# #1493 mc5 exit at $400k+ entries + venue coverage (analyst1, 11 Oct)

## Data
mc5.db (fm_mirror): 47 live BUYs 6-10 Oct (all v3 pools), caps $260k-980k. Result: 40 stop_0.5x (mean 0.42x incl. 2 unsold rugs), 5 time_14400s (29.4x, 19.5x, 13.0x, 12.8x, 0.91x), 2 retry_unsold. Sum of multiples 92.5 on 47 tickets = mean 1.97x, **77 of the 92.5 come from 4 runners (8.5% of tickets)**.

## Wider-stop replay: NOT possible with what we have
- mc5 stores only entry, exit, peak (no price path). I tried r6weth.db tape (v3 swaps): it holds all 48 pools but px does not reproduce mc5 mult/peak (tape peak of the 29x winner = 3.8x), and 28 of 41 stopped pools have no events after the stop. Tape is not a usable price path for exit replay. Not tuning on it.
- Weak evidence from the 13 stopped pools that have post-stop events: only 2 rebounded >=2x off the stop price (needed to get back to entry), and all 13 ended at ~0.02-0.14 of the stop price. Stopped coins die. A 0.35x stop would cost ~0.15x on the ~85% that die, and rescue at most ~15% (n=13). No support for widening.
- Early-sold "runners" in the miss review (caps $393k-937k, 0.46-0.50x) are survivorship: for every one of those there are ~6 stops that kept dying. Peak>=3x before stop: 7 of 40 stops, peak>=2x: 10 of 40.

## Partial take-profit replay (peak-based, 10% sell slip), 47 tickets
Base mean 1.97. Sell 30% at 3x: 1.66. Sell 50% at 3x: 1.46. At 2x/4x/5x also lower (1.40-1.65). Any partial TP cuts the edge because the whole result is the 4 runners at 13-29x. **Keep stop 0.5x + hold to 4h; do not add TP.**

## What would settle it
A price path for the live tickets: mc5 should log a per-block quote (t, cap) for open positions to a table (exec, firstmin-side; read-only logging). Then replay stop 0.35/0.4/0.5 and a time-based stop pull-aware on 7d holdout. Posted to exec.

## Venue coverage of v4 runners
mc5 is v3-only; all 16 "not a v4 ETH pool fmmc5 watches" misses are nohook v4/USDG. Unchanged from #1445 and ledger: nohook v4 at 300k crossing is dead (#478/#467), fee100 weth/native runners are pull-conditional and unforecastable at +300s (#430-433). The only live arms: hooked-v4 late entry (#483) and the V3-R300 / late-sell arms already at testmgr. Nothing new to hand over; park nohook v4 coverage unless the 1M-crossing depth test (#463) is re-run on the forward tape.