Gold reopens into a five-event central-bank week with the daily structure cracked

Gold closed last week at 4,740.90 — off 2.40% on the week and 1.51% on Friday's session — with the daily candle now sitting below both EMA20 and EMA50 even as 15m/1h/4h hold above them. That split tape, combined with managed-money still net-long +164k contracts going into a week that contains BoJ, FOMC, BoE, ECB and PCE, is the entire setup. The daily pivot at 4,714.20 is the line that decides whether Asia's 22:00 UTC reopen runs a dip-buy back to the 4,813.77 weekly pivot, or starts forcing a long-liquidation move that S2 at 4,638 starts to look reachable. Conviction is qualitative, not back-tested — and on a calendar like this one, modest.

The session

Friday's tape was wide and ugly: a 4,672.20 – 4,879.60 range — more than $200 of low-to-high — with the close down at 4,740.90 after price failed at the 4,879 prior-week high and rolled the entire day. Five-session range now sits at 4,672.20 – 4,811.00, so Friday's low is the recent floor; we are not far from it. The PAXG/USDT print on Binance closed at 4,706.85 — a 0.72% discount to spot XAU, which is meaningful for crypto-side gold and reads as risk-off / outflow rather than the small premium you usually get on bid days.

What catalysed the move: per FXStreet's read on the tape, dollar strength on the back of capped Fed-cut expectations, paired with persistent US/Iran tension, kept the bid in DXY (98.51, +0.42% week) and pulled real-yield sensitivity back into gold. The move was less about a single headline and more about a week of profit-taking off the 4,879 high running into a long-positioned managed-money book that started to flinch.

Multi-timeframe read

The disagreement is the story.

• 15m and 1h: RSI 57.5 / 56.9, both above EMA20/50. 1h MACD histogram positive (4.08) but rolling — momentum is fading from the bullish side, not flipping.
• 4h: RSI 60.5, above EMA20. The 4h is still the most bullish frame and the one bulls will lean on intraday.
• 1d: RSI 48.8, MACD histogram +8.28 but pointing down, price below EMA20 and EMA50, only EMA200 still beneath. The daily is what cracked Friday.

Read: lower timeframes are bid, the daily is in transition. That is the classic profile of a corrective move that has more work to do unless reversed quickly. The threshold for "reversed quickly" is a 4h close back above 4,778 (FXStreet pegs the 200-period EMA there) and a 1h close above the 4,750.60 daily R3.

Macro frame

DXY at 98.51, +0.24% intraday, +0.42% on the week — but still -1.09% on the month and below daily EMA20/50/200. The 1h DXY RSI is 35.4, near oversold; the dollar is not in a fresh upleg, it is bouncing inside a broader downtrend. That matters: if DXY fails the bounce, gold's downside thesis weakens.

US 10y at 4.31%, +2 bp on the day. With headline inflation prints due Thursday (PCE), real-yield repricing is the live risk in either direction. The wild card in our snapshot is the 30-day daily-return correlation between DXY and XAU: +0.00. That is not a typo — gold and the dollar have decoupled over the last month. The textbook negative correlation is gone, which means DXY is no longer the dominant driver. In a decoupled regime, gold trades its own positioning and central-bank-flow narrative more than the dollar tape, and that is consistent with the COT print: managed-money long +164k, commercials short -202k. A heavy, one-sided book.

No central-bank weekend speak hit the tape in a way our sources flagged. That void itself is interesting going into FOMC Wednesday — the market gets to write its own preview.

Two scenarios

Conviction figures are qualitative confidence, not back-tested probabilities. They reflect how the tape and positioning lean right now; they are not edge.

Buy setup

• Trigger: 1h close back above 4,750.60 (daily R3) after a hold of 4,714.20 daily pivot on the Asia open. Confirmation if 4h reclaims 4,778 (FXStreet's 200-period EMA reference).
• Invalidation: 1h close below 4,696.00 (daily S1). That kills the dip-buy.
• Target: 4,813.77 weekly pivot first, then 4,857.60 (last week's close) and 4,879.60 (prior-week high).
• Conviction: 40%.
• Rationale: Lower timeframes are still constructive, DXY is bouncing inside a downtrend, and 30-day DXY/XAU correlation at zero means dollar strength is not the same headwind it usually is. If Asia defends 4,714, the path of least resistance into a dovish-leaning FOMC is back up to the weekly pivot. The reason this isn't higher-conviction: managed-money is heavily long and needs a fundamental reason to add, not just a chart reclaim.

Sell setup

• Trigger: 1h close below 4,714.20 daily pivot, with 4,696.00 (daily S1) breaking on follow-through. A break of Friday's 4,672.20 low is the cleanest version.
• Invalidation: 1h close back above 4,750.60. That re-establishes the dip-buyers.
• Target: 4,668.70 (daily S3) → 4,638.07 (weekly S2). Stretch target into the FOMC week is 4,572 (weekly S3) only on a hawkish surprise.
• Conviction: 55%.
• Rationale: Daily MACD is rolling, daily price is below EMA20/50, and the COT has +164k of managed-money longs that are still in profit and have something to defend on a break. PAXG trading at a 0.72% discount to spot is a small but real risk-off tell from the crypto leg. The asymmetry favours the downside until the daily reclaims its EMAs, which it cannot do on Asia alone.

Levels worth marking

Pulled from the snapshot, ranked by what the tape will likely actually use:

• 4,879.60 — Friday's session high and prior-week high. Above this and the bear case is dead.
• 4,857.60 — last week's close. Mean-reversion target if buy setup runs.
• 4,813.77 — weekly pivot. The single most important reclaim level on this chart.
• 4,778 — FXStreet's 200-period EMA reference; daily-frame inflection.
• 4,750.60 / 4,741.50 — daily R3 / R2 cluster. Price is pinned here on the close; first decision.
• 4,714.20 — daily pivot. The line that decides Asia's tone.
• 4,705.10 — prior daily low/close (HLC anchor).
• 4,696.00 — daily S1. Sell-setup trigger zone.
• 4,672.20 — Friday's session low and 5-session floor. Clean break = momentum sell.
• 4,668.70 — daily S3.
• 4,638.07 — weekly S2. First serious downside magnet.

Calendar / catalysts

This is one of the most loaded weeks of the year. Per TradingEconomics:

• Tue 28 Apr, ~03:00 UTC — BoJ rate decision (currently 0.75%) + Quarterly Outlook Report. Yen is the most likely G10 mover; gold-via-yen is a real cross-channel.
• Tue 28 Apr, 14:00 UTC — US Consumer Confidence (April).
• Wed 29 Apr, 13:30 UTC — Australia CPI.
• Wed 29 Apr, 13:45 UTC — BoC rate decision.
• Wed 29 Apr, 18:00 UTC — FOMC rate decision (current 3.75%) + 18:30 UTC Powell presser. The week's anchor.
• Thu 30 Apr, 01:30 UTC — China Manufacturing PMI (50.4 expected).
• Thu 30 Apr, 09:00 UTC — Eurozone Q1 GDP flash + April CPI flash.
• Thu 30 Apr, 11:00 UTC — BoE rate decision (3.75%).
• Thu 30 Apr, 12:15 UTC — ECB rate decision (2.15%) + 12:45 UTC presser.
• Thu 30 Apr, 12:30 UTC — US Q1 GDP advance (0.5% est) + March PCE (2.8%/Core 3.0%) + Initial Jobless Claims, all simultaneous.

Three rate decisions on Thursday in a four-hour window, on top of US PCE and GDP. Then PCE is Powell's preferred inflation gauge — landing two days after FOMC. Vol-vol (vol of vol) will be elevated all week.

Week ahead

Weekend headlines that matter. No major central-bank weekend speak surfaced through the sources I pulled. US/Iran tension remains the running geopolitical bid for gold per FXStreet's framing, but it has been live for weeks and is not a fresh catalyst. The market has not yet priced any FOMC dovishness into Friday's close — Friday was sell-the-rip, not position-for-cuts. That asymmetry matters: if Powell tilts even modestly dovish on Wednesday, the reaction function is bigger than usual because the tape went into the weekend leaning the wrong way for it.

Projection, first 4–12 hours. Asia takes the 22:00 UTC reopen. Friday closed below the 4,750/4,741 daily R-cluster and right on the daily R2, so the technical setup favours a soft Asia open with a probe of 4,714.20 in the first session. Tokyo specifically — given the BoJ decision Tuesday and the yen channel — is likely to be the most directional desk. Base case: rangey 4,705 – 4,750 in the first 6–8 hours, with a slight downside lean as managed-money trims into the FOMC.

The level that changes the week. 4,696.00 in the first hour. If Asia takes out the daily S1 inside the first sixty minutes of trade, that's not a normal-Sunday-thin-liquidity flush — that's positioning being unwound, and it opens 4,672 / 4,638 for the week with FOMC as accelerant rather than circuit-breaker. Conversely, a 1h reclaim of 4,750.60 in the first session re-arms the buy setup and turns the week into a mean-reversion play toward 4,813.

Where the calendar lands in our levels. A dovish FOMC reaction targets 4,813 (weekly pivot) and 4,857 (last week's close). A hawkish FOMC reaction targets 4,672 → 4,638. A hot PCE Thursday targets the same downside; a soft PCE re-arms the upside. The BoJ on Tuesday is the cross-asset wildcard — a hawkish BoJ that drives USDJPY lower would mechanically weaken DXY and, despite the 30-day decoupling, would probably support gold; a dovish hold does the opposite.

Sources cited

• https://www.fxstreet.com/markets/commodities/metals/gold — technical levels, RSI/MACD context, geopolitical framing
• https://tradingeconomics.com/calendar — week's central-bank decisions and US data calendar

(Reuters, Kitco, BBC, DailyFX and Al Jazeera attempts failed or returned no extractable content; degraded gracefully.)

(not financial advice)