XAUUSD signal: buy the $4,048 dip, target $4,103
Gold dumped to $4,025 on the Fed statement, then reversed $57 during Warsh's press conference. Thursday's plan buys the $4,048 retracement with $3,993 at risk, targeting $4,103 and $4,158.
$4,062.43$4,028.19$4,060.16$4,042$4,025.04$4,082.34$4,078.28$4,060.54
Trade plan at a glance
- Entry
- $4,048 buy limit
- Stop loss
- $3,993
- Take profit 1
- $4,103 — 1.0R, close half
- Take profit 2
- $4,158 — 2.0R, close the rest
- Risk-reward
- 1:1 to the first target, 1:2 to the second, $55 of risk per unit
- Do not take it if
- an hourly bar closes below $4,025 before the limit fills — the event low is this trade's floor and a close under it voids the order.
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 Wednesday at $4,062.43, slid to $4,028.19 by 06:00 UTC, bounced to $4,060.16 at midday and faded to $4,042. Then the Federal Reserve held rates at 3.75%–4.00%, and inside two hours gold printed both $4,025.04 and $4,082.34 — a $57.30 range with one extreme before the statement and the other after it.
The 20:00 print is $4,078.28 on Investing.com's XAU/USD series, up 0.44% on Tuesday's $4,060.54 close. A second feed quoted $4,075 at the same moment, so call the band $4,075–$4,078.
The statement sold gold; the press conference bought it back
The decision was never in doubt — CME FedWatch had a hike priced under 10% (CNBC). The packaging was the event.
- Two dissents, both for a hike. Governor Lisa Cook and Cleveland's Beth Hammack wanted +25bp. It was Cook's first dissent since she joined in 2022.
- The statement upgraded growth, from "solid" to "expanding at a strong pace", and deleted the sentence that explained September's increase.
- It added a delay clause: "the accumulation of more information will help the Committee determine the policy stance appropriate to ensure a timely return of inflation" (statement and press conference).
Gold read that as hawkish. The 18:00 bar ran $4,067.51 to $4,025.04 — a $42.47 range — and closed $4,033.36, $34 under its own high. Then Warsh took questions at 18:30 and the same market reversed $57:
- The front end rallied while the long end sold off. The 2-year fell 5.20bp to 4.7182%, the 5-year 1.07bp, while the 10-year rose 2.70bp to 5.0466% and the 30-year 5.29bp to 5.3707%. The 2s10s spread widened from 24.94 to 32.84 basis points (InvestingLive).
- The dollar gave the day's bid back. The index climbed to 102.49 into the decision, dropped to 102.05 during the Q&A and closed 102.16, down 0.16% (FXStreet).
- His answers did it. The medicine has "worked some", "most at the Fed don't think the economy is running too hot", and price stability "doesn't hinge on any one calendar date" (InvestingLive).
The hole in that story is the long end. A 10-year at 5.05% is a real-yield headwind, and bullion ignored it for two hours. What gold traded was the front end and the dollar — the two things that price the next meeting rather than the next two years. December is still priced as a hike, somewhere between 82% and 94% depending on the feed.
Four higher lows and the highest high since 22 October
The daily bars have been repairing themselves for a week:
- Lows: $3,975.64 on 23 October, $3,978.15 on 26th, $3,988.14 on 27th, $4,025.04 today. The floor is up $49.40 in four sessions.
- Highs: $4,034.17, $4,029.81, $4,065.87, $4,082.34 — three straight sessions printing a higher high, and today's is the highest since 22 October's collapse.
- The reclaimed level held. $4,060 is where that breakdown started. Tuesday closed $0.54 above it; the 20:00 print sits $18.28 above it.
- One hour did the damage. The 19:00 bar opened $4,033.16, printed $4,027.70 and closed $4,075.83 — +$48.13 in sixty minutes.
That is a trend, so the order is a pullback buy. The caveat is that a vertical two-hour rip is not a base, and the first retracement of a spike often runs deeper than the first shelf — which is why the entry is not at the $4,060 level everybody can see.
The order filled at 05:00, and the time stop took the rest
Yesterday's $4,033 buy limit filled on the Asian slide: the 05:00 bar traded $4,030.19 and the 06:00 bar $4,028.19, so the desk was long from $4,033. The $3,979 stop was never threatened — the day's low held $46 above it — and the 17:00 time stop closed it at $4,042.90 for +$9.90, or +0.18R.
The 20:00 hour later printed $4,082.81, but TP1 at $4,087 missed by $4.19 all session. The track record books a fill and a small win; the cutoff cost the runner, not the trade.
Thursday costs a triple print at 12:30 UTC
This plan is dead by 11:30 UTC, one hour before the first US data since the decision. At 12:30 the BEA releases September PCE and the first estimate of Q3 GDP, alongside weekly claims:
- Core PCE +0.3% m/m and 3.1% y/y, from +0.2% and 3.0% (Morningstar, FactSet consensus).
- Headline PCE +0.5% m/m and 3.6%–3.7% y/y, from +0.3% and 3.4%.
- Q3 GDP around +3.0% annualised against +2.2% in Q2, with Trading Economics at 3.2% and the Atlanta Fed's GDPNow at 3.4%.
- Claims near 196,000–199,000 from 194,000 — a four-week average Warsh called the lowest "in generations".
The asymmetry is in the core number. 0.3% is in the price. A 0.1% print re-prices December lower and the next stop is the $4,145 shelf; a 0.5% print puts the hike back on the table with the 10-year already at 5.05%. The FOMC hub carries the preview this plan builds on: Wednesday was the decision, Thursday is the information.
The plan
- Instrument: XAUUSD spot, MT5 gold
- Session: Thursday 29 October, 00:00–11:30 UTC. London opens at 07:00 and the triple print lands at 12:30.
- Bias: higher while $4,025 holds, with $4,098–$4,119 the band that decides bounce or reversal
- Entry: $4,048 buy limit
- Stop loss: $3,993
- Take profit 1: $4,103 — 1.0R, close half
- Take profit 2: $4,158 — 2.0R, close the rest
- Risk-reward: 1:1 to the first target, 1:2 to the second, $55 of risk per unit
- Do not take it if: an hourly bar closes below $4,025 before the limit fills — the event low is this trade's floor and a close under it voids the order.
- Time stop: flat or unfilled by 11:30 UTC Thursday, one hour before the print.
Entry. $4,048 is 0.465 × ATR below the 20:00 print and sits on the 61.8% retracement of the post-Fed leg — $4,027.70 up to $4,082.34 puts that level at $4,048.57. It is also $12 under Tuesday's $4,060.54 close, so the reclaimed level gets tested and the order fills slightly beneath it rather than at it.
Stop. ATR(14) on daily bars is $65.11, so the 0.7 floor is $45.58. This stop is $55 away, 0.845 × ATR — wider than the floor, deliberately. Three things sit under the entry: today's $4,025.04 event low, the $4,000 round number, and 27 October's $3,988.14 higher low. A stop wedged between them is a stop inside the two most obvious magnets on the chart. $3,993 is $32.04 (0.49 × ATR) below the event low and $7 below $4,000 — the 0.1 × ATR buffer the rules require beyond a structural level, sized to survive a sweep instead of being collected by one.
The rules squeeze here. With the stop at $3,993, a 0.70 × ATR risk needs an entry at $4,038.58 — 0.61 × ATR under spot, outside the half-ATR entry rule, so an order that never trades. The desk kept the entry where price reaches it and accepted 0.845 × ATR. The reference feed is Investing.com's XAU/USD series; a broker quoting $2 higher shifts the R by 3.6%.
Targets. TP1 at $4,103 is exactly 1.0R and lands $5.12 above the 61.8% retracement of the entire 22 October collapse — $4,020.99 to $4,145.40 gives $4,097.88 — the first place a bounce has to prove itself. TP2 at $4,158 is 2.0R and $12.60 over 22 October's $4,145.40 high, the top of the shelf price fell from on its way to $3,975. It needs a trend day that finishes before noon UTC, and it will often not get one.
What the R:R demands. Half off at 1R plus a 2R runner pays +1.5R on a full winner against −1R on a loser, so the plan breaks even at 40%. If the runner is scratched at entry the average winner is +0.5R, which needs 66.7% — and with an 11:30 UTC time stop the runner is more likely to be scratched than paid. The edge here is TP1, not TP2.
Invalidation
- An hourly close below $4,025 before the fill. Today's low is the event low. Losing it voids the order rather than inviting a wider stop.
- Two hourly closes below $4,040 after the fill. The shelf has failed; exit near −0.15R instead of −1R and never move the stop away.
- A 0.5% core PCE print. December tightens, the 2-year gives back its 5bp and gold retests $4,000. The stop is $55 away for exactly this case.
- A gap above $4,103 at the Asian open. The limit never fills and the plan is void — an unfilled order is not a winner.
- A quiet night. Gold holding $4,070–$4,090 into London expires the order.
Spread and slippage. Against $55 of risk a $0.35 spread is 0.6% and a $1 spread 1.8%. The limit sits in the $4,043–$4,055 congestion that traded all afternoon before the statement, which is what makes a resting order realistic rather than hopeful.
Risk note. This is analysis, not financial advice. Leveraged gold can lose more than the deposit, and a plan is not a prediction. Size from the $55 stop, never from the $110 target. Every level is computed from the Investing.com XAU/USD daily and hourly series pulled at 20:05 UTC on 28 October 2026, with a second feed agreeing within $3.30. New York still had two hours to run, so treat $4,078.28 as a timestamp rather than a settlement.
Sources: CNBC on the decision, Warsh's press conference transcript, InvestingLive before-and-after, FXStreet on the dollar index, Morningstar on the PCE consensus.