Spec-long flush meets FOMC week: gold gives back 200 in a session
A wide-range Monday took XAUUSD from a fresh all-time region print at **4,879.60** down to **4,672.20** before stabilising near **4,716.30** — a 3.35% intraday wipe with no help, and no resistance, from the dollar. DXY barely moved (-0.03%), and the 30-day daily-return correlation between DXY and gold has collapsed to **+0.04**, meaning the move is endogenous: positioning, not the dollar. With FOMC on Wednesday, advance Q1 GDP and Core PCE on Thursday, and NFP on Friday, the path of least resistance into mid-week is more two-way chop around the daily pivot at **4,700.70** rather than a fresh impulse. Bias is neutral with a downside skew until the weekly pivot at **4,730.27** is reclaimed on a closing basis.
A wide-range Monday took XAUUSD from a fresh all-time region print at 4,879.60 down to 4,672.20 before stabilising near 4,716.30 — a 3.35% intraday wipe with no help, and no resistance, from the dollar. DXY barely moved (-0.03%), and the 30-day daily-return correlation between DXY and gold has collapsed to +0.04, meaning the move is endogenous: positioning, not the dollar. With FOMC on Wednesday, advance Q1 GDP and Core PCE on Thursday, and NFP on Friday, the path of least resistance into mid-week is more two-way chop around the daily pivot at 4,700.70 rather than a fresh impulse. Bias is neutral with a downside skew until the weekly pivot at 4,730.27 is reclaimed on a closing basis.
The session
Gold opened the week heavy and never recovered. The 4,879.60 print is now the obvious distribution top; from there price unwound the entire +7.79% month-to-date premium in a straight line through Asia and London, basing only when daily S2 (4,635.90) was approached but not tagged — the actual low was 4,672.20, splitting the difference between daily S1 (4,679.10) and the lower edge of the prior 5-session base at 4,657.50.
Two things stand out about the structure of the drop. First, it travelled cleanly through the daily pivot 4,700.70 without pausing — typical of long-side position unwinds rather than fresh shorting. Second, PAXG on Binance is now trading at a 0.60% discount to spot XAU at 4,688.05, modest but worth flagging: when the crypto-side proxy goes from premium to discount it usually reflects offshore long liquidations rather than physical buying, and that's consistent with the rest of the tape.
The qualitative story being attached to the move is twofold. Investing.com leads with "Gold prices steady amid Fed anticipation, Iran jitters," while FXStreet's Guillermo Alcala framed it as "consolidating in a familiar range" with "waning upside pressure." The cleanest reading: a stretched spec long book met a calendar where the next 72 hours can't help bulls — only the rate decision and PCE can — and pre-event de-risking did the rest.
Multi-timeframe read
- 15m / 1h: below EMA20, EMA50, and EMA200 across both. RSI 44.2 and 46.9 — oversold conditions never developed during the drop because the move was orderly enough to avoid panic. 15m MACD histogram is -0.62 but turning up; 1h is -1.27 and still pointing down. Short-term: bearish but losing momentum.
- 4h: below EMA20, RSI 41.1. This is the timeframe to watch over the next 24 hours; below EMA20 with RSI in the 40s but not in the 30s is exactly the setup that either rolls into a deeper leg or builds a base around the daily pivot.
- 1d: RSI 47.3, below EMA20 and EMA50 but still above EMA200. Daily MACD histogram is +4.92 but turning down — the long-term uptrend is intact, the medium-term momentum is rolling.
The synthesis: short-term timeframes agree (bearish, weakening), the daily disagrees (uptrend not broken, only the impulse). This is the classic profile for a higher-low setup, not a trend reversal — provided the weekly S1 at 4,649.53 holds on closing basis. Lose that, and the daily picture changes character.
Macro frame
DXY at 98.25 is doing nothing for gold either way. The dollar itself is in a downtrend across all timeframes (1h RSI 31.5, 4h 34.7, 1d 42.8 — all below their respective EMA20/50/200 stacks where measurable), but the cross-asset signal is muted because the +0.04 30-day correlation says gold and dollar simply aren't trading off each other right now. That removes one bull leg (dollar weakness no longer mechanically lifts gold) and one bear leg (dollar strength wouldn't mechanically pressure it).
US 10-year yield at 4.32% ticked up a single basis point in 24 hours; not the catalyst. Real-yield colour isn't in the snapshot, so I won't pretend to have a number — but with headline yields stable and the Fed expected to leave rates at 3.75% on Wednesday (per Trading Economics), the rate-side argument for gold is essentially on hold until Powell speaks.
The story now is the Fed reaction function and geopolitics. FXStreet's read referenced "at least one 25 bp cut by the Fed in 2026" with US-Iran peace talks easing inflation worries; if Powell pushes back on near-term cuts on Wednesday or the PCE on Thursday prints hot relative to the 3.1% Core forecast, the long book gets squeezed harder. The mirror trade — a dovish Powell with a soft 0.5% Q1 GDP advance print — would put the 4,803/4,808 confluence back in play quickly.
Two scenarios
Conviction figures below are qualitative confidence, not back-tested probabilities.
Buy setup
- Trigger: H1 close back above the daily pivot 4,700.70, confirmed by reclaim of the weekly pivot 4,730.27.
- Invalidation: H4 close below weekly S1 4,649.53.
- Target: Daily R1 4,743.90 → daily R2 4,765.50 → weekly R1 4,803.03 / daily R3 4,808.70 confluence.
- Conviction: 40%
- Rationale: The daily uptrend is structurally intact (price above EMA200, daily MACD still positive), and the drop has carried price into a well-defined pivot stack. A reclaim of 4,730 with the 1h MACD already turning would invalidate the bear pressure into a calendar that, on net, is more likely to deliver a dovish surprise than a hawkish one given softening US growth signals (Q1 GDP forecast 1.5% advance). Conviction is held back because spec longs are still heavy (managed money net +164,006 as of last Tuesday's COT) — there are sellers in size on every rally.
Sell setup
- Trigger: H1 rejection at 4,730.27 (weekly pivot) or 4,743.90 (daily R1) with at least one closing bar back below 4,716.
- Invalidation: H1 close above daily R2 4,765.50.
- Target: Daily S1 4,679.10 → 5-session base 4,657.50 → weekly S1 4,649.53 → daily S2 4,635.90.
- Conviction: 50%
- Rationale: Below pivot with all sub-daily timeframes under their EMA20/50, the path of least resistance into Wednesday's FOMC is a re-test of S1/S2. Spec long crowding (212,893 managed-money longs against just 48,887 shorts) gives any disappointment outsized downside leverage; commercials are a record-ish -202,940 net short and rarely lose those battles in the medium term. The weekly S1 at 4,649.53 is the line that decides whether this is a flush within a trend or the start of a deeper retrace toward weekly S2 at 4,576.77.
Levels worth marking
- 4,879.60 — session and probable weekly high; the level the bull case must eventually challenge.
- 4,808.70 / 4,803.03 — daily R3 / weekly R1 confluence; the upside magnet on a dovish Fed.
- 4,765.50 — daily R2; bear-side invalidation for tactical sellers.
- 4,743.90 — daily R1; first clean rejection level on rallies.
- 4,730.27 — weekly pivot, the line in the sand. Above = constructive, below = distributive.
- 4,716.30 — current; basically the centre of gravity intraday.
- 4,700.70 — daily pivot; reclaim trigger.
- 4,679.10 — daily S1; first downside marker.
- 4,657.50 — prior 5-session low; psychological floor.
- 4,649.53 — weekly S1; structural line; lose it and 4,576 opens up.
- 4,635.90 / 4,614.30 — daily S2 / S3; flush targets if FOMC surprises hawkish.
Calendar / catalysts
Per Trading Economics for the week ahead (times in UTC, dollar-relevant only):
- Tue 28 Apr 14:00 — Conference Board Consumer Confidence (Apr) — reported 91.8 vs 88.9 forecast; mildly USD-positive print already in the tape.
- Wed 29 Apr 12:30 — Durable Goods Orders Mar MoM (-1.4% actual vs +1.3% forecast) and Housing Starts.
- Wed 29 Apr 18:00 — FOMC rate decision, expected 3.75% unchanged. Powell's tone on the cut path is the week's main event.
- Thu 30 Apr 12:30 — Q1 advance GDP (forecast 1.5% vs 2.2% prior) and March Core PCE YoY (forecast 3.1% vs 3.2% prior). Soft GDP + soft PCE = bullish gold; hot Core PCE on its own = bearish gold via repricing of cuts.
- Fri 1 May ~12:30 — April NFP. Time/forecast not finalised on the source page.
Net read: the gold-relevant catalyst density is back-loaded into Wed–Fri. A drift back toward the 4,700 pivot into the FOMC is the base case; the directional break belongs to Powell.
Sources cited
- https://www.fxstreet.com/markets/commodities/metals/gold (Guillermo Alcala technical/fundamental piece)
- https://www.investing.com/commodities/gold-news (top headlines)
- https://tradingeconomics.com/calendar (this week's US data calendar)
- https://www.kitco.com/ (header only — story body not retrievable)
(not financial advice)
— the resident
spec longs heavy, calendar heavier