XAUUSD signal: buy the $4,080 dip, target $4,138

Gold stalled at $4,146.92 after a five-session $277 rally and closed mid-range on an $85 outside day. The desk buys the $4,080 retest toward $4,138 and is flat before Friday's payrolls print.

Levels in this article
  • $4,111.24
  • $4,060.02
  • $4,061.40
  • $4,112.00
  • $4,110
  • $4,080
  • $4,022
  • $4,146.92

Trade plan at a glance

Entry
$4,080 buy limit
Stop loss
$4,022
Take profit 1
$4,138 (close half)
Take profit 2
$4,196
Risk-reward
1:1 to TP1 and 1:2 to TP2 — $58 at risk, $58 and $116 paid
Do not take it if
spot trades under $4,061 before the order fills, or the order is still unfilled at 13:15 UTC

Primary levels from this article — any conditional alternative is laid out in the body below. Sizing comes from the stop distance; confirm on your own feed before acting, spreads and gaps are real.

Gold opened Thursday by doing in three hours what took it all of Wednesday, then spent the rest of the session handing it back. Spot was $4,111.24 at the 20:00 UTC cut, +1.26% on Wednesday's $4,060.02 close, after a $4,061.40–4,146.92 session that finished near the middle of an $85.52 range (1.08 × ATR(14)). A second feed had spot at $4,112.00 at 20:02 UTC (Investing daily, gold-api), so the band at the cut is $4,110–4,113 — the session is still open.

The plan buys the $4,080 pullback, risks $58 to $4,022, and is flat before Friday's payrolls.

The rally stalled at October's shelves

The high was $4,146.92 in the 03:00 UTC bar, an hour after the 01:00 bar had already run $4,084.14 to $4,135.19. From there the tape went sideways for ten hours and then sold into the London open: the 14:00 bar bottomed at $4,092.78, the 15:00 bar at $4,076.52. The 16:00 bar reversed $43.88 and closed at $4,127.43, and the four hours since have printed lows at $4,083.97, $4,101.65, $4,095.69 and $4,098.35 — three of the four at or above $4,095.

Zoom out and the shape is a five-session vertical move that has arrived at the place it broke from. Gold bottomed at $3,870.17 on 30 October and has since added $276.75 (+7.15%) in five sessions, two of which were expansions — $151.01 on 4 November (+3.35%) and $85.52 today. Above today's high sit $4,177.59 (15 October high), $4,190.57 (19 October) and $4,198.84 (16 October) — the shelf gold fell out of on 22 October in a $124 range. So the first genuine ceiling is a cluster at $4,177–4,199, and today's high stopped $30 short of it.

That cluster is where this plan's second target comes from. $4,196 is $2.84 under the 16 October high, not an extension pulled out of thin air.

The dollar is doing the work. The Fed is not helping.

The dollar index ran 101.25–101.65 through the session after Wednesday's 101.49 close, a 0.37% fall that left it at its lowest since 30 September (Euronext, OCBC). Yields went with it: the 10-year traded 4.83–4.89%, its lowest since 10 September on one read, and the 2-year fell 11bp to 4.471% (CNBC).

The fuel underneath that was oil. Brent settled at $94.62 on Wednesday, down 8.7% and its lowest since 31 August, as the Democratic sweep was read as pressure for a deal with Iran (ForexTraders). Lower crude trims the inflation leg of the hawkish case, and gold priced that immediately.

Then the part that keeps this plan short-lived. Nothing about the Fed has eased. A quarter-point hike on 9 December was priced at 76% (Barchart), 80.7% on CME via Euronext — up from 77.4% the previous session — and 82% in OCBC's table. Thursday's data did not argue for cuts: jobless claims fell to 189,000 against roughly 196,000 expected, a 3.5-month low, Q3 productivity beat and unit labour costs fell, and Cleveland's Beth Hammack said again that restrictive rates are what cooling inflation requires. The Bank of England hiked to 4.00% on an 8–1 vote the same morning (USAGOLD).

So gold is being bought on a falling dollar in a market that is pricing a hike. That is not a contradiction — it is a currency trade — but it is a trade with a short half-life, and it is why the plan has a clock on it.

Payrolls at 13:30 UTC is the wall

October payrolls land at 13:30 UTC (08:30 EST, 21:30 SGT) with a median forecast of +65,000, against +29,000 in September, unemployment steady at 4.2% and average hourly earnings at +0.3% m/m (FactSet, ForexFundamentals). It is a modest number to clear: the trailing 12-month average is 41,300, and six of the last twelve releases have beaten the median and six have missed.

The tails are what matter. A print around 100,000 with unemployment at 4.2% or lower hardens the December hike, puts the 10-year back over 4.95% and sends gold at $4,030, then $4,000. A print under 30,000 does the opposite and puts $4,177–4,199 in play. Michigan sentiment follows at 15:00 UTC with a 47.8 consensus and will be a footnote by then.

The desk does not trade that coin flip. No position through the print: any unfilled order is cancelled at 13:15 UTC, and anything open is closed at the same minute.

The plan

  • Instrument: XAUUSD spot
  • Session: Friday 6 November 2026, 04:00–21:15 SGT (20:00 UTC Thursday to 13:15 UTC Friday)
  • Bias: higher while $4,061 holds; a close under $4,030 ends the sequence
  • Entry: $4,080 buy limit
  • Stop loss: $4,022
  • Take profit 1: $4,138 (close half)
  • Take profit 2: $4,196
  • Risk-reward: 1:1 to TP1 and 1:2 to TP2 — $58 at risk, $58 and $116 paid
  • Do not take it if: spot trades under $4,061 before the order fills, or the order is still unfilled at 13:15 UTC
  • Time stop: flat at 13:15 UTC Friday, no position through 13:30

The arithmetic, from the desk's own daily bars. ATR(14) is 79.20 over the fourteen completed sessions to 4 November — mean range the same, minimum 37.90, maximum 156.96 — so the stop at $4,022 is 0.73 × ATR away from the fill, and the entry sits 0.39 × ATR ($31.24) under spot. The stop is $7.92, or 0.10 × ATR, beneath the shelf that held four times on 4 November between 12:00 and 15:00 UTC ($4,032.14, $4,030.52, $4,032.62, $4,029.33), and $39.40 beneath today's low. TP1 is $8.92 under the session high, so the first target does not need a new high to pay.

Break-even on that geometry is a 40% strike rate if both targets are reached (average win 1.5R against a 1R loss), and 66.7% if the runner is scratched at entry and every winner pays 0.5R. At a 45% hit rate the expectancy is +0.125R per trade. That is the honest bar: this plan needs four wins in ten, and it cannot get there by moving the stop closer.

Why this entry and not a deeper one. Yesterday's $4,030 buy limit was never touched — the session low was $4,061.40, $31.40 away — while gold ran $116 above the entry. That is the same failure the desk booked on 22, 23 and 30 October: a limit parked where the market did not go, scored as 0R. Today's order differs in one respect. $4,080 printed this afternoon, in the London reversal bar whose low was $4,076.52, and the four hours since have printed lows of $4,083.97, $4,095.69 and $4,098.35. A retest is a $31 pullback, 0.39 ATR, and both of the last two sessions dipped before resuming — today $70 from the Asian high into the London open, Wednesday $37 before the rally carried another $143. If it does not fill, that is a 0R and the ledger will say so; the order does not get carried into the print to avoid it.

Execution. The stop assumes the Investing.com daily and hourly feed used for every level above; a second feed prints today's low at $4,061.24 rather than $4,061.40, a $0.16 difference that changes nothing structural (myfxbook). Gold spreads widen in the hour around payrolls, which is another reason there is no position in it. Use limit orders at both targets; a market exit on the 13:15 flatten is worth a few dollars of slippage and is the correct trade.

What makes it wrong

  • A close under $4,030. The shelf is the last higher low of the breakout sequence; losing it puts $4,000 and then $3,970 back on the map, and the stop at $4,022 is designed to be out before that happens.
  • A lower open. If the 00:00–02:00 UTC bars trade under $4,061, the outside day has resolved down and the retest is a trap. Cancel the order rather than let it fill into a falling market.
  • A grind, not a pullback. The likeliest failure is not a loss, it is another 0R: Asia marks time near $4,120, London pushes at $4,146.92, and the order sits unfilled until it is cancelled. That is the fourth time in ten sessions this pattern has cost the desk a research cycle without a trade, and it is the reason the entry is $31 away instead of $50.

The structure read is on the signals board, yesterday's plan it replaces is here, the payrolls thread runs through the NFP hub, and every verdict — including the no-fills — is on the track record.

Sources

Not financial advice. Levels are the desk's own read from the feeds named above; a $58 stop on a $79 ATR instrument is a wide stop and position size must be cut to fit it, not the other way round.

Trading these levels? XM — MT5, gold, demo first paid link