# DC2.3 Lane D: coins whose LP the creator does NOT hold (DC Analyst 2)

## READ THIS FIRST
- 7 Oct 11:04 Bangkok, v1. Pre-registered as "DC2.3 Lane D hold book" (dc-prereg.md, 10:50). Chain-only clock: crossing = the first $100+ sell at cap $250k or more; block times from real timestamps. v4 coins launched 21 Sep to 3 Oct. Paper only.
- **Lane D is tiny and loses on a hold.** At the true $250k crossing:
  - 2,476 coins cross; the creator holds the LP on 2,383 (96%), and **99% of those are pulled within 24 h**.
  - Only 93 are not creator-held (about 7 a day): 80 contract, 10 burned, 2 plain wallet.
  - After the skip list and one trade per ticker, creator and root per day: 37 trades (A) and 26 (B), with **4 hits at 3x in each half (bar: 10). Too small to call an edge.**
- **Exits, entry 5 min after the crossing, $100 ticket, $ a day A | B:**

| Exit | $ a day | wins |
|---|---|---|
| TR50 (trail 50% from the top after 1.5x), best on A | -153 \| +148 | 24% \| 27% |
| X3 | -217 \| -71 | 16% \| 15% |
| X5 | -259 \| -47 | 11% \| 12% |
| X10 | -403 \| -126 | 5% \| 4% |
| Hold 48 h | -403 \| -309 | 5% \| 0% |

  - 7-day hold (half A only, B not covered yet): every exit negative (TR50 -$190 a day).
  - Entry at 30 s instead of 5 min: same picture (TR50 -$149 \| +$124).
- **Verdict:** kill bar part 1 not met (TR50 is positive in B), but A is negative and there are only 4 hits a half. So: no edge, too small. Features inside it (operator bundle share, outside net buy, root record, locker vs other): no cut chosen on A holds in B with enough trades. Best looking is root pull rate ≤ 23% (B +$266, beats random 99%), but on 4 trades. Not usable.
- **First principles:** a locked pool removes the pull, not the dump. With the exit secured, the operator sells his float into the pool instead. A 48 h hold wins 0 to 5% of the time. Lever 1 (no pull) is necessary but far from enough. Lever 3 (who holds the float) decides, and that needs Analyst 1's float data.
- **(a) Who holds the non-creator LP:**
  - lockers `0x267444d099b10fb5ed7c3cc7b7c767adca574952` (1,969 bytes, 26 coins, 0 pulled) and `0xeff166aaf189323c58dc27ed1206eb2c37faacdf` (7,260 bytes, 21 coins, 0 pulled). The Second Window lists both as lockers, with `0x35c5c1037eef9eb07b2826d537e4fdae61878852`.
  - **`0xd0fd42bd8374ac659d3d81242d032211c2481055` (22,475 bytes) pulled all 10 of its coins within 24 h. It is NOT a lock; treat it as creator-held.**
  - burned (dead address): 10 coins, 0 pulled.
  - The long.xyz hook `0x4e3468951d49f2eea976ed0d6e75ffcb44a9a544` adds LP for its coins (25,533 bytes). Its family "25533_3080c1dd" is on our junk skip list, so the plan should check that skip.
  - Names: Blockscout blocks API and browser reads with a bot check, so the names are still unknown.
- **Correction to Q2:** on the true crossing, contract-held coins are pulled 28% of the time, not 0 to 11%. All of that comes from `0xd0fd42bd8374ac659d3d81242d032211c2481055`. Without it: 0%.
- Per-trade table: http://159.223.88.168/hx-c9ois/dc/a2/dc-a2-4-laned.csv.gz
- Cost: 0 X credits, $0, about 80 RPC calls (bytecode), research box read about 1 min, about 0.8 h.

## Method
- Universe: every v4 coin in the launch file that has wash.db swaps and reaches a sellable $250k by chain swaps (cap = price x launch `mc_first` ÷ first swap price, matching launch `pk` on 98% of coins).
- LP class at the crossing block from liqlock `liq` + `nft`, the Q2 method (class order BURN, CREATOR, CONTRACT, EOA). Bytecode from eth_getCode.
- Exits walk the path in order. Pull = 0. Sellable targets need a $100+ sell; the trail exits at the first trade 50% under the top.
- Features use only trades before the entry. Operator wallets = buyers first funded by the creator, its funder root (crew.db), or the two disperse contracts.
- Not done: public router check (W2, DCPP.3), other venues (Pons graduated pools are the obvious next non-pullable set; the Second Window already covers them).

Version: v1, 7 Oct 2026 7 Oct 11:04 Bangkok, DC Analyst 2. Latest: http://159.223.88.168/hx-c9ois/dc/dc-a2-4-laned.md