Gold Slides to $4,365 as the 10-Year Hits 4.91% and Silver Loses 3.41%

Gold spot traded at $4,365.50 an ounce at 9:44 a.m. ET Thursday, down $35.30 on the session, after reaching an intraday high of $4,434 during the European morning. That is a $68.50 reversal inside a few hours, and the entire move happened in the twenty minutes surrounding the 8:30 a.m. ET producer price release.

The recovery attempt was real before it failed. XAU/USD opened the day back above the $4,400 handle, up roughly 0.5% in early trading, with spot printing near $4,414 and the four-hour chart showing $4,424. Buyers had been rebuilding since Wednesday's 1.37% advance, which snapped a three-session losing run. By 10:36 a.m. ET the metal had stabilized at $4,381.32, roughly $53 below the high.

December COMEX futures told the same story with more precision. The contract traded $4,427.40 in early trade for a 0.75% loss, then extended to $4,415.20 by mid-morning — down $45.50, or 1.02%, on the day. Tuesday's session had already knocked $77.20 off the contract, a 1.72% decline that settled it at $4,400 and pushed it further from the $4,474 level that has been acting as the ceiling.

Silver took the harder hit. Futures were down 3.41% at $66.31 an ounce in early trading, after spot had held $67.40 overnight. That puts the gold/silver ratio near 65.8, toward the low end of a range that has swung from 61.7 three months ago to 70.4 five weeks ago and back.

The dollar index sat near 98.90 after touching a session low of 98.71, and the 10-year Treasury yield climbed 8 basis points to 4.91%, its highest since 2023. Both moves came after the data, and both are direct headwinds for a non-yielding asset.

Gold has now traded either side of $4,400 for the better part of three weeks without resolving in either direction. The metal is up 21.36% over twelve months and up just 1.67% year to date — a gap that captures exactly what happened to this market in 2026.

A 5.4% Inflation Print Sent Gold Lower, Which Is the Whole Story

The Bureau of Labor Statistics released August producer prices at 8:30 a.m. ET. Headline PPI rose 0.4% month over month, matching consensus and following a flat 0.0% July reading. The annual figure came in at 5.4%, up from 4.8% in July and above the 5.3% economists had penciled in. Core PPI was forecast at 0.3% monthly and 4.6% annually, up from 0.2% and 4.2%.

Gold fell on the news. A supposedly inflation-hedging asset dropped roughly $50 on confirmation that wholesale inflation accelerated 60 basis points in a single month. That inversion is the defining feature of this market and the single most important thing for anyone forecasting the metal to understand.

The mechanism is not complicated. Gold pays nothing. Its price is a function of real yields — nominal yields minus expected inflation — and of the dollar. When an inflation print raises the probability that the Federal Reserve tightens, nominal yields rise faster than inflation expectations, real yields rise, and gold falls. The inflation data is not a signal about the purchasing power of money. It is a signal about the Fed's reaction function.

Weekly initial jobless claims printed 206,000 against 205,000 expected. That 1,000-claim miss changed nothing about the labor-market picture and removed no argument for tightening.

The full PPI release shows energy pass-through doing most of the work. West Texas Intermediate touched $100.10 a barrel Thursday, up 4.2%, with Brent at $105.37 — both at their highest since May on Strait of Hormuz disruption. Crude has climbed more than 5% in a month, and it feeds producer prices with a short lag.

That leaves gold in an awkward position. The oil shock creates the inflation that historically drove bullion higher, and simultaneously creates the central-bank response that drives it lower. In 2026 the second effect has been winning by a wide margin, and Thursday's tape is the cleanest demonstration of it so far this quarter.

Friday's consumer price index — consensus 0.4% monthly, 3.4% annually, with core seen at 2.4% — will either confirm or break that pattern.

The $4,474 Ceiling Has Rejected Every Rally Since Late August

The technical structure is a series of lower highs beneath a well-defined resistance shelf, and the sequence has been tightening.

December gold futures peaked near $4,700 on August 25 before the correction began. The contract worked down to $4,474, failed there, and has not reclaimed it. Tuesday's $77.20 drop to a $4,400 settlement extended the retreat. Wednesday produced a 1.37% bounce that stalled. Thursday's $4,434 high in spot terms was another attempt at the zone, and it lasted less than three hours.

On spot, the resistance layers stack at $4,434, then $4,450, then the $4,474 futures pivot. A daily close above $4,450 would be the first structural improvement in three weeks. A close above $4,474 restores the uptrend that was interrupted in late August and puts $4,600 back on the table — the level gold cleared in late August before the Warsh repricing.

Support is equally well mapped. The immediate shelf is $4,350 to $4,365, which is where price sits now. The 100-day simple moving average sits at $4,340, and that is the level short-biased positioning is targeting on rallies. Below it, $4,300 is the larger downside marker — gold produced a false breakdown through $4,300 earlier in this correction and rebounded, which makes the level a magnet rather than a wall.

The metal remains capped below its 200-day simple moving average on the daily chart, and daily momentum reads neutral rather than directional. That combination — price beneath the 200-day, momentum flat, resistance stack intact — describes a market in a corrective consolidation, not one in a trend.

The September 1 session is the template for what a break looks like. Gold settled at $4,325 that day, down 2.86%, when hike odds spiked toward 70% on the Jackson Hole commentary. Silver fell 3.73% to $64.13 in the same session. A hot CPI on Friday reproduces that setup almost exactly, and $4,325 becomes the first destination rather than the floor.

Every level in this range is 1.5% to 3% from the last one. That is a tight book for an asset that moved $1,200 in eight months.

Real Yields Are Doing the Damage: 10-Year at 4.91%, Dollar at 98.90

Gold's problem is arithmetic, and the arithmetic got worse Thursday.

The 10-year Treasury yield rose 8 basis points to 4.91%, its highest level since 2023, after settling near 4.857% Wednesday. That move followed the Treasury Department's announcement that it would triple its buyback of longer-dated debt to $6 billion, up from an earlier plan to at least double it to $4 billion. The market read a $6 billion operation against record issuance as inadequate, and yields rose on the announcement rather than falling.

The rest of the curve is at multi-year extremes as well. The two-year sits near 4.36%, having touched a 52-week high around 4.4% after the payroll report. The 30-year is near 5.27%. A month ago the 10-year was in the 4.78% to 4.81% band. It has repriced roughly 12 basis points higher in two weeks.

The dollar index at 98.90, after a session low of 98.71, adds the second headwind. Bullion priced in dollars becomes more expensive for every non-dollar buyer when the currency firms, and physical demand from Asia and the Middle East is price-sensitive in local terms.

There is a subtlety in this session that argues gold's decline is not purely a yield story. The metal was already reversing before the dollar firmed decisively, and the dollar itself had been under pressure from a rallying Japanese yen that pushed USD/JPY to a six-month low below 153.00 earlier in the week before rebounding toward 154.00. AUD/USD sat above 0.7200, near its highest since May 14. A broadly soft dollar with gold still falling means the selling is coming from rate expectations directly rather than from currency translation.

Historical comparison sharpens it. When the 10-year was at 4.68% in mid-August, gold pushed through $4,400 in the face of that headwind — the structural bid from central bank accumulation was strong enough to override it. At 4.91% it is not. Somewhere between those two yield levels, the marginal buyer stopped showing up.

62% Odds on a September Hike, Repriced From 45% in Four Weeks

The speed of the Fed repricing explains the entire shape of gold's August and September.

A month ago the market put the probability of a September increase at roughly 45%, leaning toward a hold after a soft July jobs report. Chair Kevin Warsh's Jackson Hole address, which reaffirmed a fixed 2% inflation target, pushed the implied odds toward 70% at one point on September 1. They settled back to 58.4% by September 7, ran to roughly 60% after the payroll report, and stood at 62% Thursday ahead of the PPI release.

The payroll data was the accelerant. August nonfarm payrolls came in at 162,000 against a consensus near 56,000 — a near-threefold beat — with the unemployment rate holding steady. That removed the labor-market justification for patience and shifted the entire debate onto inflation, which is precisely the debate Thursday's PPI print and Friday's CPI settle.

The target range in question is 3.75% to 4.00%, up from the current setting. The meeting runs September 15–16, five business days out.

The counterweight is that not everyone is convinced. A poll of economists points to no further hikes through year-end, which sets up a genuine divergence between market pricing at 62% and forecaster consensus at zero. Divergences of that size resolve violently in one direction or the other, and gold is the cleanest instrument for expressing which side is right.

The European Central Bank made its own decision Thursday, raising the deposit rate 25 basis points to 2.5% from 2.25% — a move priced at 100% odds beforehand, taken explicitly because energy-driven inflation from the Iran conflict has clouded the outlook. Two major central banks tightening into an oil shock in the same week is the macro regime gold is fighting.

Bullion has settled into a narrow range since bouncing off a floor near $4,000 in July, swinging either side of $4,400 as traders repeatedly reassess. The range has held because neither side of the Fed argument has won. Friday decides it.

Silver Down 3.41% to $66.31 and a Ratio at 65.8

Silver is doing what silver does in a rate-driven selloff: falling harder and faster than gold.

Futures were down 3.41% at $66.31 an ounce in early Thursday trading, against spot near $67.40 overnight. On September 1, silver fell 3.73% to $64.13 while gold dropped 2.86% — a consistent beta of roughly 1.3 times to the yellow metal on down sessions. On Wednesday's bounce, silver rose 2.87% against gold's 1.37%, the same relationship running the other direction.

The gold/silver ratio at 65.8 sits toward the low end of a range that has been unusually wide. Twelve weeks ago the ratio was 61.7. Five weeks ago it reached 70.4. It has now covered nearly the full range twice in three months, which is a far larger round trip than the metric normally produces. Gold is the leg giving ground in the current move, which is why the ratio has compressed rather than expanded.

The drawdown math is where silver's story separates from gold's entirely. Silver's all-time high is $121.67 an ounce, set January 29, 2026. At $66.31 the metal is down 45.5% from that peak. Gold's record of $5,589.38 spot, set January 28, 2026, sits 21.9% above Thursday's $4,365.50. Silver fell roughly twice as far from the same blowoff.

Two forces explain the gap. Silver's market is a fraction of gold's, so identical investment flows move it further in percentage terms. And roughly half of silver demand is industrial — electrical, solar, electronics — which ties it to a manufacturing cycle that a Fed hiking into an oil shock does not help.

The industrial leg also links silver to the base-metals liquidation that flattened copper equities Thursday. Freeport-McMoRan fell 7.69% to $70.37 and Southern Copper 6.49% to $195.68 after reports that Washington has not decided on refined-copper tariffs. Silver caught some of that same selling, which is why a 1.02% gold decline coincided with a 3.41% silver decline rather than a 1.3% one.

For anyone using silver as a leveraged gold expression, the asymmetry is running in the wrong direction right now.

From $5,589 to $4,365: What the January Blowoff Actually Cost

Gold's twelve-month chart and its year-to-date chart tell opposite stories, and reconciling them is essential to any forecast.

The metal is up 21.36% over the past twelve months. It is up 1.67% year to date. The 52-week range spans $3,626.46 to $5,602.23 — a band 54% wide from low to high. Both statistics are accurate and they describe completely different markets.

The reconciliation is January. Gold peaked at $5,589.38 spot on January 28, 2026, with an intraday print of $5,602.23 the following day, capping a rally in which the price more than doubled from 2024 levels. That peak surpassed the inflation-adjusted 1980 record of roughly $850 nominal in real terms — the first time the real high had been taken out in more than four decades.

Then it broke. Gold fell through the spring, reached a floor near $4,000 in July after trading below that level in late June, and chopped through midsummer. The recovery from there was sharp: from below $4,000 in late June to above $4,600 by late August, a move of roughly 15% in eight weeks. The August 25 high near $4,700 was the peak of that recovery.

At $4,365.50 today, gold sits $1,224 below the January record — a 21.9% drawdown — while remaining roughly 9% above the July floor and 20% above the 52-week low of $3,626.46.

The structural arguments that drove the doubling have not gone away. Central bank accumulation has continued. De-dollarization flows remain in place. Fiscal deterioration in the United States, with a Treasury actively trying to suppress long-end yields and a $1 trillion-plus payment proposal floated this week, is exactly the backdrop that supports monetary metals over multi-year horizons.

None of that has mattered in September. A 62% probability of a rate hike has outweighed all of it, which is the honest read of a market where the structural bid sets the floor and the rate cycle sets the price inside the range.

The $4,000 Floor and Why It Still Holds the Bear Case Together

The most useful level on the gold chart is not on the screen. It is $4,000, and it has not been tested since July.

Gold traded below $4,000 in late June, established a floor near that figure through July, and has not revisited it. Every subsequent decline — the September 1 drop to $4,325, the September 8 break to $4,400 on futures, Thursday's slide to $4,365 — has stopped hundreds of dollars above it.

That matters for scaling the downside. The distance from $4,365.50 to $4,340 at the 100-day moving average is 0.6%. To $4,300 it is 1.5%. To $4,000 it is 8.4%. A hot CPI print produces the first two easily. It does not produce the third without a genuine change in the structural picture.

The reason the floor sits where it does is flow, not sentiment. Central banks have kept adding to reserves through the entire 2026 drawdown. That buying is price-insensitive and slow, and it accumulates on weakness rather than strength, which is the exact opposite behavior of the leveraged futures money that drives daily volatility.

The physical side reinforces it. Jewelry accounts for roughly half of annual gold demand globally, with the remainder split among investment bars and coins, central bank reserves, and industrial applications. Jewelry demand is price-elastic in the wrong direction for bears — it strengthens as prices fall, particularly in Asia, and it puts a bid under corrections that has nothing to do with Fed policy.

Where the bears have the stronger argument is the composition of the 2026 selloff. Gold falling from $5,589 to $4,000 in six months was not a physical event. It was a positioning unwind, and positioning unwinds do not respect fundamental floors when they accelerate. The July low proved there was a level where physical demand and central bank buying stopped the slide. It did not prove that level cannot be tested again.

For now, the range is $4,300 to $4,474 with $4,000 as the structural backstop. Nothing in Thursday's data changed the outer boundaries.

Newmont at $127.90 With a $1,621 Cost Base and Record Free Cash Flow

The miners are where gold's 2026 story looks least like a bear market, and Newmont is the cleanest illustration.

Newmont (NEM) traded at $127.90 mid-session Thursday, up 0.66% from Wednesday's $127.06 close, inside a range of $127.55 to $130.48 on 7.58 million shares against an 8.68 million average. Market capitalization stands at $135.66 billion with a trailing P/E of 16.05 and a dividend yield of 0.80%. The 52-week range runs $76.05 to $135.29 — the stock is 5.5% off its high and up 65.72% over twelve months, with a 14.99% gain in the past month alone.

The margin structure explains it. Newmont's second-quarter realized gold price was $4,414 an ounce against byproduct all-in sustaining costs of $1,621, producing a record $2.2 billion in quarterly free cash flow on a 33% year-over-year gain in realized price. The company expects to sell roughly 5.3 million ounces of gold in 2026 from continuing mines and remains inside full-year guidance. Agnico Eagle generated $1.3 billion of second-quarter free cash flow at $1,459 AISC.

Run the sensitivity. At today's $4,365 spot, Newmont's margin per ounce is roughly $2,744 — down about $49 from the Q2 realized price, or 1.8%. At $4,000 gold, the level of the July floor, the margin compresses to $2,379, a 13% reduction. Even a full retest of the summer low leaves the world's largest producer generating enormous cash at a valuation of 16 times earnings.

The risk sits in the cost line rather than the revenue line. WTI at $99.35 and Brent at $105.37 feed directly into diesel, explosives and haulage costs, and mining is one of the most energy-intensive industries in the S&P 500. A sustained $100 crude environment raises AISC across the sector at the same moment the metal is range-bound. The combination of $4,000 gold and WTI above $95 would erase a meaningful portion of the margin expansion story that has driven the equities.

Barrick has delayed the spinoff of its American gold business, removing a catalyst that had been anticipated. Newmont's shareholder litigation over Newcrest disclosures was dismissed in late August.

GDX at $100 After a 33% Month Shows Leverage Cutting Both Ways

The VanEck Gold Miners ETF (GDX) traded around the $100 level, having opened Wednesday at $100.35 with a high of $101.48 and a low of $98.69 on 9.77 million shares against a 24.25 million three-month average. Assets under management stand at roughly $29.95 billion across 64 holdings at a fund-level P/E of 16.44 and a 30-day yield of 0.38%.

The concentration is heavy at the top. Newmont accounts for 10.79% of assets, Agnico Eagle 10.63%, Barrick Mining 7.32%, Wheaton Precious Metals 5.84% and AngloGold Ashanti 5.14%. Franco-Nevada at 4.71%, Kinross at 4.34%, Gold Fields at 4.34%, Pan American Silver at 2.57% and Northern Star at 2.48% round out the top ten. Those ten names carry roughly 58% of the fund. When Newmont and Agnico beat, GDX moves. When one stumbles, the concentration works against holders.

The trailing performance captures the operational leverage precisely. GDX advanced roughly 33% over a trailing 30-day window in August, rising from about $74 to approximately $98.50 as gold climbed from below $4,000 in late June to above $4,600 by late August. Over twelve months the fund has returned between 53% and 74% depending on the measurement window, against gold spot's 21.36%. Over ten years the gap is 317% for the miners against 253% for bullion.

The other side of that leverage showed up earlier this year. GDX has spent stretches of 2026 down year to date while bullion held gains, and in one month-long window the fund fell 10% against a 5% drawdown in the metal. Miners have largely fixed costs to extract ore, so every dollar above AISC drops to margin — and every dollar below the last price drops out of it just as fast.

The junior complex amplifies further. GDXJ carries roughly $9.8 billion in assets across 116 holdings at a 0.52% expense ratio, with exploration-stage names that lack the balance sheets to absorb a sustained pullback.

Miners historically lever gold moves roughly two to one. At $4,365 with the metal capped below its 200-day average, that leverage is a liability rather than an asset until the $4,474 ceiling breaks.

Three CPI Scenarios and What Each Does to the Level Map

Friday's consumer price index at 8:30 a.m. ET is the last major data point before the September 15–16 decision, and gold's next $200 is entirely contingent on it.

Consensus calls for headline CPI at 0.4% monthly and 3.4% year over year, with core expected to fall to 2.4%. Core is the swing variable. The Fed targets it, and a 2.4% print would be the strongest disinflation evidence of the quarter.

Scenario one: core at or below 2.3% with headline in line. Hike odds collapse from 62% toward 35%, the 10-year backs off 4.91%, the dollar softens through 98.71, and gold rallies hard. The metal has been suppressed by rate expectations rather than by any deterioration in physical demand, so the unwind would be sharp. First target $4,434, then $4,474. A close above $4,474 opens $4,600 and puts the August high near $4,700 in play within two weeks. Probability: roughly 25%.

Scenario two: core at 2.4% to 2.5% with headline at 0.4%. Nothing resolves. Hike odds stay near 60%, the range holds, and gold chops between $4,340 and $4,450 into the Fed decision itself. The meeting becomes the resolution point rather than the data. Probability: roughly 40%.

Scenario three: core above 2.6% or headline above 0.5%. Hike odds run toward 80%, the 10-year pushes through 5.00%, and gold breaks the 100-day at $4,340 immediately. First stop $4,325 — the September 1 settlement. Then $4,300, where the earlier false breakdown occurred. A daily close below $4,300 on volume targets $4,200 and re-opens the conversation about the July floor near $4,000. Probability: roughly 35%.

Weighting those outcomes gives a modestly negative skew, driven entirely by the energy component. WTI at $99.35 and Brent at $105.37 pass into headline CPI through gasoline and utilities within weeks, and both benchmarks are at four-month highs. The scenario that hurts gold most is the one crude is actively manufacturing.

Silver amplifies each outcome by roughly 1.3 times. Miners amplify by roughly two.

Verdict and Forecast: Range-Bound With a Downward Bias Until the Fed Speaks

Gold at $4,365.50, down $35.30 with December futures off 1.02% at $4,415.20, is not a broken market. It is a market that has stopped trading its own fundamentals and started trading the Federal Reserve's reaction function, and the two are pointing in opposite directions.

The evidence is Thursday's session itself. Wholesale inflation accelerated to 5.4% annually, its fastest pace of the cycle, driven by crude at $100. Gold reversed $68.50 off its intraday high on the news. An asset that rallies on inflation does not do that. An asset priced off real yields does exactly that, every time, and it will keep doing it until either the Fed stops or inflation expectations reprice faster than nominal yields.

That framing produces a clear near-term forecast. The range is $4,300 to $4,474 and it holds until Friday morning. The bias inside that range is lower, because the 100-day moving average at $4,340 sits 0.6% below spot, the 200-day caps price overhead, momentum reads neutral, the dollar is firming at 98.90, and the 10-year is at a three-year high of 4.91% with 62% odds on a hike five business days out. The path of least resistance is a test of $4,340 and then $4,325.

The medium-term picture is the opposite, and both can be true. Central banks have accumulated through the entire 2026 drawdown. The metal held $4,000 in July when positioning was far worse than it is now. Newmont is generating record free cash flow at $1,621 AISC with the stock at 16 times earnings and 5.5% off its 52-week high. Miners do not price like that when the underlying commodity is entering a bear market. Gold at $4,365 is 21.9% below a January record that was itself a blowoff, and the correction has already run seven months.

The trade, therefore, is patience rather than direction. Below $4,300 on a hot CPI, gold becomes a better buy than it is today, with $4,200 and the $4,000 structural floor as the accumulation band. Above $4,474 on a soft core print, the August high near $4,700 is the objective and the miners deliver two-to-one on the move. Between those two levels — which is where the metal has lived for three weeks — there is no edge, and Friday at 8:30 a.m. ET is when that changes.