Monday, 28 September 2026 · XAUUSD
Day 2 · Gold, Asia, London and New York (from Thailand)
+84.6 pips+£6.262/3 winsBalance £320.20
Tap to enlarge
Tap to enlargeMarket context
Gold gapped down at the Monday open from Friday's 4,286 close and drifted to Friday's low (4,254.7) in Asia. The first automatic markup only planned range breakouts; the agents' review rejected both, and I pushed to look at the levels outside the range. Gold kept falling into London (about $120 down on the day) as the US–Iran Hormuz talks stalled, oil hit $107 and traders priced in another Fed hike.
Key levels
The plan
- Reviewed Asia update: long on a failed breakdown back above Friday's low 4,254.7; short on a real break below the week low 4,244.4; short on a rejection at 4,265–4,268.
- After the $20 crash through the week low (no retest), don't chase. Wait for a pullback into the broken level and a 1m lower high.
- Below 4,229 there was no recent history, so the targets came from June–August levels: 4,195–4,203, then 4,166–4,170.
- London (reviewed levels): sell a clean 5m close below 4,160–4,166, retest failing with a 1m lower high. Stop 4,167, target 4,125 in front of the strong 4,116–4,122 zone.
What happened
At 07:31 Thai a $20 5m candle crashed through the week low to 4,229, so I skipped it (lesson 3). A failed-breakdown long at today's low triggered on the 07:55 close, but the 1m never made a higher low, so I stayed out, and it broke lower. Price then pulled back into the broken lows at 4,223–4,226: the 08:15 5m candle rejected 4,225.7 and the 1m made lower highs, so I sold at 4,221.03. I trailed the stop down behind the 5m swings into profit (4,224.6, 4,218.9, then 4,214.4). Price made 4,206 and pulled back to retest 4,214.4, which took the trailing stop at 08:53 for +66 pips. London: the first 5m close under 4,160 was only 4,159.84, a touch, so I waited. The 08:25 UK candle closed 4,158.26 (clean, normal size), the retest made 1m lower highs, and I sold twice (4,158.86 and 4,159.73). I closed the first at 4,157.11 for a small profit in case it reversed. At +50 pips on the second I moved the stop to breakeven (4,159.66); a normal $4–5 pullback tagged it to the cent, then gold carried on down past 4,153. New York: no trade. Gold whipsawed about $30 each way around the open (4,171 → 4,136 → 4,160 → 4,127), and every setup either failed a rule or ran away without a pullback.
Trades
Rejection at the broken lows (4,223–4,226), 1m lower high. Trailed the stop into profit; stopped by the retest wick at 4,214.4. +£5.01 before £0.04 commission.
Break and retest of 4,160–4,166 (July weekly high), 1m lower high. Layered entry 1 of 2. Closed early to protect profit in the chop. +£1.32 before £0.04 commission.
Break and retest of 4,160–4,166 (July weekly high), 1m lower high. Layered entry 2 of 2. Reached +50 pips, stop moved to breakeven, tagged by a $4–5 pullback before gold fell further. +£0.05 before £0.04 commission.
Trades I skipped (and why)
- 13:05–13:25 UK: Sell at the 4,158.5–4,166 zone after a $19 spike into it. No closed 5m back below 4,157.5; the stop would have been about $12. Price then faked out above the zone to 4,171.
- 13:25 UK: Sell the failed breakout back into the zone. The signal candle was a $10.5 dump and price had already fallen $12: chasing. No pullback came.
- 14:30 UK: Sell the 5m close below 4,138.5 with a 1m lower high. New York open window (no trades 14:30–14:45) and an $8.47 candle. It whipsawed $22 straight back up.
- 15:00 UK: Sell the rejection at 4,158.5–4,166. The 5m never closed cleanly below 4,156, then dropped $10 with no pullback; not enough room left to the next level.
- 15:35 UK: Sell the clean break of the 4,136.5–4,141 shelf. Valid close (4,131.92), but no 1m lower high on a pullback; price ran into the 4,116–4,122 support instead.
Lessons
- The best trades were at the key levels outside the morning range, not the range breakouts (lesson 9).
- The 1m higher-low filter kept me out of a losing long at today's low.
- Targets must be real levels. When price breaks into fresh ground, look further back (June–August) instead of using 2x the risk (lesson 10).
- My final trailing stop sat exactly on the level price was retesting (4,214.4), so a normal retest wick took it. Trail beyond the swing, not on the level (lesson 8).
- A close a few cents past a level is a touch, not a break. Waiting for a clean close (about $1.50 through) kept me out of the first fake break.
- Breakeven exactly at the entry gets tagged on a normal pullback. On a strong trend day, move the stop to breakeven only once price is past the next swing, or put it beyond the last 1m lower high instead.
- Big candles are relative: measure wick to wick and compare with the last 14 five-minute candles. Skip anything over about 1.5× that (about $8 in Asia/London, $10–11 in a busy New York session).
- The New York open rule paid for itself: the 14:30 break looked perfect and reversed $22 inside the no-trade window.