34 trades closed, 26 wins, 8 losses — 76.5% win rate, one of the best of the week by that measure. Net P&L still came in at −$87.94. BTCUSDm went a perfect 5-for-5 (+$112.08) and ETHUSDm went a perfect 4-for-4 (+$28.21). Both good days on their own. GBPUSDm erased more than that by itself.
GBPUSDm closed 4 wins and 3 losses. The four wins were profit-lock exits worth $6.30, $9.10, $6.30, and $7.00 — $28.70 total. The three losses were full stop-losses at −$43.05, −$45.50, and −$44.80 — −$133.35 total. Net for the symbol: −$104.65, on a 57% win rate. This is the same shape as GBPUSDm's opposite-direction good day two days ago (Sep 16, +$62.65, 6-for-6) — the mechanics don't change, only whether the stops get touched.
XAUUSDm told a similar story at smaller scale: 11 wins to 4 losses (73.3%) but net −$81.64, because most of the wins were locks in the $4–$15 range while the losses included a −$61.84 stop that alone outweighed six or seven of them. The day's single biggest win, by contrast, shows what happens when a trade is given room: a BTCUSDm long ran to a genuine $103.42 peak before locking 70% of it for +$70.00 — the largest locked win logged yet, and a preview of what the trigger scaling fix (four days out at this point) was built to make more common.
The XAUUSDm stop that did the most damage held only 20 minutes: entered at 4391.035, peaked at essentially breakeven (+$0.06), and reversed to a full −$61.84 stop — no room to lock anything before it turned.