traded at $4,045.89 an ounce Tuesday, with at $4,046.20, both grinding toward the $4,000 handle through the European session. Futures opened at $4,083, up 0.1% from Monday’s settlement, then gave the entire gain back and more, printing $4,027 by 7:24 a.m. Eastern.
Monday had looked like the start of something. Spot gold gained 1.39% to close at $4,108.91 as the dollar softened and Treasury yields fell, with adding 2.76% to $59.77. The metal pushed above $4,100 intraday and failed to hold it. That failure is the entire setup for Tuesday.
Gold has opened below $4,100 in every session since July 14. That is a two-week ceiling built at a round number, and it has been tested and rejected repeatedly. The path of least resistance sits lower while that pattern holds.
Silver has taken the harder hit, trading at $57.53 an ounce, down 1.49% from $58.40 Monday, with another read putting the close at $57.68 for a 1.18% decline. The ratio between the two metals is widening again, which is what happens when industrial demand assumptions come under pressure alongside monetary ones.
The driver is not complicated. The dollar sits at a one-month high, with pinned near the mid-1.1300s and at fresh July lows around 1.3270. A firmer dollar raises the cost of bullion for every buyer outside the United States, and it does so mechanically rather than sentimentally.
What makes Tuesday unusual is that the dollar strengthened while Treasury yields fell. sits at 4.628%, down a basis point, at 4.306%, and flat at 5.121%. Falling yields normally support a non-yielding asset. They are not supporting it today because the dollar bid is coming from risk aversion — a 10.84% collapse in South Korea’s benchmark and a broad semiconductor rout that has sent capital into cash rather than into metal.
That distinction matters for what happens next. Gold is not being sold as a hedge failure. It is being sold because the currency it is priced in is the preferred hedge this week, and because a policy decision lands in roughly 30 hours that could confirm or reverse that preference.
The Dollar at a One-Month High Is Doing Most of the Damage
The currency channel is the cleanest way to understand Tuesday’s tape. Dollar strength has been the single most reliable predictor of gold weakness through 2026, and the correlation has tightened as the metal has become more rate-sensitive and less event-sensitive.
The sequence that broke gold’s January momentum ran through the dollar. Prices had become stretched through January, and the appointment of a new Federal Reserve chair with a hawkish reputation triggered a sharp correction into month-end. Energy prices then took over. The escalation of the US-Iran conflict pushed crude sharply higher, which pushed inflation expectations higher, which forced markets to price a higher-for-longer rate path that had not been in the curve. The dollar surged to a 13-month high on that repricing.
Bullion fell roughly 28% from its January peak on that chain of events, and the causation runs almost entirely through real yields and the currency rather than through any deterioration in gold’s underlying demand picture.
Tuesday adds a wrinkle. has now collapsed — down 3.71% to $85.08 and down about 3% to $80.11, extending Monday’s 8.7% Brent decline as the US-Iran pause held for a third session. Energy deflation should, mechanically, soften the inflation impulse and reduce the case for further tightening. That is gold-positive.
The market is not trading it that way yet, and the reason is timing. Falling oil takes months to filter through core inflation prints. A Federal Reserve decision lands tomorrow. Traders are not going to front-run a disinflationary transmission channel that will not appear in the data until the fourth quarter, when a hawkish statement could arrive in 30 hours.
The dollar is therefore holding its bid on two separate legs: the policy leg, which is about Wednesday, and the risk leg, which is about semiconductors. Both would need to reverse for gold to reclaim $4,100 convincingly.
For context on how much this currency channel matters, physical demand from Asia has remained firm throughout. That firmness has not been enough to offset a dollar at monthly highs. When the price-setting mechanism sits in the currency market, physical tightness becomes a floor rather than a driver.
A Fed Decision With No Dot Plot and 35.8% Odds of a Hike
The Federal Open Market Committee opened its two-day meeting Tuesday with the target range at 3.50% to 3.75%. The statement arrives Wednesday at 2 p.m. Eastern, followed by a press conference at 2:30.
Market-implied odds of a quarter-point increase stand at 35.8%, up from 25.77% a week earlier. Other readings across the week have clustered between 30% and 38%, with hold probability running from 62% to roughly 70%. The direction of the repricing matters more than the level: hike odds have tripled in a fortnight, and there is no meaningful probability assigned to a cut at any point in this meeting’s pricing.
Roughly 80% odds of a September increase are now embedded in the curve. That is the number doing the damage to gold, not the July decision itself.
This meeting produces no Summary of Economic Projections and no dot plot. That removes the usual anchor traders use to map the path beyond the immediate decision, which leaves the press conference as the only forward-looking input. The current chair has committed to reducing forward guidance and declined to submit individual projections at his first meeting in June, where nearly half of policymakers indicated support for a hike later in the year.
The specific language to watch is the balance-of-risks paragraph. June’s framing described inflation risks as too high. Any softening of that phrasing would be read as dovish and would immediately compress real yields — the mechanical input that sets gold’s opportunity cost. Any hardening would confirm the September path and push the metal toward the $4,000 handle.
Inflation data supports both cases. The June consumer price print at 3.5% came in cooler than the trajectory implied, after readings ran 2.4% in January and February, 3.3% in March, 3.8% in April, and 4.2% in May. That deceleration is the strongest argument for a dovish hold. Against it, the June labor report showed just 57,000 payrolls against forecasts near 110,000 — weak enough to argue against tightening, but also weak enough to raise questions the committee would rather not answer publicly.
The data calendar behind the decision is dense: ADP employment Tuesday, second-quarter GDP and jobless claims Thursday alongside core PCE, then Chicago PMI and Michigan inflation expectations Friday.
Thirty-One Sessions Below the 200-Day and an Active Death Cross
The long-horizon technical structure is unambiguously damaged, and the numbers state it plainly.
Gold’s 200-day moving average sits near $4,496. That average was broken in early June, and by July 20 the metal had logged 31 consecutive trading days beneath it — the longest sustained stretch below that indicator since 2022. Price currently sits roughly 10% below the line.
A death cross is active, with the shorter averages having crossed beneath the longer ones and the MACD histogram in negative territory. The 50-day average has converged into the $4,200 region, where it now overlaps with both a long-term uptrend line and a shorter descending trendline. That confluence makes $4,203 the pivot that defines whether this is consolidation or continuation.
The shorter frames tell the same story in miniature. The 20-day average sits near $4,068. On the two-hour chart, the 50-period exponential average sits at $4,070 and the 200-period at $4,077 — gold failed to reclaim either on Monday’s rally, and that failure is why Tuesday opened weak. The four-hour 50-period average sits at $4,065.78, currently acting as first resistance.
Momentum is soft without being extreme. The four-hour relative strength index reads 44.8, below the neutral 50 line and below its own signal line, but well above oversold. On the daily frame, RSI has been oscillating around 33 during the worst of the selling. The metal is not stretched enough for a mechanical bounce, which means any recovery has to be driven by news rather than by positioning.
Volatility remains contained. The average true range has been running near 2.1% of price, and gold’s beta against the S&P 500 sits at 1.00 — a neutral reading that undercuts the safe-haven framing. An asset moving one-for-one with equities is not currently functioning as a portfolio hedge.
The structure that would change this view: a daily close above $4,203, followed by a successful retest. That would put the 200-day at $4,496 in play as the first serious upside objective, roughly 11% above spot. Nothing shorter than that alters the trend.
The Levels That Matter: $4,021 Below, $4,066 Above
Immediate support is layered tightly beneath current price, which is why the tape has felt heavy rather than collapsing.
The first shelf sits at $4,021 to $4,025, which contains this week’s low of $4,022. Directly beneath that is $4,004, the 78.6% Fibonacci retracement of the recent advance, and then the $4,000 psychological level itself. Those three levels occupy a $25 band, which makes them function as one zone rather than three.
Gold has defended $4,000 repeatedly. It broke below the level on June 24 for the first time since November 2025, then recovered. A higher swing low was established at $3,959 in mid-July, and the metal touched $3,942 earlier this month before bouncing. That sequence — lower lows arrested, then a higher low printed — is the only genuinely constructive pattern on the chart.
Below $4,000, the next liquidity pocket sits at $3,950, then a broader secondary support band at $3,800 to $3,850 that represents the late-2025 breakout zone. That is the level where the structural case would need re-examination rather than restatement.
Overhead resistance is stacked and dense. The first hurdle is $4,066 to $4,077, where three separate moving averages converge. Above that is $4,100, which has capped every session since July 14. Then $4,150, then the $4,200 to $4,230 pivot zone where the 50-day average and two trendlines intersect. Beyond that, $4,300 and $4,375 come into view before the 200-day at $4,496.
Six resistance levels between $4,066 and $4,496. That is a heavy ceiling, and it explains why the metal has been unable to convert any of July’s bounces into trend.
Daily range projections put Tuesday’s expected band at $4,007.83 to $4,157.41 — a $150 span that captures both the support cluster and the first two resistance levels, which is an honest reflection of how binary Wednesday makes the setup.
The trading structure that follows: neutral to slightly bearish below $4,066, with $4,000 the line that keeps a rebound alive. A confirmed break above $4,100 would be the first genuine evidence that buyers have regained control.
Twenty-Eight Percent Off the January Peak, and Down Just 7% on the Year
Scale matters here, and the drawdown framing tends to overstate the damage.
Spot gold set a record at $5,595.46 on January 29, 2026, after surging nearly 30% inside a single quarter. That advance built on 2025, when spot prices rose 64.5% — the strongest annual performance since 1979. The 52-week range runs from roughly $3,287 at the low to $5,586 at the peak.
At $4,046, gold sits about 27.7% below that January high. Measured from the top, the correction looks severe. Measured from January 1, gold was down only about 7% year-to-date as of July 20, because the record was set inside the first month of the year and everything since has been the unwinding of a parabolic move rather than a destruction of the underlying trend.
Quarter-to-date, the second quarter delivered gold’s worst quarterly decline since 2013. That is the statistic that captures how violent the repricing has been.
The cause is well documented and largely singular. The escalation of the Middle East conflict pushed energy prices sharply higher, which lifted inflation expectations, which forced a hawkish recalibration of Federal Reserve policy that markets had not priced. Higher expected rates raised the opportunity cost of holding a non-yielding asset, and a dollar that surged to a 13-month high compounded it.
Nothing in that chain involves a deterioration in gold’s demand structure. Central bank purchasing continued through the decline. Asian physical demand set records. Mine supply rose 2% year-over-year to 1,231 tonnes in the first quarter, which is a supply response to high prices rather than a demand signal.
The reason this distinction is worth drawing carefully: corrections driven by rate expectations reverse when rate expectations reverse. Corrections driven by demand destruction do not. Gold has just experienced the first kind, and the input that caused it — energy-driven inflation — has now reversed sharply, with Brent down more than 20% from its recent peak in a matter of days.
The transmission lag is the problem. It takes quarters, not weeks.
298 Tonnes of ETF Gold Is Underwater — a Ceiling, Not a Floor
The most underappreciated constraint on any gold recovery sits inside the exchange-traded fund complex, and it is quantifiable.
Roughly 298 tonnes of gold held inside ETFs is currently priced below its holders’ average cost basis at gold near $4,000 an ounce. That is up from 270 tonnes when gold traded above $4,250. At current prices, it represents approximately $38 billion of metal held by investors whose rational response to any recovery is to exit near breakeven rather than hold for further upside.
This is a structural ceiling rather than a floor, and the mechanism is worth stating precisely. Gold ETFs are physically backed. When investors redeem, the fund sells metal into the spot market. A sustained wave of redemptions is not merely a repositioning event — it is supply-generative. Every dollar of price recovery toward those entry points brings a tranche of that 298 tonnes closer to a rational exit.
The flow record shows how the overhang built. Global gold ETFs recorded $18.7 billion in net inflows during January, roughly 120 tonnes, the strongest single month ever recorded, with assets under management reaching $669 billion after a 20% single-month gain. Asia accounted for 51% of that at $9.6 billion, with mainland China alone contributing $6 billion.
Then the conflict hit. North America posted its largest monthly outflow on record at $13 billion, ending a nine-month inflow streak. US-listed funds recorded record outflows of 85 tonnes in March, erasing the 69 tonnes that had flowed in earlier in the quarter. May produced net outflows of 16 tonnes globally, with redemptions continuing into the first half of June before a $1.1 billion inflow week broke a four-week losing run.
The offsetting consideration is positioning. Global gold ETF holdings remain well below their pandemic-era peak despite the entire 2025 accumulation cycle and the second-quarter selloff. Institutional positioning is not stretched. That means the rebound, when the catalyst arrives, has room to run rather than needing to absorb a crowded long.
Both facts are true simultaneously: there is a 298-tonne supply overhang above current prices, and there is unused capacity above that.
Asian Buyers Absorbed What Western Funds Sold
The regional split inside 2026’s flow data is the most important structural development in the gold market, and it changes who sets the marginal price.
Physically backed gold ETFs recorded roughly $8 billion in global inflows across the first half of 2026 despite the price collapse. Asian funds alone drew a record $12 billion. The arithmetic means Western redemptions were substantial and were more than offset by Eastern accumulation.
The physical market tells the same story with more force. First-quarter bar and coin demand rose 42% year-over-year to 474 tonnes. Mainland China surged 67% to a record 207 tonnes. India posted its strongest first quarter since 2013, up 34% to 62 tonnes. Those are buyers accumulating physical metal into a falling price while Western funds trimmed paper exposure.
Same metal, same price, opposite decisions.
The structural read is that Asian consumers and emerging-market central banks are becoming the dominant drivers of global gold demand rather than Western institutional allocators. That is a regime change in price formation, and it has consequences for how the market behaves.
Asian physical buyers are price-sensitive in the opposite direction from Western funds — they buy dips as an accumulation opportunity rather than selling them as a stop-out. A market whose marginal buyer strengthens on weakness has a different volatility profile than one whose marginal buyer capitulates on weakness. It falls more slowly and rises less explosively.
The forward view supports continuation. Bar and coin demand is expected to feature more heavily through the remainder of 2026 as high prices, a lack of viable alternative investments in some markets, inflation fears, and elevated uncertainty attract both savers and speculators. Asian demand is projected to remain the key source of investment strength.
Jewellery is the offsetting weakness. Tonnage demand continues to slip as record prices and regional tax policy bite, though value-terms spending has held up as higher prices offset weaker volumes. China’s VAT policy change remains an obstacle for jewellery specifically, funnelling purchases toward lower-premium bars and coins — a pattern also increasingly visible in India.
Firm physical demand from Asia was cited as a supporting factor for gold Tuesday even as the dollar bid overwhelmed it.
