Back to session review

Session review · XAUUSD · Oct 2, 2026

Both Coil Edges Failed

On October 2, XAUUSD broke both sides of the prepared coil and closed back inside it after a soft US employment report. The Whipsaw call and Neutral / Wait lean were right, but none of the three scenarios completed within its own trigger and invalidation rules. The next preparation should require a retest before treating a single post-data H1 close as durable.

Grading the exact preparationGOLD Session Analysis: October 2, 2026, NFP Decides the CoilOpen the original prep

Session vs preparation

H1 · UTC · XAUUSD
  • Observed candles
  • Reviewed session
  • Preparation levels
XAUUSD — session vs prep levelsH1
Shaded band: the reviewed session. Price lines: the original preparation’s levels. Times shown in UTC.

Called vs actual

Every element is assessed against the preparation as written.

Grade card

10 of 14 correct
  1. Day-type callCorrect
    Called
    Event-suspended through the 12:30 UTC employment cluster, then Whipsaw for the completed day
    Actual
    Gold broke both 4,139.17 and 4,192.67 intraday, but neither break held into the close
  2. LeanCorrect
    Called
    Neutral / Wait until a fresh post-data break held with price and driver confirmation
    Actual
    Price closed 34.83 below the open, but back inside the original coil after both directional breaks failed
  3. Conditional leanPartial
    Called
    Long after 14:30 UTC on a 3 to 15 dollar H1 close above 4,192.67 and a held retest, or short on the matching break and failed retest below 4,139.17
    Actual
    The upside close was 24.85 above the trigger and failed its next retest. The later downside close and failed retest qualified, but the move stopped at 4,125.00 and reclaimed 4,139.17 before the close
  4. Lead scenarioIncorrect
    Called
    First move fails, then two-sided rotation, 34%, with no operative lead because both directional branches carried 33%
    Actual
    The broad shape was two-sided, but the post-data H1 close above 4,195.67 invalidated the branch, and the cross through 4,177.10 came later than its 60 to 90 minute trigger window
  5. Scenario map qualityPartial
    Called
    Three nearly equal branches covered first-move failure, bullish resolution, and bearish resolution
    Actual
    The map anticipated both directions, but no branch completed within its own trigger, timing, and invalidation rules
  6. Driver stackCorrect
    Called
    The employment report should outrank unverified real-yield and dollar inputs, with price confirmation required after the blackout
    Actual
    Payrolls printed at 29,000 against 52,000 forecast, unemployment rose to 4.2% against 4.1%, and wage growth was 0.1% against 0.3%. Gold's delayed rally then failed completely
  7. 4,370.94 to 4,375.83 resistance bandCorrect · not activated
    Called
    Major supply only if a broader repair reached it
    Actual
    The session high was 4,227.40
  8. 4,291.40 to 4,315.55 resistance bandCorrect · not activated
    Called
    Broken structure and supply above the immediate repair levels
    Actual
    Price never approached the band
  9. 4,244.28 decision edgeCorrect · not activated
    Called
    Acceptance above it would weaken the sequence of lower H4 highs
    Actual
    The daily high stopped 16.88 below it
  10. 4,218.95 resistanceCorrect
    Called
    First bullish target after a valid break of 4,192.67, with rejection keeping the repair incomplete
    Actual
    Price reached 4,227.40, then reversed below 4,218.95 in the next hour and later crossed the full coil
  11. 4,192.67 upper triggerCorrect
    Called
    Bullish resolution required a 3 to 15 dollar post-blackout H1 close above the level and a held retest
    Actual
    The 15:00 UTC candle closed at 4,217.52, too far beyond the permitted displacement, and the next candle closed back below the trigger
  12. 4,177.10 anchor pivotCorrect
    Called
    Repeated crossings would support rotation, while holding one side alone would not confirm direction
    Actual
    Price crossed the pivot in both directions and the late selloff carried through it without producing a durable outer-edge break
  13. 4,139.17 lower triggerPartial
    Called
    A displaced post-blackout close and failed retest would activate bearish resolution toward 4,110.50
    Actual
    The 18:00 UTC candle closed at 4,127.12 and the next hour rejected near 4,139.17, but price never reached 4,110.50 and later reclaimed the trigger
  14. 4,110.50 major floorCorrect · not tested
    Called
    Acceptance below it would open price discovery beyond the mapped range
    Actual
    The session low was 4,125.00

Showing 14 of 14 elements

The tape

How the session unfolded, in chronological order.

The tape

  1. Open, first 60 minutes

    Gold opened at 4,177.57 and held near that level through the first recorded H1 candle. The session turned sharply lower at 03:00 UTC, falling from 4,181.73 to a 4,153.31 close. The next hour reached 4,133.64, below the prepared 4,139.17 lower edge, but recovered to close at 4,151.24. By 06:00 UTC, price had returned to 4,177.58. The first downside break was an Asian sweep, not a confirmed bearish trigger.

  2. Mid-session

    London pushed into the upper half of the coil. Gold reached 4,196.67 at 09:00 UTC and 4,196.28 at 11:00 UTC, but both moves closed back below 4,192.67. At 12:30 UTC, Nonfarm Payrolls printed at 29,000 against 52,000 forecast, unemployment rose to 4.2% against 4.1%, and monthly wage growth slowed to 0.1% against 0.3%. The first response was contained. Gold dipped to 4,165.91 during the 13:00 UTC hour and recovered to 4,182.39, still inside the coil.

  3. Late / close

    The delayed move arrived at 15:00 UTC. Gold surged to the session high at 4,227.40 and closed the hour at 4,217.52, above both 4,192.67 and 4,218.95. That break did not hold. The next hour closed at 4,190.88, and the 17:00 UTC candle fell through 4,177.10 to close at 4,158.85. Selling extended to 4,125.00 at 18:00 UTC, with an H1 close at 4,127.12 below the lower trigger. The follow-through stopped there. Price recovered above 4,139.17 by 21:00 UTC and settled at 4,142.74. The daily range was 102.40, about 1.19 times the preparation's 86.29 D1 ATR, while the close remained inside the original 4,139.17 to 4,192.67 coil.

What we learned

Evidence that should shape the next preparation.

What we learned

The moderate surprise was not that both sides traded. The preparation called Whipsaw. It was the sequence and force of the failures: a post-data H1 close above the lead branch's invalidation reversed through the full coil, then a qualified bearish trigger also failed before reaching its first target.

  1. Precondition checks

    Keep the completed-day Whipsaw call separate from the Event-suspended pre-release phase. That distinction worked, and the same structure should remain in the next payroll preparation.

  2. Tradability standard

    Define a held H1 close as a close plus a successful retest or a second confirming close. One close at 4,217.52 looked decisive, but the next hour finished back below 4,192.67.

  3. Driver-stack ordering

    Treat the tape's refusal to hold a rally after uniformly soft employment data as a warning that the unverified yield and dollar response did not support the price move. Do not promote the macro narrative above failed price confirmation.

  4. Tradability standard

    Add a follow-through requirement before projecting 4,110.50 from a bearish trigger. The close at 4,127.12 and failed retest below 4,139.17 qualified mechanically, but the market stopped at 4,125.00 and reclaimed the edge.

  5. Precondition checks

    Broaden the first-move-failure clock on payroll days when the first response is muted. The decisive upside break arrived 2.5 hours after the release, and the full rotation through 4,177.10 took another two hours. Session-analysis grading check: pass. Reviewed prep: 2026-10-02-xauusd-session-preparation

A correct forecast does not establish an executable or profitable trade. Every scored preparation stays in the public record.

See the full track record
Read the full original review

Session Summary

On October 2, 2026, XAUUSD broke below the prepared coil in Asia, recovered through the London session, surged above the upper edge after the US employment report, and then reversed through the entire range. The confirmed D1 open was 4,177.57 and the close was 4,142.74. The Whipsaw call and Neutral / Wait lean matched the completed day, but the nominal 34% first-move-failure branch breached its own invalidation and missed the timing of the reversal.

Session:            GOLD NFP Double Failure Review
Symbol:             XAUUSD
Window:             00:00 to 23:00 UTC
Day type called:    Whipsaw
Day type actual:    Whipsaw
Lean outcome:       Correct
Regime:             Two-sided event whipsaw with failed breaks at both coil edges
Preparation:        Partially accurate
Surprises:          Moderate

Grade Card

Prep elementCalledActualVerdict
Day-type callEvent-suspended through the 12:30 UTC employment cluster, then Whipsaw for the completed dayGold broke both 4,139.17 and 4,192.67 intraday, but neither break held into the closeCorrect
LeanNeutral / Wait until a fresh post-data break held with price and driver confirmationPrice closed 34.83 below the open, but back inside the original coil after both directional breaks failedCorrect
Conditional leanLong after 14:30 UTC on a 3 to 15 dollar H1 close above 4,192.67 and a held retest, or short on the matching break and failed retest below 4,139.17The upside close was 24.85 above the trigger and failed its next retest. The later downside close and failed retest qualified, but the move stopped at 4,125.00 and reclaimed 4,139.17 before the closePartial
Lead scenarioFirst move fails, then two-sided rotation, 34%, with no operative lead because both directional branches carried 33%The broad shape was two-sided, but the post-data H1 close above 4,195.67 invalidated the branch, and the cross through 4,177.10 came later than its 60 to 90 minute trigger windowIncorrect
Scenario map qualityThree nearly equal branches covered first-move failure, bullish resolution, and bearish resolutionThe map anticipated both directions, but no branch completed within its own trigger, timing, and invalidation rulesPartial
Driver stackThe employment report should outrank unverified real-yield and dollar inputs, with price confirmation required after the blackoutPayrolls printed at 29,000 against 52,000 forecast, unemployment rose to 4.2% against 4.1%, and wage growth was 0.1% against 0.3%. Gold's delayed rally then failed completelyCorrect
4,370.94 to 4,375.83 resistance bandMajor supply only if a broader repair reached itThe session high was 4,227.40Correct (not activated)
4,291.40 to 4,315.55 resistance bandBroken structure and supply above the immediate repair levelsPrice never approached the bandCorrect (not activated)
4,244.28 decision edgeAcceptance above it would weaken the sequence of lower H4 highsThe daily high stopped 16.88 below itCorrect (not activated)
4,218.95 resistanceFirst bullish target after a valid break of 4,192.67, with rejection keeping the repair incompletePrice reached 4,227.40, then reversed below 4,218.95 in the next hour and later crossed the full coilCorrect
4,192.67 upper triggerBullish resolution required a 3 to 15 dollar post-blackout H1 close above the level and a held retestThe 15:00 UTC candle closed at 4,217.52, too far beyond the permitted displacement, and the next candle closed back below the triggerCorrect
4,177.10 anchor pivotRepeated crossings would support rotation, while holding one side alone would not confirm directionPrice crossed the pivot in both directions and the late selloff carried through it without producing a durable outer-edge breakCorrect
4,139.17 lower triggerA displaced post-blackout close and failed retest would activate bearish resolution toward 4,110.50The 18:00 UTC candle closed at 4,127.12 and the next hour rejected near 4,139.17, but price never reached 4,110.50 and later reclaimed the triggerPartial
4,110.50 major floorAcceptance below it would open price discovery beyond the mapped rangeThe session low was 4,125.00Correct (not tested)

The preparation got the most important restraint right: this was a two-sided event day that did not reward an unconditional directional lean. Its weakness was precision. The scenario map described the session's general shape, but its exact triggers and invalidations left no branch that can be scored as fired.

The Tape

Open, first 60 minutes: Gold opened at 4,177.57 and held near that level through the first recorded H1 candle. The session turned sharply lower at 03:00 UTC, falling from 4,181.73 to a 4,153.31 close. The next hour reached 4,133.64, below the prepared 4,139.17 lower edge, but recovered to close at 4,151.24. By 06:00 UTC, price had returned to 4,177.58. The first downside break was an Asian sweep, not a confirmed bearish trigger.

Mid-session: London pushed into the upper half of the coil. Gold reached 4,196.67 at 09:00 UTC and 4,196.28 at 11:00 UTC, but both moves closed back below 4,192.67. At 12:30 UTC, Nonfarm Payrolls printed at 29,000 against 52,000 forecast, unemployment rose to 4.2% against 4.1%, and monthly wage growth slowed to 0.1% against 0.3%. The first response was contained. Gold dipped to 4,165.91 during the 13:00 UTC hour and recovered to 4,182.39, still inside the coil.

Late / close: The delayed move arrived at 15:00 UTC. Gold surged to the session high at 4,227.40 and closed the hour at 4,217.52, above both 4,192.67 and 4,218.95. That break did not hold. The next hour closed at 4,190.88, and the 17:00 UTC candle fell through 4,177.10 to close at 4,158.85. Selling extended to 4,125.00 at 18:00 UTC, with an H1 close at 4,127.12 below the lower trigger. The follow-through stopped there. Price recovered above 4,139.17 by 21:00 UTC and settled at 4,142.74. The daily range was 102.40, about 1.19 times the preparation's 86.29 D1 ATR, while the close remained inside the original 4,139.17 to 4,192.67 coil.

What We Learned

The moderate surprise was not that both sides traded. The preparation called Whipsaw. It was the sequence and force of the failures: a post-data H1 close above the lead branch's invalidation reversed through the full coil, then a qualified bearish trigger also failed before reaching its first target.

  • Precondition checks: Keep the completed-day Whipsaw call separate from the Event-suspended pre-release phase. That distinction worked, and the same structure should remain in the next payroll preparation.
  • Tradability standard: Define a held H1 close as a close plus a successful retest or a second confirming close. One close at 4,217.52 looked decisive, but the next hour finished back below 4,192.67.
  • Driver-stack ordering: Treat the tape's refusal to hold a rally after uniformly soft employment data as a warning that the unverified yield and dollar response did not support the price move. Do not promote the macro narrative above failed price confirmation.
  • Tradability standard: Add a follow-through requirement before projecting 4,110.50 from a bearish trigger. The close at 4,127.12 and failed retest below 4,139.17 qualified mechanically, but the market stopped at 4,125.00 and reclaimed the edge.
  • Precondition checks: Broaden the first-move-failure clock on payroll days when the first response is muted. The decisive upside break arrived 2.5 hours after the release, and the full rotation through 4,177.10 took another two hours.

Session-analysis grading check: pass.

Reviewed prep: 2026-10-02-xauusd-session-preparation