Headline return vs. what you'd actually make

A wallet's headline return assumes you fill at the same prices, at the same instant, paying nothing. You do none of those things. Once you subtract trading fees, the slippage of moving real size, and the latency between the leader's trade and yours, a copier typically keeps far less of the gap than the headline implies — often 20–40% less. Here's where it goes, and how to estimate what's actually left.

The headline number is a best case you'll never get

When a leaderboard shows a wallet up some percentage, that figure is computed on the wallet's own fills — its prices, its timing, its zero-fee accounting of paper PnL. It is the most flattering possible version of the result. A copier sits one step behind every trade and pays three costs the headline ignores.

Cost 1: fees

Polymarket charges a taker fee on trades, and a copier — arriving after the leader, taking whatever liquidity is there — pays it on the way in and the way out. On a strategy that turns over capital frequently, fees compound into a real drag. CopyGrade computes returns net of fees, matched FIFO, rather than counting gross paper PnL, because the gross number is one a copier never actually banks.

Cost 2: slippage

The leader's fill price was available for the leader's size. When you copy — especially into a thin prediction market — your order walks the book. The deeper the market and the smaller your size, the less this bites; in shallow markets with real capital, it bites hard. Slippage is the difference between the price you saw and the price you got, and it always runs against you.

Cost 3: latency

You see the leader's trade after it happens, your bot reacts, and your order lands later still. In that gap the price has often already moved toward where the leader pushed it — so you buy higher and sell lower than they did. We model this as an adverse drift that scales with the delay: the longer your latency, the more of the leader's edge has evaporated before you arrive. Edge that only exists at zero latency isn't edge you can copy.

How big is the gap?

It depends on the wallet — turnover, market depth, and how much of the "edge" was really just being first. A low-turnover wallet trading deep markets loses little. A high-frequency wallet whose edge depends on speed can lose most of it. The point is not a single magic discount; it's that the headline is the ceiling, and your real result is some distance below it. Treat any wallet whose case rests entirely on the headline number with suspicion.

How to estimate the real number

Don't guess — replay it. The Copy Simulator runs the wallet's actual, timestamped fills through a FIFO backtest scaled to your capital, then subtracts modelled fees, slippage, and your chosen copy latency. The output is an honest equity curve: what copying that wallet, under your conditions, would actually have returned — not the leaderboard's best case. You can compare two wallets under one identical set of conditions, which is the comparison the leaderboard can't make for you.

The Copy Score bakes the same discipline into its risk-adjusted-return factor: it scores the post-fee, latency-aware number, so a wallet whose edge evaporates once you account for the real costs of copying scores lower than its headline suggests.

CopyGrade is analysis-only — it never executes trades or holds funds. Not financial advice.