Gold flushes 150 points with DXY asleep — read the unwind, not the trend

Gold gave back 151 dollars off the session high while DXY moved twelve basis points. That asymmetry tells you what's driving today's tape: managed-money longs trimming into a two-day FOMC, not a regime change. The 30-day daily-return correlation between gold and DXY currently sits at +0.04 — effectively zero — so anyone framing today as a "dollar move" is reading the wrong instrument. The level that matters now is the weekly pivot at 4,730.27, sitting almost exactly where Monday's tape closed.

The session

Spot pushed up to 4,879.60 earlier, then walked all the way back to 4,672.20 before settling around 4,728.10 — a 207-dollar intraday range, which is wide even for current vol. Per an Investing.com analyst quote on the day, "Gold prices moved in a tight range on Monday as the dollar dithered amid some hopes for a U.S.-Iran de-escalation," with a separate framing of "peace deal priced in as inflation keeps pressure on prices." Translation: geopolitical risk premium got unwound at the highs, and there was nothing on the macro tape to catch the knife.

What's notable is what didn't happen. DXY closed +0.12% on the day. The US 10-year yield added two basis points to 4.31%. A genuine "stronger dollar" sell-off prints a much larger DXY move; a "yields are back" sell-off prints 10-year +8 to +10 bp. Neither happened. The selling is internal to gold.

PAXG (the tokenised proxy on Binance) trades at 4,698.45, a -0.63% discount to spot — a meaningful gap. When the crypto-traded proxy leads spot lower, you're typically looking at leveraged longs hitting stops in a thinner book, not a fundamental bid disappearing.

Multi-timeframe read

• 15m — RSI 49.9, MACD histogram -1.12 and rolling, price below the 20EMA but holding the 50/200. Short-term momentum is bearish but not exhausted. A first bounce off the 15m 50EMA is the obvious tactical buy that the desk would consider fading rather than chasing.
• 1h — RSI 51.8, MACD histogram 0.90 but turning, above the 20EMA but below the 50. Squeezed between EMAs. The cleanest "indecision" frame on the board.
• 4h — RSI 43.1, below the 20EMA. The weakest read on the chart. The 4h has been giving up support for several bars now and is the timeframe most aligned with the bear scenario.
• 1d — RSI 48.0, MACD histogram still positive at 5.64 but rolling, below the 20 and 50 EMAs, above the 200. Pullback inside an intact higher-timeframe uptrend. The 200EMA holding below price is the line that says correction, not top.

Where the timeframes agree: every single MACD histogram is rolling lower. Where they diverge: the 15m and 1h still defend short-term EMAs while the 4h and 1d have given them up. That signature — fast frames stretched, slow frames intact — is what a corrective pullback in a bull market looks like, not a reversal. Until the daily 200EMA breaks, that read holds.

FXStreet's own technical desk reads this similarly: "the broader uptrend remains intact, even as momentum cools." Investing.com's technical summary disagrees — daily is Strong Sell, weekly Neutral, monthly Strong Buy. That spread of timeframe scores is the trade.

Macro frame

DXY at 98.40, +0.12% on the day, +0.36% on the week, but -1.50% on the month. On the daily the dollar sits below its 20, 50 and 200 EMAs with RSI 44.5 — its own modest downtrend. The 10y at 4.31% printed a 2bp move that is not signal. With the FOMC meeting starting Tuesday and the decision Wednesday, the rates market is sitting on its hands.

The single most important macro number for today's gold tape is that +0.04 daily-return correlation. Gold has been trading on its own narrative — central-bank demand, geopolitical premium, and now positioning — not on USD direction. Scenarios that hinge on "if the dollar rallies X..." don't translate cleanly here. The dollar isn't driving.

On real yields: with nominal 10y at 4.31% and breakevens around the recent 2.4–2.5% area, real 10y likely sits near 1.8–1.9% — historically a regime in which gold should struggle. That it hasn't is itself the story, and points back to non-US monetary demand as the marginal bid through this rally.

Per FXStreet's overlay, the two non-monetary catalysts hanging over the tape are US–Iran negotiations (with Strait of Hormuz access as the binary risk) and Israel–Hezbollah escalation. Both can put the safe-haven bid back in gold without moving DXY — exactly the kind of asymmetric driver this market has been trading on.

Two scenarios

Note: the conviction percentages below are qualitative confidence reads, not back-tested probabilities.

Buy setup

• Trigger: Reclaim of the weekly pivot 4,730.27 with a 1h close above 4,743.90 (daily R1 / approximate 1h 50EMA confluence).
• Invalidation: 1h close below 4,679.10 (daily S1) or daily close below weekly S1 4,649.53.
• Target: First 4,765.50 (daily R2), then 4,803.03 (weekly R1) on a stretch.
• Conviction: 45%
• Rationale: The daily uptrend structure is intact above the 200EMA. With DXY refusing to bid and the FOMC priced for an unchanged hold at 3.75%, a dovish-leaning press conference Wednesday is the natural catalyst for a reclaim. This setup is timing-dependent, not thesis-dependent — the higher-timeframe trend is already on its side.

Sell setup

• Trigger: 1h close below 4,679.10 (daily S1) with the 4h MACD histogram confirming negative.
• Invalidation: 1h close back above 4,743.90 (daily R1).
• Target: First 4,649.53 (weekly S1, confluent with last week's low 4,657.50), then 4,635.90 (daily S2) on follow-through.
• Conviction: 40%
• Rationale: Managed-money net long at +164,006 contracts (CFTC, 21 April — gross long 212,893 vs short 48,887) is heavy. Into a two-day FOMC, the pain trade is more length getting flushed. The 4h is already below its 20EMA with RSI sub-45, and PAXG's -0.63% discount to spot says retail/leveraged length is leaning the same way.

(The remaining ~15% of outcomes is a 4,679–4,744 chop into Wednesday's print with neither setup triggering cleanly. That's the FOMC-week base case and it's not unlikely.)

Levels worth marking

Level	Note
4,879.60	Today's session high — the level any bounce has to deal with
4,803.03	Weekly R1
4,765.50	Daily R2
4,743.90	Daily R1, rough 1h 50EMA area
4,730.27	Weekly pivot — the line for the week
4,722.30	Last week's close, prior daily HLC anchor
4,700.70	Daily P
4,679.10	Daily S1
4,672.20	Today's session low
4,657.50	Last week's session low
4,649.53	Weekly S1 (confluence with the above)
4,635.90	Daily S2

The cluster between 4,649 and 4,657 is the high-conviction support shelf — weekly S1, last week's low, and the bottom of the prior 5-session range all sit inside that band. That's where dip buyers should defend. Below it, the next real shelf is weekly S2 at 4,576.77, and the tape would need to break for that to come into play.

Calendar / catalysts

Per TradingEconomics, scheduled US events through Thursday (treat all numbers below as forecast/scheduled, not prints):

• Tue 28 Apr — CB Consumer Confidence; Richmond Fed Manufacturing. Mid-impact, unlikely to override the FOMC frame.
• Wed 29 Apr 12:30 UTC — Durable Goods MoM (forecast -1.4% vs prior +0.5%).
• Wed 29 Apr 18:00 UTC — FOMC rate decision (forecast 3.75%, unchanged hold). 18:30 UTC press conference. This is the week.
• Thu 30 Apr 12:30 UTC — Q1 GDP advance (forecast 1.5% vs prior 2.2%); March PCE headline (forecast 3.3% YoY) and Core PCE (forecast 3.1% YoY); Initial Jobless Claims. Double-header risk event the day after the FOMC.

Base case for the next 48 hours: gold ranges 4,680–4,770 into Wednesday and resolves on Powell's tone. Any geopolitical print between now and then — particularly Iran or Israel–Hezbollah headlines flagged by FXStreet — re-prices the risk premium that just got unwound.

Sources cited

• FXStreet — gold technical & catalysts: https://www.fxstreet.com/markets/commodities/metals/gold
• TradingEconomics — calendar: https://tradingeconomics.com/calendar
• Investing.com — technical summary and analyst pull-quote: https://www.investing.com/commodities/gold

(Reuters and Kitco fetches failed at request time; ForexFactory returned 403. The qualitative read above is built off the three sources that did return.)

(not financial advice)