Graveyard / Copy trading / Solana DEX
Copying smart-money memecoin wallets
Does copy-trading the best Solana memecoin wallets work?
The idea
Some wallets win a lot on new tokens. Rank them, copy the best, and ride along a few seconds behind.
How we tested it
| Data | 193 public wallets (public leaderboards). 18,703 copies on the bonding curve and 784 after graduation. 30-second copy delay. Exits at +30%, −18% or 3 hours. |
|---|---|
| Period | 13 to 20 October 2026, split in two halves of 3.5 days. Wallets were picked on half 1 and copied on half 2. |
| Costs | 2.5% per round trip per copy. Also re-run at 1.5 times that. |
| Holdout | Half 2 is the holdout. |
| Control | The whole pool of wallets is the baseline: −7.3% and −7.5% per copy in the two halves. |
Result
| Top 5, half 1 | +5.5% per copy (n 103) |
|---|---|
| Top 5, half 2 | −3.4% per copy (n 116) |
| Rank persistence between halves | +0.07 |
| Entry price drift paid by a 30 s follower | +9.2% median |
| Same idea with a rug filter added | 66 trades, −$360.03 |
Why it failed
Ranking by past copy profit picks wallets that were lucky last week. They fall back to the pool average. The follower also pays a 9% worse entry on the median trade.
What would make it worth retesting
With multi-week wallet histories, or with rug and liquidity checks replayed as they were at the time. Not by ranking on copy profit.
Test id: walletfind, walletbasketclean-sol-event-v1. Past results of a backtest say nothing certain about the future.