Is Polymarket copy trading profitable?
Update, July 2026: the third edition of our data report makes the answer more negative — across 1,649 wallets the median fee-adjusted edge fell to −2.7% and only 22 (1.3%) pass every test. The June figures below stand as the dated snapshot they were.
For most wallet choices, no. In our June 2026 snapshot of 1,297 actively-traded Polymarket wallets, the median wallet's fee-adjusted edge was −0.8% — negative before a copier adds slippage, latency, or bot fees — and only 28 wallets (2.2%) passed every test a copier should apply. Copy trading on Polymarket can be profitable, but only as a selection problem solved well: the base rate of good choices is about one in fifty, and the average choice loses money by construction.
Why does the average copy lose money?
Three stacked reasons, each documented in our data:
- The median wallet has no edge to copy. 70% of active wallets have a negative realistic post-fee edge; the median sits at −0.8%. Copying transmits the wallet's result to you — including the negative ones.
- The copier keeps less than the leader made. Fees, slippage, and copy latency mean a copier typically banks 20–40% less than the headline return, and bot fees of around 1% per side come out of what's left. A thin edge becomes no edge in transit.
- The records that attract copiers are the least trustworthy. 53% of active wallets carry a farming flag — and on the leaderboard, where copiers shop, it's 72%. The best-advertised records include wallets built specifically to dump on their followers.
So who actually makes money copying?
The copiers who treat it as due diligence with an execution step, not passive income. The profitable path runs through four gates, in order:
| Gate | What it filters |
|---|---|
| Vet the wallet | 97.8% of active wallets fail at least one test |
| Simulate the copy | wallets whose edge dies in transit at your latency and size |
| Choose the bot deliberately | fee structures that eat more than the edge |
| Size from the drawdown | ruin during the losing stretch that will come |
Skip any gate and the base rates above are your expected outcome. Run them all and you're choosing from the 28 — with realistic costs priced in and a stop rule written down.
The honest caveats
Two, and they're load-bearing. First: the 2.2% were worth copying as of the June 2026 snapshot. Edges decay, behaviour changes, farming findings surface on previously clean wallets — which is why alerts on score decay matter as much as the initial pick. Second: the Copy Score is a documented opinion built from public trade history, and it's young — we publish every model change and validate it against outcomes in public. A high score is a vetting summary, not a prophecy; nothing here is a promised return.
"Is it profitable?" has the same answer as most trading questions: not on average, occasionally in the tail, and the work is in finding the tail honestly. The numbers behind every figure above are in the full data report, and the complete walkthrough — shortlist, vet, bot, size, monitor — is in how to copy trade on Polymarket.
CopyGrade is independent and analysis-only — it never executes trades or holds funds. Not financial advice.