Gold Slides To $4,378 After A 9.6% August As Bear Market Confirms Below $4,481.78 — $4,530 Or $4,225 Next

Gold Slides To $4,378 After A 9.6% August As Bear Market Confirms Below $4,481.78 — $4,530 Or $4,225 Next

December futures (GC=F) fell $55.40 to $4,426 while silver dropped 2.69% to $64.76 and the gold/silver ratio widened to 67.55 | That's TradingNEWs

Itai Smidt 9/1/2026 12:06:14 PM
Commodities GOLD XAU/USD XAU USD

Key Points

  • Spot gold hit a two-week low at $4,375, a 7% drop from last week's $4,700 peak.
  • A second close below $4,481.78 confirms a bear market 20% under the $5,602.23 January record.
  • Fed funds futures price 67% odds of a September 16 hike and 60bp of tightening over 12 months.

Spot gold (XAU/USD) cut through support on Tuesday and printed a fresh two-week low at $4,375, trading $4,378 into the European afternoon. December futures (GC=F) sit at $4,426.10, down $55.40 or 1.24%, after opening the session at $4,498.70 — a 0.4% gap higher that lasted a matter of hours before sellers took it apart. Yahoo's spot quote had the metal at $4,432.20 at 7:56 a.m. ET before the break lower.

That is a 7% decline from last week's peak near $4,700 in four sessions.

The cause is a single repricing. CME FedWatch puts a 25 basis point hike at the September 15-16 FOMC meeting at 66.4%, with 33.6% on a hold. One week ago the split was 39.6% hike and 60.4% hold. Fed funds futures now imply 67% odds of a move on September 16 and 60 basis points of tightening over the next twelve months, against a current target range of 3.50% to 3.75%.

Silver is taking it harder. XAG/USD trades $64.76, down 2.69% from Monday's $66.55, and is now down 8.90% year to date. December silver futures (SI=F) opened at $67.28 and slid to $65.37 by 8:08 a.m. ET. The gold/silver ratio climbed to 67.55 from 66.84 on Monday — the high-beta metal underperforming, which is what happens when the driver is rates rather than inflation fear.

The tension in this tape is the whole story. Gold closed August up 9.6%, its best month since January. It also confirmed a bear market during the same stretch: Monday's settlement at $4,449.24 was the second consecutive close below $4,481.78, the level 20% under the record high of $5,602.23 set on January 29, 2026. Year to date, gold is up 0.97%. Over twelve months it is up 27.16%.

A metal that gained 9.6% in a month and is flat on the year is a metal that spent 2026 giving back a parabolic January. The August rally was a position for a Federal Reserve hold. That position died in one speech at Jackson Hole, and what is left underneath is the level at $4,315 to $4,330.

Monday's $4,449.24 Close And The 20% Line At $4,481.78

Spot gold settled Monday at $4,449.24, down $5.75 or 0.13%. The session low was $4,396.52, the lowest print since mid-August, and the metal recovered $52.72 into the close. Friday's decline was the largest single-day drop since June 10.

The number that matters from that close is $4,481.78. That is the price sitting exactly 20% below the January 29 record at $5,602.23, and Monday's settlement was the second consecutive close beneath it. Two closes below a 20% drawdown line is the standard technical definition of a confirmed bear market, and gold has now met it while simultaneously posting its strongest month since January.

Both facts are true and they describe different time horizons. The monthly gain describes a bounce from near $4,000 at the start of August. The bear market describes a seven-month distribution from $5,602.23. Traders positioning off the monthly candle are trading the bounce. Traders positioning off the drawdown are trading the trend. The 200-day simple moving average at $4,530 is where those two views meet, and price is currently $152 below it.

The 52-week range frames the scale of what has happened. The low is $3,579.72, the high is $5,602.23 — a $2,022.51 band, or 56% of the low. Gold has traded a full bull-to-bear cycle inside twelve months. At $4,378 it sits 39.4% above the 52-week low and 21.8% below the 52-week high, close to the midpoint of a range that has no obvious anchor other than the moving averages running through it.

The performance table sharpens the picture. Down 1.74% on the day. Down 5.25% on the week. Up 6.70% on the month. Up 0.97% year to date. Up 27.16% over twelve months. Every timeframe shorter than a month is negative and every timeframe longer is positive — the signature of a bull market that stopped making highs in January and has been converting momentum into range ever since.

For a forecast, that structure argues against extrapolating either the August rally or the January peak. It argues for trading the levels.

The 7% Round Trip From $4,700 In Four Sessions

Gold rose nearly 14% in a three-week rally through late August, reaching its highest level since mid-May near $4,700. It has given back 7% of that in four sessions.

The sequence is precise. The rally topped near $4,700 last week. Friday delivered the largest single-day drop since June 10 following the Jackson Hole keynote. Monday opened flat, fell to $4,396.52, and clawed back to settle at $4,449.24. Tuesday opened December futures 0.4% higher at $4,498.70, failed immediately, and drove spot to $4,375.

Round-trip moves of that speed tell you what kind of buying produced the original advance. A 14% three-week rally that surrenders half its gain in four days was not accumulation. It was positioning for a policy outcome that did not arrive. When the outcome flipped, the position unwound at the pace it was put on.

What drove the August advance was a specific set of data. Three prints in one week — jobs, CPI and PPI — all came in soft, and September hike odds fell from 50% to 31%. Gold rallied 10% from near $4,000 on that repricing. Treasury's expansion of long-dated bond buybacks added a second leg to the argument by capping long-end yields.

Both supports have now reversed. Hike odds went from 31% to 66.4%. And the Treasury buyback effect has faded — 30-year yields have already recovered around two-thirds of the decline that intervention produced, with the long bond back at levels last seen in 2007.

The metal is trading exactly as it should against that. Gold has no yield. Its entire valuation rests on the opportunity cost of the alternative. When the alternative repriced from 3.50%-3.75% and falling to 3.50%-3.75% and rising, with 60 basis points of tightening priced over twelve months, the discount applied to a zero-coupon store of value widened mechanically.

The important detail for the next two weeks: none of the moves in either direction have been driven by physical demand. They have been driven entirely by the rates path. That makes the September 16 decision, and the data leading into it, the only variable that matters.

Warsh, 3.7% PCE And The Speech That Repriced The Metal

The Federal Reserve chairman's Jackson Hole address on Friday, August 28 — his first major speech since taking the chair in May — is the single input that turned gold.

He put PCE inflation at 3.7% over the past year and 4.1% annualized over the past six months. He said the summer's better-than-expected readings do not tell him underlying trends have meaningfully improved. He described the 2% inflation target as firm and fixed, and stated that financial conditions are not currently restrictive. His formulation on the mandate was the line the market traded: the committee must be confident underlying inflation is moving to target clearly and at sufficient speed, otherwise the Fed has work to do.

He did not explicitly call for a hike. He did not need to. Odds moved from 31% at the start of the month to 39.6% before the speech to 66.4% today.

He followed up Monday morning at the G20 finance ministers' meeting, 100 days into his tenure, describing secular stagnation as a description of a period long past and the current environment as one of secular growth and a global investment surge driven by AI capital spending.

Inflation at 3.7%, financial conditions not restrictive, growth accelerating, labor at full employment with unemployment at 4.1% — that is a complete case for tightening, delivered in two appearances four days apart. There is no reading of that combination that is friendly to a non-yielding metal.

The dissenting view has weight. The chairman has restated the inflation mandate at every appearance, including a July pledge that the Fed will not waver on 2%, and markets read that one as insufficiently committed. What changed on Friday was tone rather than substance, and 27 percentage points of repricing came out of tone alone. Repricings built on tone can unwind on data.

The data arrives fast. JOLTS Tuesday at 7.3 million expected against 7.359 million. ISM manufacturing at 55.2 against 55.6, with prices paid at 71.2 against 71.1. The August employment report Friday at 55,000 payrolls with unemployment at 4.1%. And the August CPI on September 10, six days before the decision.

Real Yields At Multi-Decade Extremes Are Doing The Damage

The rate story hitting gold is not confined to the Federal Reserve. Every major sovereign curve broke to a multi-decade extreme on the same session.

The US 10-year Treasury reached 4.786%, the highest since January 2025. The 30-year hit its highest level since 2007. Japan's 10-year government bond struck 3.00% for the first time since 1996. Germany's 10-year Bund pushed to 3.3546%, a level last seen in 2011, with the 2-year Bund at 2.9496%. The 10-year gilt trades 5.23% after UK yields jumped 10 basis points. The French 10-year OAT sits at 4.21%.

Gold competes against all of them. A global repricing of the risk-free rate raises the carrying cost of a zero-coupon asset everywhere at once, and it does so in the currency of every major holder simultaneously. This is not a dollar-specific headwind that a weaker dollar could offset — it is a term-premium shock that hits the metal in euros, yen, sterling and dollars in parallel.

The composition of the yield move complicates the picture. Yields are rising on fiscal concern and an energy-driven inflation impulse rather than on growth. The US debt load has passed $40 trillion. Corporate issuance tied to AI infrastructure is competing for the same capital. Japan's ministries are preparing a record initial budget request. That configuration is, on a multi-year view, the debasement argument gold was built to express — every basis point of term premium demanded for holding sovereign paper is a reason to own a fixed-supply alternative.

That argument is correct on the horizon that matters least right now.

In the immediate term, the mechanism runs the other way. Nominal yields at multi-decade highs with the front end pricing 60 basis points of tightening produces rising real yields, and rising real yields are the single most reliable historical headwind for gold. Treasury's expanded long-dated buyback program was designed to cap this, and 30-year yields have already retraced two-thirds of the move that program bought.

The debasement trade pays after the rate shock is absorbed, not during it. September 16 is when the market learns which phase it is in.

Why $92 Brent Is Not Rescuing Gold This Time

Brent trades $92.04 and WTI $87.96 after two oil tankers were struck by projectiles in the Strait of Hormuz overnight, one Saudi and one South Korean-owned. Sunday brought the first publicly acknowledged US strike on Iranian positions since late July, hitting two rocket launchers on Larak Island, and Iran retaliated against American bases in Jordan. Crude is up 50% year to date.

That is a textbook safe-haven setup, and gold fell 1.74% into it.

The reason is that the war is no longer a haven story for the metal. It is an inflation story, and an inflation story is now a rate story. Every escalation in Hormuz pushes crude higher, crude pushes headline inflation higher, higher inflation pushes hike odds higher, and higher hike odds push real yields higher. The chain runs directly from tanker strikes to a lower gold price, and it has been running that way for weeks.

The transit data quantifies the supply risk. Kpler counted five commodity transits through the Strait on Monday against a 10-day average near 14, with no liquid tankers among them. Traders monitoring cargoes still put 6 million to 8 million barrels a day moving through, with Iranian channels describing traffic along an approved route as proceeding on a limited basis with vessels paying tolls. Tehran keeps the Strait closed; Washington maintains a counter-blockade of Iranian ports. Six months in, the conflict sits in a stalemate.

Brent has swung through a $17 range in August alone on stop-start peace efforts. That volatility is the reason gold cannot price the geopolitical premium cleanly — every escalation is followed by a de-escalation headline, and the metal has been whipsawed by both.

The condition under which oil becomes gold-positive again is a break in the correlation between crude and hike odds. That requires either the Fed to signal it will look through energy-driven inflation, or an escalation severe enough to threaten growth rather than just prices. Neither is in the tape today.

For now, $92 Brent is a reason to sell gold, not buy it. That is the inversion traders positioned for a haven bid have been fighting all month, and it has cost them 7% in four sessions.

Dollar Strength And A Yen Through 160 Compound The Squeeze

The dollar has recovered most of its recent pullback as policy expectations skew toward further tightening, and gold is priced in dollars.

The yen broke through 160, past the level that has previously drawn official intervention. The Treasury Secretary went further than his prior framing at the G20 in Asheville, saying he holds information the market does not and expects both the Japanese government and the Bank of Japan to take steps producing a stronger yen, after previously stating that Abenomics has run its course.

That combination — a strong dollar driven by rate differentials, plus a Japanese currency at intervention territory — is a double headwind. A rising dollar mechanically lowers the dollar price of gold for any given level of global demand. And a yen at 160 makes gold expensive for Japanese buyers in local terms, suppressing one of the physical demand channels that has historically absorbed dollar-price weakness.

The same dynamic hit Bitcoin, which faded from $79,184 to $77,767 on Tuesday, and equities, with Nasdaq 100 futures down 1.11%. Gold, Bitcoin and the Nasdaq all traded the identical intraday shape: a firm open followed by a sustained fade. Three assets with nothing structurally in common, one variable.

The cross-asset read is useful for what it rules out. If gold were falling while equities rallied, the explanation would be risk appetite rotating out of havens. If gold were falling while Bitcoin rallied, the explanation would be a shift within alternative stores of value. Everything falling together with the dollar firm and yields at multi-decade highs isolates the cause precisely: a discount-rate shock, not a preference shift.

That has a practical implication for the forecast. Gold does not need a haven bid or a demand catalyst to stabilize. It needs the rate path to stop repricing. If Friday's payroll number comes in at or below the 55,000 consensus with unemployment holding at 4.1%, hike odds compress, the dollar softens, and the metal has room to retrace toward $4,450. If payrolls beat, the sequence extends and $4,330 gets tested inside a session.

Central Banks Bought A Record 289 Tonnes In Q2 — And Proved Price-Sensitive

The structural bid under gold is sovereign, and the quarterly demand data shows both its strength and its limit.

Central bank net purchases reached 289 tonnes in Q2 2026 — the strongest second quarter on record since 2010 and the strongest quarter since Q4 2024. Buying was broad-based. Poland led with 51 tonnes, lifting first-half purchases to 82 tonnes and its stockpile to 632 tonnes against a stated 700-tonne target. The People's Bank of China added 33 tonnes, its largest quarterly increase since Q4 2023, taking H1 purchases to 40 tonnes and reported holdings to 2,346 tonnes. Uzbekistan added 16 tonnes, Kazakhstan 15, Jordan 6 and the Czech Republic 6. On the sell side, Russia reduced holdings by 22 tonnes and Turkey — the largest Q1 seller — slowed to 4 tonnes.

The revision underneath that number is the part traders should weight. The Q1 2026 estimate was cut from 244 tonnes to 57 tonnes. That is a 187-tonne downward revision, and it lands in the exact quarter when gold printed its all-time high at $5,602.23 on January 29.

Read the two quarters together and the pattern is unambiguous: central banks bought 57 tonnes when gold was at record levels and 289 tonnes after it had corrected. Total H1 2026 purchases came to 345 tonnes, below the prior year's first-half pace precisely because of that Q1 collapse.

Sovereign demand is real, it is structural, and it is price-sensitive. That is the correct framing for a forecast. It does not chase. It bids weakness.

The practical consequence is that central bank buying functions as a floor rather than a catalyst. It will absorb supply into $4,300 and below. It will not carry price back toward $5,000. Any forecast that treats reserve accumulation as an upside driver misreads what the Q1 revision demonstrated.

For the current tape, that floor sits below the market. It supports the $4,225 to $4,330 zone as a place where selling gets absorbed. It does nothing to prevent price reaching it.

ETF Outflows Of 45 Tonnes And A 942-Tonne Demand Print

The investor side of the demand equation ran the opposite direction from central banks through Q2, and the aggregate number was ugly.

Global gold demand fell to 942 tonnes in Q2 2026, the lowest level since Q3 2021. The decline came from weaker jewellery demand and investor outflows from gold-backed exchange-traded funds, which recorded 45 tonnes of net redemptions during the quarter, concentrated in North American products.

That produced the defining asymmetry of the year: sovereigns accumulating at a record quarterly pace precisely while private investors sold. Coins moved from Western ETF vehicles into official reserves at prices between roughly $4,000 and $4,600.

The ETF trend began reversing in August. The largest US gold trust recorded approximately $637 million in net inflows on August 7 alone, and the metal's 9.6% monthly gain came with participation from the fund complex rather than against it. That reversal is the piece most at risk from the rate repricing. ETF flow is the most rate-sensitive component of gold demand — it is discretionary, marked daily, and held by allocators who compare it directly against a 4.786% Treasury.

The jewellery weakness deserves separate weight. Physical consumption at these price levels has been suppressed for four consecutive quarters, and a 942-tonne total demand print is the confirmation. That removes the demand channel that historically cushioned corrections. When jewellery buyers are absent at $4,400, the marginal absorber of a $200 decline is a central bank operating on a strategic timeline rather than a price trigger — which means the absorption happens, but slowly and at a lower level.

The scenario tree from the flow side is narrow. Continued ETF creations at the August pace stabilize the metal in the $4,375 to $4,530 band into September 16. A resumption of the Q2 redemption pattern removes the discretionary bid entirely and hands the tape to the $4,310 to $4,330 support zone.

Given 67% hike odds and 60 basis points of tightening priced over twelve months, the second path carries the higher probability into the decision.

Technical Structure: RSI Under 50, MACD Below Zero, 200-Day At $4,530

The daily chart has flipped. Momentum indicators that were neutral-to-positive last week have crossed into bearish territory in four sessions.

The Relative Strength Index is testing levels below the 50 line, having sat at 54 as recently as Monday. The MACD has slipped below zero. That combination — RSI losing the midline and MACD crossing negative in the same window — is the standard confirmation that a correction has become a downtrend on the daily timeframe.

The moving average stack tells the same story. The 200-day simple moving average sits at $4,530, and price at $4,378 is $152 beneath it. The 100-day moving average sits at $4,370.48, and price is now testing it directly. The 20-day simple moving average, which had been acting as support at $4,430.23, has been broken and now sits overhead as resistance.

Losing the 100-day is the next technical event. Above it, the structure can still be characterized as a correction inside an uptrend. Below it, with the 200-day already lost and the 20-day flipped to resistance, the metal has no remaining moving average support until the lower Bollinger band near $4,136.78.

The bear market confirmation adds a layer that momentum alone does not capture. Two consecutive closes below $4,481.78 — the 20% drawdown line from the $5,602.23 record — establishes the primary trend as down. In that regime, rallies into resistance are distributions rather than accumulations, and the burden of proof sits with buyers on every attempt.

The one structural positive is that this decline has been orderly. Friday produced the largest single-day drop since June 10, but there has been no gap, no liquidation cascade, and Monday's $52.72 recovery off $4,396.52 showed dip buyers are still functioning. That is the difference between a repricing and a capitulation, and it argues for a controlled retest of support rather than a vertical break.

Trigger levels to watch on the daily close: $4,370.48 for the 100-day, $4,330 for the mid-August low band, and $4,481.78 for any reclaim of the bear market line.

Downside Map: $4,330, $4,225 And The Air Beneath

The support structure is well defined and the levels are close together, which makes them tradable.

Immediate support is the mid-August low zone at $4,310 to $4,330. That band held twice during August and produced the launch pad for the 14% three-week rally toward $4,700. It is the single most important level on the chart right now, and price at $4,378 sits $48 above it.

The 100-day moving average at $4,370.48 is the intermediate marker. Price is testing it directly at $4,375 to $4,378. A daily close beneath it removes the last moving average providing structural support and opens the mid-August band immediately.

Below $4,310, the next reference is the August 6 low near $4,225. That is a $153 drop from the mid-August band with no tested intermediate level, which makes it a fast move if $4,310 gives way on a closing basis.

Beneath $4,225, the chart thins considerably. The lower Bollinger band sits near $4,136.78 and the psychological $4,000 handle is where the August rally originated. A move to $4,000 from $4,378 is 8.6% and would put gold negative on the year, erasing the 0.97% year-to-date gain entirely.

The catalyst that produces the downside sequence is specific. A payroll print materially above 55,000 on Friday, or an August CPI on September 10 that fails to show cooling, pushes hike odds from 66.4% toward certainty and takes the 10-year through 4.80%. That configuration tests $4,310 inside a session and $4,225 within the week.

The offsetting consideration is where sovereign demand sits. Central banks bought 289 tonnes in Q2 with gold trading substantially below the January high, and the Q1 revision showed they stand aside at elevated prices. The $4,225 to $4,330 zone is where that bid is most likely to be active, which argues the decline gets absorbed there rather than accelerating through it.

Position sizing accordingly: $4,310 is a level to respect, not a level to fade blindly.

Upside Map: $4,450 Turned Resistance, $4,530 And The $4,700 Ceiling

The recovery path requires clearing three levels in sequence, and each has become harder than it was a week ago.

The first is $4,450. That area functioned as support through the second half of August and has now flipped to resistance following the break. Former support turned resistance is the most reliable rejection level on any chart, and it will test every rally attempt first. Monday's close at $4,449.24 sat exactly on it, and Tuesday's failure from $4,498.70 confirmed the flip.

Above $4,450 sits $4,481.78 — the 20% drawdown line. Reclaiming it on a closing basis would technically exit the bear market designation and is the level bulls need most, because it changes the regime rather than just the momentum reading.

The third is the 200-day simple moving average at $4,530. Price has not closed above it since the break, and a daily close through $4,530 would restore the intermediate uptrend structure and put the late-August high near $4,700 back in play. From $4,378, that is a 3.5% move to the 200-day and 7.4% to the prior peak.

Beyond $4,700, the upper Bollinger band near $4,723.68 marks where buyers historically hesitate. Anything above that reopens the path toward $5,000 and, on a longer horizon, the $5,602.23 record.

The catalyst required is the mirror of the downside case. A payroll miss on Friday, or an August CPI on September 10 showing genuine disinflation, compresses the 66.4% hike probability and softens the dollar. That sequence gets gold back to $4,450 quickly and to $4,530 if the Fed holds on September 16.

The probability weighting favors the downside for now. Fed funds futures imply 60 basis points of tightening over twelve months, not a single hike. That is a full cycle being priced, and unwinding it requires more than one soft data point. It requires a change in the inflation trajectory itself, with PCE at 3.7% year over year and Brent at $92.04 arguing against one.

Rallies into $4,450 are sales until $4,481.78 closes green.

Silver Down 2.69% And The Ratio At 67.55 Confirm A Rates Trade

Silver is the confirming instrument, and it is confirming a rate-driven decline rather than an inflation-driven one.

XAG/USD trades $64.76, down 2.69% from Monday's $66.55, and is down 8.90% year to date against gold's 0.97% gain. December futures (SI=F) opened at $67.28 and slid to $65.37 by 8:08 a.m. ET. The gold/silver ratio rose to 67.55 from 66.84 on Monday.

That widening ratio is the tell. When precious metals rally on inflation fear or currency debasement, silver outperforms — its higher beta and industrial demand leverage the move, and the ratio compresses. When they fall on rising real yields, silver underperforms because the industrial component gets hit alongside the monetary component. A ratio moving from 66.84 to 67.55 in one session while both metals fall is the signature of a discount-rate move.

Silver's own technical structure is deteriorating on the same schedule. The 20-day exponential moving average sits at $65.71 and price at $64.76 has broken beneath it. The August 9 low at $62.19 is the next major cushion, a further 4% down. On the topside, a decisive break above the June 17 high at $71.56 is what the metal needs to extend, and that is 10.5% away.

The longer arc puts the current level in context. Silver reached an all-time high of $121.67 on January 29, 2026 — the same session gold peaked at $5,602.23. At $64.76 it sits 46.8% below that record, a far deeper drawdown than gold's 21.8%. The metal is up 15.48% over the past month and 62.83% over twelve months, which is the same bull-then-bear structure gold shows, amplified.

For the gold forecast, silver's behavior narrows the scenarios. As long as the ratio is widening, the driver is rates and gold's support levels are the operative reference. A compressing ratio with both metals rising would signal the debasement bid returning, and that would be the first genuine evidence that the August rally had a foundation beyond a policy bet.

Watch $62.19 on silver alongside $4,310 on gold. They will break together or hold together.

Forecast: Bearish Into September 16 With $4,330 As The Line

Weighting the evidence produces a defined distribution into the Federal Reserve decision.

The base case is a controlled retest of $4,310 to $4,330 over the next one to two weeks, and it carries the highest probability. The reasoning is direct: fed funds futures price 67% odds of a hike on September 16 and 60 basis points of tightening over twelve months, the 10-year sits at 4.786%, the 30-year is at 2007 levels, the dollar is firm with the yen through 160, and momentum has crossed bearish on the daily chart with RSI under 50 and MACD below zero. Gold has no yield and every one of those variables raises its carrying cost.

The bear extension triggers on a daily close below $4,310. That opens the August 6 low near $4,225 with no tested intermediate support, and beneath $4,225 the chart thins toward $4,136.78 and the $4,000 handle where the August rally began. The catalyst is a payroll beat on Friday above the 55,000 consensus, or an August CPI on September 10 that fails to cool, either of which pushes hike odds toward certainty.

The bull case requires reclaiming $4,481.78 on a closing basis to exit bear market territory, then $4,530 at the 200-day simple moving average to restore the intermediate uptrend, with $4,700 as the target beyond. That path needs the September hike priced out, which means soft payrolls, soft CPI, and a hold on September 16.

The structural floor beneath all three: central banks bought 289 tonnes in Q2, the strongest second quarter since 2010, and demonstrated through the Q1 revision from 244 tonnes to 57 tonnes that they bid weakness rather than chase strength. That sovereign demand is most likely to be active in the $4,225 to $4,330 zone, which argues a decline into that band gets absorbed rather than accelerating through it.

The structural risk: gold ETFs shed 45 tonnes in Q2, total demand printed 942 tonnes — the lowest since Q3 2021 — and jewellery consumption has been suppressed at these price levels for four quarters. The discretionary bid is thin and rate-sensitive.

Verdict: bearish into September 16. Sell rallies into $4,450 with a stop above $4,481.78. Cover into $4,330 and reassess. Below $4,310 on a daily close, $4,225 is the target and the year-to-date gain of 0.97% goes with it.

That's TradingNEWS