# A3.13 The operator's exit (Analyst 3)

## Read this first (07 Oct 11:50 Bangkok)
- **Verdict: KILL (clean).** Selling when the coin's own bundle wallets have sold X% loses to the live 50% trail in every book and half except one tiny cell. The chosen X (10%, best on half A) gives: live book A +$156 a day, B -$4,098; kit300k A -$620, B -$5,498. It beats random exit times 35.5% / 0% (live) and 0.5% / 0% (kit300k), against a 95% bar.
- **Why (mechanism):** the bundle has usually ALREADY sold by the time we hit 5x. In 26 of 41 coin trades with a named bundle, the bundle's sold share at our 5x was at 10% or more, often 60% to 100%. So the rule mostly turns into "sell at 5x", and that cuts the biggest runners: 0x8e41fd25e85e624560fa50ff9f117a02cd30b8b3 29.15x to 4.48x, 0x9ba4d84ac1571a0788b2bb7a6412e73529d12dc6 28.00x to 4.53x, 0xc38c332012a9116dcade0c7f43b5d3c71799a9f8 16.17x to 4.68x, 0xee049979960d5e42b389b941ec45790ca7b6e1f7 16.18x to 4.74x. The operator's bundle sells into the first leg. The second leg is paid for by outside buyers, not by the bundle, so the bundle's selling says nothing about the top.
- It does save money on the coins that fade after 5x (for example 0x0f3e825cdafbacdc552899a1185e5998f2c2ec9c 2.86x to 4.78x, 0x0d974113699e786f4fc56d760ec7f5270b147a88 2.67x to 4.75x), but those gains are small next to the runners it cuts.
- Why this is different from the killed drip-seller, net-buy and pull-tell exits: those read anonymous flow. This one keys on the operator's OWN wallets, named before the run from how they were funded. It still fails, for the reason above.

## Results ($1k tickets, $ a day on the kit-hold coins only; the rest of each book is unchanged)
| X (bundle sold) | live book A: trades, op exits, live, op, change | live book B | kit300k A | kit300k B |
|---|---|---|---|---|
| 10% (chosen) | 7, 3, $1,787, $1,943, +$156 | 8, 5, $9,625, $5,527, -$4,098 | 24, 7, $5,157, $4,537, -$620 | 24, 12, $21,792, $16,294, -$5,498 |
| 25% | 7, 2, +$42 | 8, 5, -$4,098 | 24, 6, -$737 | 24, 12, -$5,351 |
| 50% | 7, 2, +$42 | 8, 5, -$4,098 | 24, 5, -$564 | 24, 12, -$5,625 |
- Beats random (200 draws, same coins, exit at a uniform time between the 5x and the live exit), X = 10%: live A 35.5%, live B 0%, kit300k A 0.5%, kit300k B 0%.

## Pre-registered rule (published 11:00 Bangkok, before any result)
- Coins: every kit coin that hit 5x under KIT HOLD in my live wide-gate book (15) and in the A3.12 kit300k trades (48). Same engine for both: tick-walk fills at block end.
- Bundle: the coin's buyers while its cap was under $300k (router buys resolved to the signer). Each was traced with an archive ETH balance search (paid dRPC on research-box; key read at use, never printed): fresh = zero ETH 3 days before the buy; funder = sender of the first funding tx (direct, or the signer of a disperse call). Bundle = every pre-$300k buyer funded by the coin's top funder. Coins with fewer than 3 bundle wallets keep the live trail.
- Sold share at time t = tokens sent out of bundle wallets to anyone outside the bundle, divided by tokens received by the bundle so far (from the token's Transfer logs).
- Exit: after the 5x, sell all at the first moment the sold share reaches X (at the 5x if it is already past X), else the live 50% trail / pull / 48 h.

## Tracing numbers
- 15,768 pre-$300k buyers across 50 pools. 7,257 had 20 txs or fewer; 7,189 were fresh (zero ETH 3 days before); the funding tx was found for 7,113.
- 41 of 63 coins have a bundle of 3+ wallets (median about 100 wallets; the largest is 888). 22 coins have no usable bundle (0 to 2 wallets from the top funder), mostly half B Kit B coins whose early buyers were funded by many different senders.
- Data gap: a few bundles show a sold share above 1 (up to 21x). Those wallets received the coin by a path the Transfer filter did not count as "in" (for example, minted to them directly). Those coins trigger at the 5x either way, so the verdict does not change.

Analyst 3, 07 Oct 2026 11:50 Bangkok. Latest: http://159.223.88.168/hx-c9ois/sw/sw-a3-13-operator-exit.md