Gold at $4,193 Fights to Hold the $4,139 Double Bottom as Newmont NEM Drops 4% and Gold Fields GFI Craters 12.9%

Gold futures for December delivery traded at $4,193.20 an ounce late Tuesday morning, up $24.80 or 0.59%, after opening at $4,150.10, down 0.4% from Monday's close. Earlier in the U.S. session the contract printed $4,189 and $4,190.60. Spot gold was quoted at $4,148.65 at 6:32 a.m. New York time, up $21.47 from Monday's $4,127.18 close. The bounce is real and it is small. It sits on top of Monday's collapse, when spot lost $144.60 in a single session to $4,141.30, the lowest level since August 5, with a 24-hour range from $4,139.69 to $4,287.41. That is a $147.72 intraday swing, the metal's worst single day in weeks.

The numbers around the price are worse than the price. Gold is down 4.23% over the past week and 6.65% over the past month. Its year-over-year gain has shrunk to 9.5%, the lowest annual growth rate since daily tracking of this metric began, and one CFD benchmark has it at 7.59%. The all-time intraday high of $5,597.23 was set on January 29, 2026. Tuesday's price is $1,404 below that, a 25.1% drawdown. The 52-week low is $3,600, set in the fall of 2025 before the parabolic run into January. Gold has now given back more than a third of the move from that low to the peak and is trading closer to the low than the high on a percentage basis.

The proximate cause is the bond market. The 10-year Treasury yield traded at 5.264% Tuesday, up 2 basis points, near levels not seen since 2007. The 30-year sat at 5.589%. Monday's intraday high on the 10-year was 5.27%. The Federal Reserve raised the federal funds target range to 3.75% to 4.00% on September 16, its first hike in three years, and the market now prices roughly a 70% probability of a second hike in October. A 5.26% nominal yield against a Fed PCE inflation projection of 3.7% for 2026 is a real yield above 1.5%, and every basis point of real yield is a basis point of opportunity cost on an asset that pays nothing.

The thesis for this forecast: gold at $4,193 is a metal that has lost its rate-cut premium, not its structural bid. Central banks bought a record 288.9 tonnes in the second quarter, ETFs took nearly $2 billion in the first week of September, and the sell-side year-end target of $4,900 has not been cut. The question is whether the $4,150 low holds through Wednesday's PCE and Friday's payrolls, because if it does, the reflexive bounce becomes a base. If it does not, $4,000 is the next stop and the structural bid gets tested at a level that has not traded since March.

Monday's $145 Crash: Oil, Yields, and the Iran Headline That Broke the Range

Monday's session is the one to understand because it set the level gold is now defending. Spot gold opened near $4,285, drifted through the London session, and then dropped $144.60 in a matter of hours to $4,141.30 by 9:10 p.m. Hanoi time, 10:10 a.m. New York. The move started when Brent crude surged past $107 on reports that Iranian officials doubted a deal could be reached before the U.S. midterm elections in November, following the President's rejection of Tehran's latest proposal to reopen the Strait of Hormuz. Higher oil revived inflation concerns, the 10-year pushed to a 5.27% intraday high, and gold got sold as the market repriced the October Fed meeting from a coin flip to a likely hike.

The mechanics matter. Gold had been trading in a $4,250 to $4,350 range for most of September, having built what looked like a floor at $4,300 to $4,400 on record central bank demand and nearly $2 billion in ETF inflows ahead of the September 16 Fed decision. That floor was roughly 22% below the January peak, and the argument for it was that the Fed hike was fully priced. Monday proved the hike was priced but a second hike was not. The 4% drop was the market discovering that the October meeting is live, and that gold at $4,300 had been assuming the Fed was one and done.

Silver went with it. Silver futures traded at $61.24 Tuesday, down 0.78%, after a heavier Monday decline that took the gold-silver ratio to 68.5. Silver has an industrial component tied to electronics and solar demand, and the same OpenAI training pause that hit semiconductor stocks Monday reads through to silver's industrial bid. Platinum and palladium traded lower in sympathy.

The dollar was flat Tuesday, which is the one piece of good news. Monday's gold selloff came with a firm dollar and yields near 19-year highs, a combination that has historically produced the worst outcomes for bullion. Tuesday's stabilization came with WTI falling 1.79% to $90.94 and Brent dropping to $103.97 on renewed back-channel talks between Washington and Tehran. Gold rose 0.59% because oil fell and the inflation-hike-yield chain loosened by one link. That is the tell: gold is trading as a rates instrument right now, and rates are trading as an oil instrument. Reopen the strait and gold rallies. Escalate and gold breaks $4,100.

The Rate Shock in Full: 5.264% Ten-Year, 70% October Hike Odds, and 4.1% Fed Funds Through 2027

The single largest input to the gold price today is the path of the federal funds rate, and that path just got steeper. On September 16 the FOMC voted 12-0 to raise the target range 25 basis points to 3.75% to 4.00%. The Summary of Economic Projections showed 16 of 18 participants expecting at least one more hike this year. The median projection for the end of 2026 rose to 4.1% from 3.8% in June, and the 2027 projection rose to 4.1% from 3.6%. Eight participants project further hikes in 2027, six see rates unchanged, four see cuts. The FOMC does not expect PCE inflation to return to 2% until 2029.

Market pricing has moved ahead of the dots. Immediately after the September 16 press conference, the CME FedWatch tool put October hike odds at 49%. By Tuesday morning, after Monday's oil spike and yield surge, that probability was roughly 70%. The chance of at least one more hike by year-end was 87% two weeks ago and is higher now. Every increment in that probability is a direct hit to gold, because gold's fair value on a rates-only model is a function of the expected real yield over the next two years, and that expected real yield has risen more than 50 basis points in six weeks.

The Treasury curve has done the rest. The 5-year yield crossed 5% for the first time since 2007 last week. The 10-year at 5.264% and the 30-year at 5.589% are multiyear highs. The iShares 20+ Year Treasury Bond ETF TLT fell more than 1% Monday. Global government bonds are in a rout that some of the largest fixed income managers now call overdone, but a bond market that is overdone is still a bond market that is selling, and until it stops selling the real yield keeps rising.

The counterargument is that the Fed's own projections make the bull case for gold on a longer horizon. A 4.1% policy rate through 2027 with PCE inflation at 3.7% in 2026 and 2.3% in 2027 is a real policy rate of 0.4% this year rising to 1.8% next year. That is restrictive but not punishing, and a central bank that admits it cannot get inflation to target for three more years is a central bank that is tolerating above-target inflation, which is the environment in which gold made every one of its major bull runs. The Fed hiked in September 2026 for the same reason it hiked in 1979: energy-driven inflation it cannot control with domestic policy. Gold's 25% drawdown from the January peak is the market pricing the hikes. Gold's 9.5% year-over-year gain is the market refusing to price the inflation away.

Central Banks: 288.9 Tonnes in Q2, China's 21-Month Streak, and the $4,900 Year-End Target That Has Not Moved

The structural bid underneath gold has not changed in the past two weeks, and it is worth setting out in full because it is the reason the sell-side has not cut its targets. Central banks purchased a record 288.9 tonnes in the second quarter of 2026, a 62% to 74% increase year over year, buying aggressively into the price decline from the January peak. China extended its official buying streak to 21 consecutive months through July, and gold now accounts for 8% of China's foreign exchange reserves, up from 5.5% a year ago. The Bank of England's gold holdings on behalf of foreign central banks rose 63 tonnes in July, far exceeding the decline in the New York Fed's custodial holdings, which points to sovereign accumulation that does not show up in official reports.

The largest sell-side house on the metal estimates central bank purchases of 44 tonnes in July and puts the three-month seasonally adjusted trend at 91 tonnes per month. Its year-end 2026 fair-value forecast is $4,900 per ounce, assuming average central bank purchases of 50 tonnes per month this year and 40 tonnes per month in 2027, alongside a recovery in Western ETF demand. That forecast was reaffirmed on September 23, five days before Monday's crash, with an explicit warning of "greater two-sided volatility" along the way. At $4,193 the metal is 16.9% below that target with three months to go.

The bear read on central bank demand is that the headline pace cooled in the first quarter, when net reported purchases were only 16 tonnes and central banks sold 129 tonnes on a gross basis, headlined by Türkiye's 60-tonne sale in March. The second-quarter rebound to 288.9 tonnes suggests Q1 was the aberration, and a record 45% of central banks surveyed plan to increase gold reserves in the next 12 months. Central banks hold roughly 29,500 tonnes, 17.2% of all above-ground gold. Gold now accounts for a larger share of central bank reserves than U.S. Treasuries for the first time since 1996. Nearly $14 trillion is invested in gold across official holdings, bars, coins, ETFs and over-the-counter positions.

The mechanism is simple and it is why the structural bull case survives a 25% drawdown. A central bank does not buy on Monday and sell on Friday. Every tonne that goes into a sovereign vault comes off the market permanently, and 91 tonnes a month is roughly 3% of annual mine supply being removed every 30 days. China's September announcement that it will serve as custodian for foreign sovereign gold reserves is designed to create new emerging-market demand rather than shift existing holdings from London or New York. The World Gold Council's third-quarter central bank purchase report is due in late October. If it shows another 250-tonne-plus quarter, the $4,150 floor holds. If it shows a return to the Q1 pace, the floor does not exist.

ETF Flows: $2 Billion Into GLD and GLDM in Five Days, and Whether It Survives the Hike

The Western investor bid was the missing piece of the gold rally for most of 2025, and it arrived in force in early September. SPDR Gold Shares GLD took in $1,378 million over five trading days ending September 9, and SPDR Gold MiniShares GLDM added $590 million, combined inflows approaching $2 billion in the week before the Fed decision. GLD holds 1,167 tonnes, the largest private gold hoard on earth and the third-largest of any kind behind the U.S. and German central banks. iShares Gold Trust IAU holds 523 tonnes. The COMEX Gold Trust holds 440 tonnes. Together the top ten ETFs and closed-end funds control more than 3,000 tonnes.

Gold ETF holdings and the gold price moved together for five months into September, and then decoupled: holdings kept rising as price pulled back from $4,400 toward $4,150. That divergence is either the most bullish thing on the chart or the most dangerous, depending on which way it resolves. If ETF investors are buying the dip because they believe in the $4,900 target, then Monday's crash was a transfer of gold from leveraged futures longs to unleveraged ETF holders, which is structurally healthy. If ETF investors are late to a trade that has already topped, then the $2 billion that came in at $4,300 is now 3.5% underwater and will become supply on any bounce toward breakeven.

The pre-hike inflow was the bet that the September 16 decision was one and done. That bet is now being tested by 70% October hike odds. The key variable is whether ETF inflows hold after the hike, and the honest answer is that the data since September 16 is mixed: holdings have not fallen, but the pace of creation has slowed, and Monday's 4% drop is the kind of session that produces redemptions two days later once retail sees the statement. Watch the GLD holdings print through Friday. A decline of more than 10 tonnes would confirm the Western bid is cracking. Flat or rising holdings through the PCE and payrolls prints would confirm the dip is being bought.

The medium-term risk was laid out in June by one of the larger bank research desks: if the Fed's hiking cycle extends, the most likely outcome is a flip to more sustained Western ETF outflows, a persistent headwind for gold particularly if central bank buying intensity dips at the same time. That is the bear scenario in one sentence, and both of its conditions are now live. The bull scenario is the mirror image: central banks keep buying 50 to 90 tonnes a month regardless of the Fed, and the Western investor eventually recognizes that a 4.1% policy rate with 3.7% inflation is not a reason to sell an inflation hedge.

The Gold Miners Got Smoked: Newmont −4%, Barrick −4.6%, Gold Fields −12.9%

The equity side of the gold trade is where the damage is concentrated, and the numbers are brutal. Newmont NEM fell 4.04% Monday and is now testing its 50-day moving average near $116, with a decisive break below that level expected to accelerate downside into October. Newmont reported a record $2.2 billion in free cash flow last quarter, has a UBS target of $155, and reports third-quarter results on October 22, and none of that mattered Monday. Barrick Mining B dropped 4.6% to $40.92, with its market cap at $67.7 billion, despite resolving a labor dispute at its Loulo-Gounkoto mine in Mali the same day. Barrick has been trading below its 20-day and 200-day averages on the Toronto listing and is in the middle of a restructuring that includes a planned IPO of its North American assets.

The mid-tier and South African names were worse. Gold Fields GFI collapsed 12.88% to $35.18, down $5.20, on 7.08 million shares against a 3.66 million average, taking its market cap to $31.3 billion. Coeur Mining CDE fell 7.02% to $17.87 on 36.8 million shares. Aris Mining ARIS dropped 7.55% to $17.50. Perpetua Resources PPTA lost 7.11% to $21.82. The VanEck Gold Miners ETF GDX tracked the group lower. For context, Gold Fields' 12.88% drop on a 4% gold move is a beta of more than 3, which is what you get when a miner's all-in sustaining cost is $1,800 an ounce and the metal drops $145: the margin compression is a much larger percentage of the profit than the price move is of the revenue.

There is a second layer to the miner selloff that is not about gold. The same yield shock that hit bullion hit every leveraged equity, and gold miners are capital-intensive businesses with floating-rate debt and multi-year development pipelines that get discounted at the 10-year. Monday's NYSE breadth was 3.55 decliners for every advancer. Gold Fields and Coeur fell alongside Modine, Aurora Innovation, Cleveland-Cliffs and Archer Aviation, none of which have anything to do with gold. The miners were sold as high-beta small and mid-caps in a risk-off tape, and the gold move was the excuse.

The miner-to-metal divergence is now the widest of the year, and it cuts both ways for the forecast. If gold holds $4,150 and bounces to $4,300, the miners bounce 8% to 15% because the same leverage works in reverse. If gold breaks $4,100, Newmont loses the 50-day at $116 and the group has another 10% of downside. Newmont's October 22 earnings call is the next fundamental catalyst; a realized price near $4,250 for the quarter against $4,600 in Q2 will compress margins but still produce free cash flow above $1.5 billion. The market is not pricing that. It is pricing $4,000 gold and a second hike.

The Technical Map: $4,139 Is the Line, $4,000 Below, $4,300 and $4,400 Above

The chart is now clean because Monday's crash cleared out three weeks of congestion. Support first. $4,150 is the Monday low on a closing basis and the September 29 open; it is also the August 5 low, which makes it a double bottom on the daily if it holds. $4,139.69 is the intraday low from the 24-hour window into Tuesday's open and the exact number that has to hold for the double-bottom read. $4,100 is the round number and the level at which the two-month range fully breaks. $4,000 is the psychological floor and the level the sell-side used as its mid-2026 base case before the January spike rendered it obsolete; it has not traded since March. Below $4,000 the next reference is the $3,600 52-week low, which would be a 36% drawdown from the January peak and the kind of correction that ends bull markets.

Resistance next. $4,250 is the lower bound of the September range and the first level a bounce has to reclaim to invalidate Monday's breakdown. $4,287.41 is the 24-hour high into Tuesday and the level from which the crash started. $4,300 to $4,400 is the zone that the structural bulls called a floor two weeks ago and that has now become a ceiling; the ETF money that came in during the first week of September has its average entry in this band. $4,600 is where the metal traded on August 25 when the $4,900 year-end target was published, and it is the level that would need to be reclaimed for that target to be plausible by December. $4,900 is the target itself, 16.9% above Tuesday's price. $5,000 is the round number that most sell-side houses had as a 2026 destination in January. $5,597.23 is the all-time high.

The moving averages are the problem. Gold is below its 20-day, 50-day and 100-day averages after Monday, and the 50-day is now falling. The 200-day, somewhere near $4,450 given the parabolic first quarter, is well overhead. On the weekly chart the metal has lost the uptrend line from the October 2025 lows and is testing the 50-week average for the first time since the rally began. RSI on the daily is in the low 30s after Monday's session, which is oversold but not at the extremes that mark durable bottoms; the June low in Bitcoin and the April 2025 low in gold both came with daily RSI under 25.

The pattern that matters is the double bottom at $4,139 to $4,150. If it holds through Friday's payrolls print and gold closes the week above $4,200, the target on the pattern is a measured move back to $4,400, the top of the range. If it fails, the target is $4,000 by mid-October with a probable overshoot to $3,900 on the same forced-selling dynamics that took Gold Fields down 12.88% in a day. The odds are close to even, and they will be decided by a 10:00 a.m. data print Wednesday.

Oil, Hormuz, and the Inflation Chain That Runs Straight Through Gold

Gold's correlation to oil has flipped sign this year, and understanding why is the key to the forecast. In a normal cycle higher oil means higher inflation means gold rallies as a hedge. In this cycle higher oil means higher inflation means the Fed hikes means real yields rise means gold falls. The transmission mechanism has inverted because the Fed has proved it will tighten into an energy shock rather than look through it. Monday's sequence was Brent above $107, the 10-year to 5.27%, October hike odds to 70%, gold down $145. Tuesday's sequence was WTI down 1.79% to $90.94, the 10-year flat, gold up $24.80.

The Strait of Hormuz is the fulcrum. U.S. and Iranian officials held separate indirect talks with mediators Monday, and Iran's foreign minister said Tehran expected Washington's formal response to its ceasefire proposal Tuesday. That proposal offered to reopen the strait within a week in exchange for lifting the port blockade, releasing frozen assets and waiving oil sanctions. The President initially rejected it and called reports of sanctions relief a "HOAX," but regional officials say back-channel work continues. Oil flows out of the Persian Gulf have reverted to more than 90% of pre-war levels as Iran loses control of the strait and U.S. isolation tactics squeeze its economy, which explains both Iran's motivation to deal and Washington's reluctance to compromise.

The conflict is in its eighth month. Brent settled near $98 Monday and traded at $103.97 Tuesday after briefly surging to $99.76 overnight; WTI's range across the same 24 hours was $90.94 to $94.18. Every dollar of oil is worth roughly 2 basis points on the 10-year and 3 basis points on October hike odds at current sensitivities, and every 10 basis points on the 10-year is worth roughly $30 to $40 on gold. That chain means a ceasefire that takes Brent to $85 is worth about $150 to $200 on gold, which would put the metal back at $4,350 to $4,400. An escalation that takes Brent to $115 is worth the same amount in the other direction, which would put it at $4,000.

The irony for gold bulls is that the metal's best short-term outcome is peace, and peace is the outcome that historically reduces safe-haven demand. That contradiction is only apparent. Gold at $4,193 has no war premium left in it; that was wrung out over eight months as the market learned the conflict produced Fed hikes rather than flight to safety. What gold has is a rate premium in reverse, a discount for the two hikes the market expects, and peace removes that discount. The war premium comes back only if the conflict escalates to a point where the Fed cannot hike, which means a demand shock large enough to threaten growth. That is the tail scenario, and it is also the scenario in which gold goes to $5,000, because a Fed that cannot hike into 4% inflation is the 1970s.

The Week's Data: PCE Wednesday and Payrolls Friday Decide the Double Bottom

Tuesday's macro slate is the first test. The August JOLTS report was scheduled for 10:00 a.m. ET with consensus at 7.23 million job openings against 7.271 million in July. The Conference Board's September consumer confidence was due at the same time with consensus at 90.1 against 89.4 in August. The University of Michigan's final September reading had already shown one-year inflation expectations jumping to 4.6% from 4.0%, the highest since June. The July S&P Case-Shiller home price index landed at 9:00 a.m. Wednesday brings the ADP employment change and, critically, the August PCE price index, the Fed's preferred inflation gauge. Thursday is jobless claims. Friday is September nonfarm payrolls and the unemployment rate.

The FOMC's September projections have PCE inflation at 3.7% for 2026 and the unemployment rate at 4.1%. A PCE print Wednesday that shows core running above a 0.3% monthly pace confirms the October hike and takes gold through $4,139. A print at 0.2% or below gives the doves an argument and gives gold a reason to reclaim $4,250. Payrolls Friday is the second gate: a number above 150,000 with unemployment at 4.1% or lower is a hike; a number under 75,000 with unemployment ticking to 4.3% is a pause. The Fed has said explicitly that labor market data, private sector earnings and capital investment show the economy strengthening, and it hiked on that basis. It needs to see that strength fade to stop.

The other data point that matters for gold is not on the U.S. calendar. The World Gold Council's third-quarter central bank purchase report is due in late October, and one of the mid-tier research desks has argued it matters more than the September rate hike. If Q3 central bank purchases match Q2's 288.9 tonnes, the structural floor is confirmed regardless of what the Fed does, because 290 tonnes a quarter is 1,160 tonnes a year, which would be the highest annual total on record and roughly a third of global mine supply. If Q3 reverts toward the Q1 pace of 16 tonnes net, the structural argument collapses and gold is a pure rates trade until it does not.

Several FOMC members and regional Fed presidents speak this week, and the House Speaker's meeting with AI executives Wednesday is a wild card for risk sentiment. The pattern to watch is whether any Fed speaker walks back the October hike. Two weeks ago the chair offered limited forward guidance and said the decision was "overdetermined." If that language holds through the week, October is a hike and gold has to find its floor without help from the Fed.

Bull Case: Double Bottom at $4,139, Reclaim $4,300, Target $4,600 by Thanksgiving

The bull case starts with what did not break. Gold at $4,193 is 25% below its January high and 16% above its 52-week low, and it has held $4,139 through the worst single-session decline in weeks, a 5.27% ten-year print, a 70% October hike probability, and a 4% crash in the miners. Central banks bought a record 288.9 tonnes in Q2 and are on a 91-tonne-per-month trend. Western ETFs took $2 billion in five days three weeks ago and holdings have not fallen. The sell-side's $4,900 year-end target was reaffirmed five days before the crash. The dollar is flat, not rising. Oil is falling, not rising. And the Fed itself does not project inflation back at 2% until 2029.

The trigger is a weekly close above $4,250. That reclaims the bottom of the September range, invalidates Monday's breakdown, and confirms the $4,139 to $4,150 double bottom as a base rather than a way station. From there the measured move on the pattern is $4,400, the top of the range and the zone the structural bulls called a floor. A reclaim of $4,400 puts the 200-day moving average near $4,450 in play, and above the 200-day the metal is back in an uptrend with $4,600, the August 25 level, as the next objective. $4,600 by Thanksgiving is a 9.7% gain from Tuesday's price. $4,900 by year-end, the sell-side target, is 16.9%.

The macro path is a ceasefire. A Hormuz deal that takes Brent to $85 to $90 removes the energy input to the inflation forecast, takes October hike odds from 70% toward 40%, takes the 10-year from 5.26% toward 5.0%, and takes gold up $150 to $200 on the rates channel alone. Add the ETF money that would return on a confirmed pause, and $4,400 is a two-week target. The Fed does not need to cut for gold to rally from here. It needs to stop hiking, and a peace deal is the most direct route to that outcome.

The structural path is the Q3 central bank report in late October. A print above 250 tonnes confirms that sovereign buyers absorbed the entire drawdown from $5,597 to $4,150 without pausing, which means the marginal seller through the correction was leveraged Western money that is now largely out. That is the definition of a washout. Gold has made every major low of the past 25 years when the last leveraged long has been liquidated and the last sovereign buyer is still buying. Monday's 4% crash on record central bank demand looks like exactly that moment. The upside case is $4,600 by late November and $4,900 by year-end, with $5,000 as the first-half 2027 target if the Fed pauses in December.

Bear Case: Lose $4,139, Fall to $4,000, and Test the March Levels

The bear case starts with what has already happened. Gold is down 6.65% in a month and 4.23% in a week, and it lost $145 in a session when oil rose $7. The 10-year is at 5.264% and the 30-year at 5.589%, both multiyear highs, and the global bond rout has not turned. The Fed's dot plot has 4.1% policy through 2027. October hike odds went from 49% to 70% in two weeks. Year-over-year gold returns have compressed to 9.5%, the lowest on record for the tracking series, which means the momentum bid is gone. The miners are pricing $4,000 gold, and the miners have led the metal at every major turn of the past two years.

The trigger is a daily close below $4,139. That breaks the double bottom, the August 5 low, and the two-month range in one move, and it opens $4,100 immediately and $4,000 within days, because there is no support on the daily chart between the August low and the round number. $4,000 has not traded since March, before the Iran conflict intensified, and a test of it would be a 4.6% decline from Tuesday's price. Below $4,000 the March and February levels between $3,850 and $3,950 are the next references, and a full retracement toward the $3,600 52-week low would be a 36% drawdown from the January peak.

The macro path is a hot PCE Wednesday and a strong payrolls Friday. Core PCE above 0.3% monthly and payrolls above 150,000 confirm the October hike, push the 10-year through 5.30%, and push October hike odds toward 90%. In that environment gold has no rates support, no momentum support, and a Western ETF base that is 3.5% underwater and looking for an exit. The June research call from one of the large bank desks laid it out: an extended hiking cycle produces sustained Western ETF outflows, and if central bank buying intensity dips at the same time, gold has no bid at all. Q1 2026, when net central bank purchases were 16 tonnes and Türkiye sold 60, is the template for that scenario.

The structural risk is that the Q3 central bank report in late October disappoints. Sovereign buyers have been treated as a permanent bid, but they are price-sensitive too: Türkiye sold 60 tonnes into the January peak, and the Q1 pause coincided exactly with gold above $5,000. If Q3 shows a return to that pattern, the $4,900 year-end target gets cut to $4,300 or lower, and the market reprices gold as a rates instrument with no structural floor. The downside case is $4,000 by mid-October, $3,900 on an overshoot, and a range of $3,800 to $4,200 through year-end if the Fed hikes in October and signals December.

What the Gold-Silver Ratio, the Dollar, and Bitcoin Are Saying About the Bottom

Three cross-asset signals help triangulate whether Monday was a low. First, the gold-silver ratio. With gold at $4,193 and silver at $61.24, the ratio sits at 68.5. Silver fell less than gold Monday on a percentage basis and was down only 0.78% Tuesday, which means the ratio compressed during the selloff. Silver outperforming gold during a precious-metals decline is unusual and historically bullish for the complex, because it implies the industrial and speculative bid in silver is holding while the rates-driven selling is concentrated in gold. In the 2013 and 2022 gold corrections the ratio expanded as silver got hit harder; the opposite happening now suggests this is a gold-specific rate repricing, not a precious-metals liquidation.

Second, the dollar. The dollar index was flat Tuesday and was firm but not surging Monday. Gold's worst historical outcomes come when the dollar and real yields rise together, as they did through most of 2022. Right now yields are rising and the dollar is not, which means the foreign-currency gold price is falling less than the dollar price. For central banks that buy in local currency, and for the Chinese and Indian retail markets, gold at $4,193 with a flat dollar is a shallower correction than the headline suggests, which supports the structural bid.

Third, Bitcoin. Bitcoin traded at $84,254 Tuesday, up 0.75%, on the same tape that had gold up 0.59%. Both non-yielding assets bounced together on lower oil, and both had been sold together Monday. The correlation is the point: gold and Bitcoin are both trading as rate instruments this week, which means neither is being bought as a hedge against the other or against equities. When the correlation between gold and Bitcoin is this high, it is because a single macro variable, the real yield, is driving both. That variable turns when the Fed pauses. Until then the correlation holds and gold cannot decouple upward.

The composite read is that Monday's low has the cross-asset characteristics of a washout: silver held, the dollar did not spike, and the selling was concentrated in the most rate-sensitive precious metal. It does not have the characteristics of a durable bottom: RSI is not at extremes, the moving averages are all overhead, and the catalyst that would reverse the trend, a Fed pause, is not on the calendar. That is a bounce, not a bottom. The bottom needs a data print.

Verdict: Neutral to Bearish Into PCE, Buy the Reclaim of $4,300, Target $4,600

Gold at $4,193 is a hold with a bearish tilt into Wednesday's PCE print and a buy on a confirmed reclaim of $4,300. The structural bull case is intact and it is the strongest it has been in decades: record central bank demand at 288.9 tonnes a quarter, a 91-tonne-per-month trend, $2 billion of ETF inflows in the first week of September, gold overtaking Treasuries in reserve portfolios for the first time since 1996, and a Fed that projects inflation above target through 2028. The $4,900 year-end target from the largest sell-side house on the metal has not been cut. None of that has changed since Monday.

What has changed is the price of the Fed. October hike odds went from 49% to 70% in two weeks, the 10-year went from 5.10% to 5.264%, and gold went from $4,400 to $4,150 to reflect it. The metal has repriced two hikes, not one, and it did so in a single 4% session that took the miners down 4% to 13%. That repricing is complete only if the October hike is the last one. If Wednesday's PCE and Friday's payrolls put December on the table, gold has another $150 to $200 of downside to $4,000, and the double bottom at $4,139 fails.

The forecast: gold trades a $4,100 to $4,300 range through Friday's payrolls, with the direction of the break decided by the data. A weekly close above $4,250 confirms the double bottom and targets $4,400 in two weeks and $4,600 by late November, a 9.7% gain. A daily close below $4,139 targets $4,000 by mid-October, a 4.6% decline, with an overshoot to $3,900 possible on forced selling. The odds are close to even into PCE and tilt bullish after it if the Hormuz talks produce a deal, because oil is the single input that moves everything else.

The trade is to wait. Gold has spent eight months proving it cannot rally into a hiking Fed, and the Fed has spent two weeks proving it is still hiking. The next durable low comes when the Fed pauses or when the Q3 central bank report proves the sovereign bid does not care what the Fed does. Both are October events. Until then $4,300 is the level that turns a bounce into a base, $4,139 is the level that turns a correction into a bear market, and $4,193 is a price in between that will not last the week.

That's TradingNEWS