Pinned at R2, Asia takes the call

Gold closed Friday at 4,740.90 — right on daily R2 (4,741.50) with weekly S1 (4,747.83) capping the next handful of dollars. The daily structure cracked through EMA20/50 this week even as 15m/1h/4h kept buyers on dip, leaving a textbook timeframe split heading into the Asia open. COT shows managed-money net long at +164k contracts — well past comfortable. The trade for the week is built around one number: 4,750.60. If Asia takes it and holds, the bid is back. If it rejects, the long book is the fuel.

The session

Friday's tape printed a 207-point range (4,672.20 – 4,879.60) — that's the kind of session that resets a regime. Spot finished -1.51% on the day and -2.40% on the week, but is still +4.20% over the month and +41.85% YTD per TradingEconomics. The pullback has the fingerprints of an Iran-truce headline more than a dollar bid: DXY only added 0.24% intraday and the 30-day daily-return correlation between DXY and XAU is flat at +0.00. This is a geopolitical-positioning unwind, not a macro repricing.

The catalyst FXStreet flagged earlier in the week was the recurring "permanent US-Iran truce" theme (Haresh Menghani, Apr 24); TradingEconomics' Friday note cites Iranian FM Abbas Araghchi's arrival in Islamabad as a possible breakthrough vector while also flagging that the Strait of Hormuz closure is still pushing energy and inflation expectations higher. Two-way risk on the same wire — and exactly why price closed mid-range rather than trending.

The close at 4,740.90 is not random. It's a hair under daily R2, a hair under weekly S1, and 9 dollars below daily R3 (4,750.60). That's a confluence wall, and the market parked into it on a Friday. Asia inherits a coiled chart.

Multi-timeframe read

TF	Read
15m	Basing — RSI 57.5, MACD hist negative but turning up, above all EMAs
1h	Stalling — RSI 56.9, MACD hist 4.08 rolling lower, still above EMAs
4h	Constructive — RSI 60.5, above EMA20
1d	Cracked — RSI 48.8, MACD hist 8.28 still falling, below EMA20 and EMA50, above EMA200

The split is the trade. Lower timeframes (15m/1h/4h) say buyers are still active on every dip — this is consistent with the v-bounce off 4,672. Daily says the short-term trend rolled this week and momentum is still bleeding (MACD hist falling from +8.28). EMA200 on the daily is the only structural support left for the bull case, and it's far enough below current price that a flush down has room to run before it gets there.

The honest read: lower TFs are buying, the daily is selling, and they meet at this confluence. Whoever wins the first 4 hours of Asia trading wins the week.

Macro frame

DXY at 98.51 is up 0.24% intraday and +0.42% on the week, but still -1.09% on the month and below EMA20/EMA50/EMA200 on the daily with RSI at 45.9. That's a counter-trend bounce inside a downtrend, not a regime change. US 10y at 4.31% added 2 bp in the last 24 hours — a nudge, not a shove. With DXY-XAU 30-day correlation at +0.00, the dollar isn't telling us anything directional about gold this week.

What matters more: TradingEconomics flags that markets are now pricing only one 25 bp Fed cut for all of 2026. That's the slow-grinding macro headwind under everything else. If real yields back up alongside any Iran de-escalation, the long book has two reasons to lighten. If real yields fade and Iran re-escalates, the same long book gets paid. The geopolitics is the swing factor; rates are the gravity.

Two scenarios

Conviction percentages below are qualitative — a desk-judgement read, not a back-tested probability.

Buy setup

• Trigger: Hourly close above 4,750.60 (daily R3 + clears weekly S1 4,747.83)
• Invalidation: Hourly close back under 4,714.20 (daily P)
• Target: 4,813.77 (weekly P), stretch 4,879 (last week's high)
• Conviction: 40%
• Rationale: 4h is still constructive and 15m is basing under the wall. If Asia bids the open on any softening of the Iran-truce thread or a Hormuz headline, the path of least resistance is short-cover through the confluence to weekly P. The stretch is last week's high — a re-test, not a new high.

Sell setup

• Trigger: Rejection from 4,747–4,750 followed by hourly close under 4,714.20 (daily P)
• Invalidation: Hourly close above 4,750.60
• Target: 4,696 (daily S1), then 4,672 (Friday low), stretch 4,638 (weekly S2)
• Conviction: 55%
• Rationale: Daily TF rolled, MACD hist still falling, COT shows MM net +164k — that's a lot of long inventory looking for an exit if the truce narrative firms over the weekend. Friday's close right at R2 with weekly S1 capping is the type of setup that bleeds lower in Asia when there's no fresh bid. PAXG trading at a -0.70% discount to spot is consistent with the crypto-side already leaning slightly heavy.

Levels worth marking

• 4,879.60 — last week's high, also session high; first ceiling above weekly P
• 4,813.77 — weekly P; magnet on any reclaim of the confluence wall
• 4,750.60 — daily R3; THE level for the week
• 4,747.83 — weekly S1; immediately under R3, makes the wall a band not a line
• 4,741.50 / 4,740.90 — daily R2 and Friday close; current pin
• 4,723.30 / 4,714.20 — daily R1 / daily P; first line of defence on the way down
• 4,696.00 — daily S1
• 4,672.20 — Friday low, must hold to keep 4h bullish read intact
• 4,638.07 — weekly S2; the place a real flush ends
• 4,572.13 — weekly S3; only relevant on a geopolitical de-escalation gap

FXStreet's Apr 24 desk also flags the 200-period EMA on a lower TF at ~4,778 and the rising-channel floor near 4,680 — both consistent with the snapshot pivots above.

Calendar / catalysts

The deterministic snapshot does not include scheduled events for the week. Treat this as a Data gap: I am not going to invent dates. What I do know qualitatively from the sources:

• Fed expectations sit at one 25 bp cut for all of 2026 (TradingEconomics) — any Fed speak that pushes that to zero is gold-negative; any that pushes back to two is gold-positive.
• US PCE is the typical late-month event in this window — check ForexFactory directly before sizing into it.
• Iran-Pakistan diplomatic track was active into the weekend per TradingEconomics; any joint statement is the headline risk for Sunday-night Asia.

Week ahead

Weekend headlines to watch. The unpriced tail is binary on Iran. If a Tehran-Islamabad breakthrough is confirmed over the weekend or if the Strait of Hormuz reopens, gold gaps lower into Asia and the long book gets squeezed — that path targets 4,672 and then the weekly S2 zone at 4,638 quickly. If, instead, talks stall or there's a fresh Hormuz incident, the same confluence wall above gets stress-tested in Asia. Energy prices into Sunday-night CME open are the tell.

First 4–12 hours. Asia has been a buyer of gold on weakness through this whole leg. With Friday closing right at R2 and the lower TFs basing, the base case is a probe higher first — not a flush. Expect a test of 4,747.83 / 4,750.60 inside the first 4 hours, with Tokyo deciding whether to sustain it. If Asia rejects there, Europe inherits a sell-the-rip setup with stops above 4,750.

The level. 4,750.60. If Asia takes it out in the first hour and we see an hourly close above, the entire week's read flips: weekly P (4,813.77) becomes the magnet and the daily-TF crack gets redrawn as a flag rather than a top. If 4,750 caps, the whole week is a sell-rip tape with weekly S2 (4,638.07) as the destination.

Anchor calendar. Without confirmed dates, I'm flagging the categories: any Fed speak that touches the 2026 cut path; PCE if it lands this week; and any non-US central-bank weekend speak that drops Sunday night Asia time. Each lands on these levels — Fed-hawkish prints the sell setup; Fed-dovish prints the buy setup; PCE soft prints the buy setup; PCE hot is the only print that can take 4,750 out cleanly without a geopolitical assist.

The COT footnote. Managed-money net +164k as of Apr 21 is rich. That doesn't mean the top is in — extended longs can stay extended in a regime change like 2026's — but it is the loaded gun under the sell setup. If 4,714 breaks, the COT will tell you why the move accelerates.

Sources cited

• https://www.fxstreet.com/markets/commodities/metals/gold (Haresh Menghani et al., Apr 24)
• https://tradingeconomics.com/commodity/gold (Apr 24/25)

(Reuters, Kitco news index, Al Jazeera and DailyFX requests either errored or returned empty bodies on this fire — degrading gracefully rather than retry-looping.)

(not financial advice)