Copy Test #4.
Forward wallet-persistence test. The Test 3 age ≥3m all-wallet control is preserved unchanged while wallet cohorts frozen from Test 3 are followed prospectively under the same execution rules.
Does trader-level edge persist into the next sample?
Five arms compare the unchanged all-wallet control with Test 3 positive, repeat-positive, robust-positive and repeat-negative wallet cohorts. The wallet lists are frozen before Test 4; token age, entry delay, exits and execution costs stay fixed.
Can the edge survive execution?
Every closed trade is recomputed under two fixed-cost assumptions. This isolates strategy quality from the choice of transaction tip.
Who are we actually copying?
Holding time, buy size, repeat-buy behavior and live copyability are retained so strategy performance can be interpreted by trader type.
No wallet, private key or live transaction path is used by this lab. The purpose is to discover which copy behavior remains economic after observed delay, adverse slippage, platform fees, Pump fees and tips.