Copy-trading a caller sounds simple: buy what they call, sell when it pumps. Whether it works depends on three numbers, and most people only look at the first.
1. Hit rate is not return
A caller's hit rate tells you how often a call reached a target multiple at some point after the call. It says nothing about whether you captured that peak. In our data the median call peaks around 1.5× but the median call is worth a fraction of its call price a few days later. The gap between those two numbers is where copy-traders lose money.
2. Exit discipline decides the outcome
Model two followers of the same caller. One takes profit at 2× and holds the misses; the other holds everything. The first banks every call that touched 2×; the second rides most of them back down. Run both in the copy backtester and the difference is usually larger than the difference between callers.
3. Costs compound
Every copy pays slippage and fees on the way in and out. On thin memecoin pools 2–5% per side is normal. A strategy that makes 8% per call before costs can be flat after them.
A simple expected-value check
Expected multiple per call ≈ hit rate × target + (1 − hit rate) × average multiple of misses − costs. If that is below 1, no amount of conviction fixes it. The copy-PnL leaderboard computes this for every caller with enough measured calls.
Educational only. #DYOR #NFA