How to build a Polymarket copy-trading portfolio
To build a Polymarket copy-trading portfolio, copy a small set of vetted, low-overlap wallets instead of betting everything on one — then judge the group by its blended drawdown and how much its members trade the same markets, not by its best member. Diversification only works if the wallets aren't making the same bets. A basket of two wallets that both pile into the same election is one bet held in two accounts.
Why copy more than one wallet?
Because a single wallet is a single point of failure, even a good one. A vetted edge can decay, a wallet can drift into markets it has no edge in, and a farming-risk finding can surface on a previously clean record. Spreading across a few independent edges means one wallet going cold doesn't sink your whole copy allocation.
The catch is supply. Copyable wallets are rare — in our June 2026 snapshot only 2.2% of active wallets cleared every vetting test (the numbers). So a real portfolio is a handful of carefully chosen wallets, not a long list. Padding a basket with mediocre wallets to feel diversified just dilutes it toward the field, where the median wallet has negative post-fee edge. Quality of selection beats quantity, every time.
The overlap trap: diversification that isn't
Two wallets that trade the same markets are not diversification — they are the same exposure twice. If both your wallets are long the same election outcome or the same BTC price market, a single resolution moves them together, and the "two-wallet" portfolio behaves like one oversized position. This is the most common mistake in copy-trading portfolios: stacking wallets that look different but bet the same.
So check overlap before you add a wallet, not after. The Baskets tool surfaces a market-overlap percentage for the group — how many markets two or more members are both in — and a high number means your diversification is an illusion. When two candidates look similar, put them head-to-head with Compare to see whether they're genuinely different edges or near-duplicates.
Judge the basket by its blended drawdown
A basket's drawdown is not the average of its members' drawdowns — correlated wallets draw down together, so the combined worst-case can be nearly as deep as the worst single member. Averaging the headline numbers hides this; you have to compute the drawdown of the merged record. Low-overlap members, by contrast, rarely hit their worst stretches at the same time, so the blended drawdown lands well inside the worst member's:
That gap — −36% versus −22% — is the entire point of a portfolio, and it comes from low correlation, not from the number of wallets. Baskets grades a set as a unit: a blended Copy Score, the combined drawdown from the members' merged trade history, and the overlap percentage. Size from that combined drawdown the same way you'd size a single wallet — from the worst loss you can sit through, not the return you hope for.
How to build one, step by step
Build selection-first, then construction. The order matters: a perfectly balanced basket of mediocre wallets is still mediocre.
- Shortlist vetted wallets. Start only from wallets that clear the full vetting checklist — real post-fee edge, survivable drawdowns, farming-clean. This is the rare 2.2%, so expect a short list.
- Cut the overlap. Drop near-duplicates; for any two candidates that look alike, compare them side by side and keep the one with the cleaner, more distinct edge.
- Size each from its drawdown. Cap every wallet so its worst historical stretch stays inside your tolerance, and never let one wallet be the whole allocation — see position sizing.
- Grade the group as a unit. Assemble the survivors in a basket and read the blended drawdown and overlap %, not the headline score. The tool holds up to eight wallets — most copiers need far fewer.
- Monitor and rebalance. Alerts fire when a member starts farming or its score decays; act on the stop rules you wrote down before any money moved, not on the feeling of the day.
A portfolio controls how you lose when a wallet goes wrong — it can't turn negative-edge wallets into a positive-edge basket. Selection first, construction second; the basket is only ever as good as the wallets you were disciplined enough to leave out.
CopyGrade is analysis-only — it never executes trades or holds funds, and a Copy Score is a documented research opinion, not a statement of fact about any trader or financial advice.