Running · results December 2026
The November study
Do the wallets at the top of a public profit leaderboard still make money for someone copying them in the following month, after delay, fees and slippage?
The pilot replayed 3 wallets. This study screened 30, chosen the way a follower would choose them, and follows the 12 that could be studied fairly. Every rule was written down before their results existed.
How the wallets were chosen
The top 30 wallets on a public profit leaderboard (its monthly view) were recorded by hand on 3 November 2026, in the leaderboard’s own order. No names or handles were recorded.
Before anything is measured, three checks on each wallet’s October trades decide whether it can be studied fairly:
- Bot-like activity: more than 5,000 transactions on a single day. No person could copy that.
- Too few trades: fewer than 50 positions bought and sold.
- Too much missing data: more than 20 % of swaps that can’t be priced.
Wallets that fail a check are left out and listed with the reason. Nobody is swapped in.
30 screened, 12 studied
- 12passed every check and are being followed
- 12could not be analysed: other accounts send them so many automated transactions, up to thousands an hour, that their history costs more to download than the study’s budget allows
- 3had too many trades that can’t be priced
- 2made too few trades
- 1signed more than 5,000 transactions on a single day
The second group includes most of the highest-ranked wallets. So the results will describe the 12 wallets that could be studied, not the very top of the leaderboard, and the write-up says so.
What is measured
Every trade the wallets make from 4 November to 4 December 2026 is replayed for a follower with a 100 EUR budget. The main result assumes an instant copy through a copy-trading bot, with its 1 % fee and normal slippage. Copying by hand 10, 30 and 60 seconds late is measured as well.
Timeline
- 24 OctoberStudy rules fixed and fingerprinted (SHA-256), before any wallet was chosen
- 3–4 NovemberLeaderboard recorded; 30 wallets screened on their October trades; the 12 selected were fixed and timestamped
- 4 Nov – 4 DecThe month being replayed
- December 2026Results published here
Changes are shown, not hidden
Four changes to the plan were made before any results were seen, and each is recorded with its date and reason:
- The leaderboard was recorded on 3 November instead of 1 November.
- The top 30 were recorded instead of the top 50.
- The replay starts on 4 November, after the leaderboard was recorded, so no wallet is chosen partly on the days it is then judged on.
- The bot check counts only transactions a wallet signs itself. Transactions other accounts send that merely mention a well-known wallet are not its own activity, and a wallet whose history is too large to download within budget is listed separately instead of being called a bot.
What will be published
Results for each wallet under a label such as W01, which wallets were left out and why, and how much of each wallet’s trading could be priced. Never wallet addresses, names or a ranking of who to copy.
This is research on past on-chain data. It is not investment advice and does not recommend buying, selling or copying anything.