufactured. Wash trading and bot patterns are common enough that checking is not paranoia. Watch for a wallet trading against itself or a small cluster of related wallets, round-the-clock activity with no human gaps, hundreds of tiny fills that inflate volume without moving PnL, and near-perfect win rates on illiquid markets nobody else is trading. Replix screens copied wallets for exactly these patterns before you follow them, but the tells are visible on-chain to anyone who looks.
Copying several traders at once. Diversification helps here, because forecasting edges are category-specific — someone excellent on US politics is not automatically good on Champions League markets. Each trader gets their own allocation, your per-trade cap applies to all of them, and your daily loss limit is measured across the whole portfolio rather than per trader. Three to five is a sensible range; twenty traders means you've reconstructed the market average and are paying fees for the privilege.
Risk management: the part that decides your result
Copy trading doesn't remove risk — it helps you manage it. A 65% win rate still means roughly one trade in three loses. What separates a survivable strategy from an expensive one is whether the losses are bounded before they happen.
Three limits do most of the work:
- Maximum per trade. A hard ceiling on any single copy, independent of what the trader risked or what your sizing mode calculated. A common starting point is 1–3% of your copy-trading capital. It's the only limit that protects you from a single catastrophic conviction bet.
- Daily loss limit. When realized losses for the day cross the threshold, copying pauses until tomorrow. This exists to break correlated bad days — a news event that moves every position at once. Somewhere around 5–10% of capital is a reasonable first setting.
- Drawdown stop. If total capital falls by a set percentage from its peak, copying halts entirely and waits for you. Set it at a level where you'd genuinely want to stop and reassess — for most people between 15% and 25%.
Set these before you copy anyone, not after the first bad week. Replix suggests defaults at onboarding, and you can loosen them — deliberately, seeing what you're doing.
Trader degradation. Forecasting edges decay. A trader who read the primaries perfectly may be lost when the calendar moves to macro. Replix monitors every copied trader against their own historical baseline, and when recent performance falls clearly below it, copying auto-pauses and you get an alert with the numbers, rather than discovering the decay in your balance a month later. You then decide: resume, halve the size, or drop them.
Ring-fence the capital. Fund the trading wallet with the amount you have allocated to copy trading, and nothing more. It bounds your exposure to any bug, any bad week and any mistake of your own, and it makes your results legible — you can see exactly what the strategy did without it being tangled up with the rest of your holdings.
Liquidity discipline. Markets with an empty book are filtered out before a copy is placed. Nothing ruins a good signal faster than a fill 8¢ away from the trader's price.
How to start
Four steps, and the first three are the ones that matter.
- Set up your wallet. A dedicated trading wallet is created for you, with signing handled by Replix — a custodial arrangement during the beta — so the bot can execute without a confirmation for every trade. Fund it with USDC on Polygon; it converts to the market's dollar collateral automatically. There is no minimum imposed by Replix, but be realistic: your position sizes need to be large enough that fees and spread don't eat the edge, and your bankroll large enough that a per-trade cap of a few percent still produces meaningful positions.
- Choose traders. Work through the catalog with the previous section in hand: win rate read against entry price, PnL against capital, at least 90 days of history, no wash-trading tells. Start with two or three in categories you can sanity-check yourself.
- Set your limits. Per-trade maximum, daily loss limit, drawdown stop. Accept the suggested defaults if you're unsure — they are deliberately conservative, and you can widen them once you have your own data.
- Let the bot copy. Positions are mirrored in real time and every action arrives as a Telegram notification: what was copied, at what price, and where you stand. Pause any trader, or all of them, at any moment.
Access is currently by application — Replix is in closed beta and the bot isn't public yet. You can apply for access with your Telegram username; beta participants get a 7-day free trial when the product launches publicly. The security model and the full FAQ are on the main site.
FAQ
How much money do I need to start? There's no minimum set by Replix. Practically, you want enough that a 1–3% per-trade cap still buys a position worth the fees — for most people that means starting with a few hundred dollars rather than twenty.
Can I stop copying at any time? Yes. Pause an individual trader, or all copying, instantly from the bot. You can also withdraw your balance to your own address whenever you want, whether copying is running or not.
What happens if a trader I copy goes quiet or stops trading? Nothing bad — no trades means no copies, and your capital sits idle. You'll see the inactivity in their profile and can reallocate to someone else. Traders who stop posting results also stop meeting the catalog's activity criteria.
How is this different from copying trades manually? Speed and consistency. Manually you're watching the leaderboard, spotting a fill, checking the market and placing an order — minutes later, at a repriced level, and only when you happen to be awake. A bot does it in seconds, at every hour, and applies the same size and stop rules to every trade instead of the ones you feel strongly about.
Does Replix guarantee profit? No. No copy-trading tool can. Past performance never guarantees future results, and copying a profitable trader means copying their losing trades too. Replix is built to bound the downside, not to promise upside.
Who holds the keys to the trading wallet? Replix does. Signing is handled on your behalf, which is what makes hands-off execution possible, and the bot trades only within the limits you set. This is custodial by design during the closed beta, so fund the wallet with your copy-trading capital only, the way you would treat any exchange account; withdrawals go to the address you specify. At public launch we're moving to a smart wallet where ownership stays with you and Replix holds trading-only permissions.
Is Replix connected to Polymarket? No. Replix is an independent tool that reads public market data and places orders on your behalf. It is not affiliated with or endorsed by Polymarket.
What if I want to copy a trader Replix doesn't list? Catalog coverage grows during the beta, and beta participants can request wallets. Tell us in the application form — early requests shape what gets built first.