Gold Fell Below $4,000 as Oil Surged. What Changed for Metals Investors?

Spot gold fell 4.07% to US$3,995.80 and silver 5.45% in the week to 23 October 2026, while December Brent rose 6.24% and the US 10-year real yield fell 8 basis points. Our Tactical Macro Conditions Score slipped from 46 to 45/100, into the Cautious band, with its direction indicator at Stable. We separate the week’s evidence from what a metals holding is for.

Published · 17 min read

Key Takeaways

Gold ended the week below US$4,000 an ounce on the spot series we use, while December Brent crude rose more than 6%. Silver fell harder than gold. Yet the US 10-year real yield fell again and the dollar index barely moved. This review covers the week of 19 to 23 October 2026 and compares Friday 23 October with Friday 16 October. Metal prices are US-dollar spot readings from one provider’s daily history, Investing.com’s, not LBMA benchmark prices, futures settlements or what a coin or bar cost at a dealer.

US$3,995.80 Spot gold on 23 October 2026, Investing.com’s daily history: down 4.07% from US$4,165.50 on 16 October. Over the same week spot silver fell 5.45%, to US$53.9199, December Brent crude rose 6.24%, to US$116.72 a barrel, and the 10-year real yield fell 8 basis points, to 2.66%. A provider’s spot reading below US$4,000 is not a claim that every gold contract or benchmark closed there.

That combination challenges two easy explanations: that geopolitical and inflation risk automatically lifts bullion, or that a lower competing yield guarantees a recovery. Neither describes this week. The more useful question is not whether gold is a hedge, but a hedge against which risk, over what period and at what position size. A short-term trade, a long-term diversifying holding and a preference for owning metal outright serve different purposes, and confusing them makes a difficult week harder to read.

How far did gold and silver fall this week?

Spot gold fell 4.07% and spot silver 5.45% between 16 and 23 October on Investing.com’s daily history, while December Brent crude rose 6.24%. Both Treasury yields fell, and the dollar index ended almost where it began. On these readings the gold-silver ratio, the number of ounces of silver one ounce of gold would buy, rose from about 73 to about 74.

Measure16 October23 OctoberChange
Gold, spot, US$/oz4,165.503,995.80−4.07%
Silver, spot, US$/oz57.025753.9199−5.45%
Brent, December, US$/bbl109.86116.72+6.24%
US 10-year Treasury yield5.09%5.00%−9 bp
US 10-year real yield2.74%2.66%−8 bp
US Dollar Index101.84101.81−0.03%
US$1 in C$, daily average1.41541.3993−1.1%

Sources: Investing.com, XAU/USD, XAG/USD and US Dollar Index daily history; Reuters and Bloomberg settlement reports for December Brent; US Treasury, Daily Treasury Par Yield Curve Rates and Daily Treasury Par Real Yield Curve Rates, 2026; Bank of Canada, daily exchange rates. Percentage and basis-point changes are our arithmetic. These are different observation conventions, not synchronised prices or returns after costs.

The benchmark matters most around a round number. US$3,995.80 is one provider’s daily spot observation, not a claim that the December futures contract, the LBMA Gold Price or a dealer’s quote was identical. Our explainer on who sets the gold price sets out why those numbers differ.

Equities did not move as one. Over the week the S&P 500 lost 0.5% and the Nasdaq composite 0.7%, but the Dow Jones Industrial Average gained 0.4% and the Russell 2000 2.0%: weak metals did not coincide with every major share index falling. Credit offered a different caution. The ICE BofA US high-yield spread was wider on Thursday 22 October than on the previous Friday; Thursday was the latest observation published when we read the series, so this is not a Friday comparison.

Share indexClose on 23 OctoberChange over the week
S&P 5007,623.68−0.5%
Nasdaq composite25,955.34−0.7%
Russell 20002,775.90+2.0%
Dow Jones Industrial Average52,215.16+0.4%

Source: the Associated Press’s closing table for 23 October 2026, index levels and weekly changes as reported.

Why did oil rise, and what does it mean for metals?

Oil rose on renewed concern about attacks on Saudi energy infrastructure and a wider regional conflict, according to Bloomberg’s report of 23 October, which described Houthi claims of attacks on the Saudi Red Sea port of Yanbu and Aramco evacuating its Dhahran headquarters. December Brent settled at US$116.72, up 3.1% on the day; the Associated Press called it Brent’s highest level since June 2022. Bloomberg also noted that crude flows from the Persian Gulf had recovered while oil-product shipments from the region remained constrained. A futures-price surge is not, by itself, a measure of barrels already lost.

For investors the possible channels run in different directions. More expensive energy can raise concern about purchasing power and encourage demand for protection. It can also squeeze spending and margins, or make policymakers less comfortable easing rates. These are mechanisms to weigh, not a verified account of what caused gold’s decline.

Friday’s business survey made the policy question harder. S&P Global’s flash US Composite PMI was 57.6 in October, down from 58.4 in September but still signalling strong expansion. New orders and hiring remained firm, and businesses reported selling prices rising at the fastest pace since August 2022. The survey collected responses from 9 to 22 October; it is a preliminary business indicator, not an official GDP figure.

Households described a less comfortable economy. The University of Michigan’s final October consumer sentiment index was 48.0, against 48.1 in September: essentially unchanged, not a fresh collapse. Year-ahead inflation expectations rose from 4.6% to 4.7%, and long-run expectations from 3.4% to 3.5%.

Canada’s September consumer price index, published on 19 October, showed the same need to look beneath a headline. Annual inflation rose from 3.0% to 3.1%, but excluding gasoline it slowed from 2.4% to 2.2%. Excluding gasoline is one particular comparison, not a synonym for every measure of core inflation.

Our reading is that neither a clean disinflation story nor an imminent-collapse narrative fits all of this. Resilient activity, household strain and price pressure can coexist, which is a reason to weigh competing risks rather than assume every release points to the same policy decision.

The week’s evidence did not point one way

Read both columns: Lower real yields did not produce a metals recovery, and strong activity did not remove inflation concerns. No single input explains the week.

Did lower real yields help gold this week?

Only in one respect: the fall in the 10-year real par yield from 2.74% to 2.66% reduced one competing return available outside bullion, but it remained well above zero and gold still fell. This is the Treasury’s inflation-protected yield curve, not the ordinary bond yield minus the latest CPI figure.

The dollar index was essentially flat over the week rather than an additional headwind. Together those observations make a one-factor explanation inadequate. They do not tell us which other influence dominated each trade, or how gold would have performed without the fall in yields. The Learning Academy’s counterarguments section is relevant here: a constructive reason to own metals should be tested against opportunity cost and the possibility of prolonged underperformance, not treated as a promise of an immediate gain.

What did the Macro Conditions Score record for 23 October?

The dashboard’s recorded Tactical Score for Friday 23 October was 45/100, one point below its 46/100 for 16 October and one point below the floor of the Neutral band, so it now sits at the top of the Cautious band. Its direction indicator read Stable, against Deteriorating a week earlier. In between, the score read 48 on 19, 20 and 21 October and 45 on 22 October. The Long-Term Structural Score was unchanged at 61/100, Bullish, in its 19 October update. The Macro Conditions Score reflects macro conditions historically associated with precious metals performance. It is a systematic framework, not investment advice.

The Tactical Score at each run, 16 to 23 October
  • 16 Oct: 46 pts
  • 19 Oct: 48 pts
  • 20 Oct: 48 pts
  • 21 Oct: 48 pts
  • 22 Oct: 45 pts
  • 23 Oct: 45 pts

One point, one band: Neutral runs from 46 to 60 and Cautious from 26 to 45, so a one-point fall moved the reading across the boundary.

The week’s change was small: up two points early in the week, down three on 22 October.

Source: GoldSilverPortfolio Macro Conditions Score dashboard (Engine V2), recorded tactical snapshots for each weekday run from 16 to 23 October 2026, read from its record on 24 October 2026.

Two blocks moved, in opposite directions. The fiscal and Treasury stress block fell from 8 to 5 of its 10 points: it scores the 10-year term premium, its change over 13 weeks, and whether over four weeks the 10-year yield rose, the dollar fell or both, and its term-premium reading eased from 1.07 to 0.99 percentage points (observations for 9 and 16 October). The momentum block rose from 13 to 15 of its 30 points. The core levels stayed at 19 of 50, and the confirmation basket stayed at 6 of 10 even though three of its signals changed: the VIX component lost a point, the copper-gold ratio component lost one and the credit-spread component gained one, taking the raw basket from 15 to 14 of 24.

What adds up to 45
  • Tactical Score, 23 October 2026: 45 pts
  • Core levels: 19 pts
  • Momentum: 15 pts
  • Fiscal and Treasury stress: 5 pts
  • Confirmation basket: 6 pts

Momentum added two points and fiscal stress took away three. The core levels, which include the real yield, did not move.

Source: GoldSilverPortfolio Macro Conditions Score dashboard (Engine V2), recorded reading for 23 October 2026, as stored at that day’s run. Blocks: core levels 50 points, momentum 30, fiscal and Treasury stress 10, confirmation basket 10.

Block16 October23 OctoberMaximum
Core levels191950
Fiscal and Treasury stress8510
Momentum131530
Confirmation basket6610
Tactical Score4645100

Does the direction indicator agree with the score?

On 23 October, yes. The momentum block’s direction index was −0.02, inside the band the engine reads as Stable, and a one-point fall in the score is consistent with that. Underneath, its five components pulled against each other. The real-yield component swung from neutral to a clear positive, +0.62, as the 10-year real yield fell, and the dollar component went from a drag of −0.27 to almost nothing. Inflation expectations remained a small drag, −0.10, and the two-year yield component, the engine’s proxy for repricing of the Federal Reserve’s path, a negligible one. Gold’s own price momentum went the other way, from −0.33 to −1.00, the largest drag the component can register. The dashboard named gold momentum and inflation expectations as the strongest drivers.

The disagreement was a week earlier. On 16 October the direction index was already −0.15, inside the Stable band, yet the indicator still read Deteriorating while the score rose seven points. That is by design: the engine changes state only when the same new state appears at two consecutive runs. The 16 October record holds Stable as the pending state, and the run of Monday 19 October confirmed it.

Input dates matter too. The dashboard runs at about 20:00 UTC and reads the real yield from FRED’s DFII10 series, whose latest value at Friday’s run was for 21 October. It therefore scored 2.76%, not the Treasury’s 2.66% for 23 October. Both readings sit inside the 2.5% to 3.5% band in which the real-yield level input earns 2 of its 20 points, so the timing changed nothing in the core levels; the fall did reach the momentum block, which reads the real yield’s change over two to four weeks. The futures-positioning input is the CFTC report for 20 October, and the term premium is dated 16 October. Our review of the week to 16 October set out the same banding and timing.

A Cautious reading is not a forecast and not an instruction. It records that the measured near-term conditions were fractionally less favourable on 23 October than a week earlier, after a week in which the inputs moved in both directions. Headlines earn no points, and we have not adjusted the published reading.

Were futures traders selling before the drop?

Only slightly, and the data stop on Tuesday. The CFTC’s disaggregated futures-only report shows that on 20 October managed money held a net long position of 104,206 COMEX gold contracts, down 2,883 from 13 October. Silver’s net long fell by 2,114 contracts to 3,505. Both groups remained net long.

The underlying changes differed. In gold, longs fell by 3,777 contracts to 120,970 and shorts by 894 to 16,764. In silver, longs rose by 378 to 16,180, but shorts rose more, by 2,492 to 12,675. A smaller net long does not mean every investor sold, and gold and silver contracts represent different quantities of metal. More importantly, these Tuesday observations predate Thursday’s and Friday’s trading. They cannot establish who sold during the later decline, whether anyone faced a margin call, or whether money moved from gold into any particular asset.

This extends the distinction in our article on September’s fund buying: fund holdings, futures exposure and prices are related measurements, not entries in one physical-metal account. Our Who Holds Gold Futures chart lets readers explore the categories separately.

Silver’s India headline: a permit is not a purchase

Reuters reported on 22 October that India had issued permits to import around 6,500 tonnes of silver over the next six months, according to four bullion dealers who received them. Dealers also said premiums over official domestic prices had fallen to 45 cents an ounce, from up to US$2.50 the previous week, as jewellery and investment demand stayed weak.

Those are reported permits and dealer observations, not confirmed shipments or a measured addition to global consumption, and we have not aggregated the permits ourselves. The useful distinction is between access to supply, actual buying and the premium paid locally. Our premium and break-even calculator addresses the separate question of what an investor pays above spot; its calculations exclude shipping, insurance and payment-method surcharges.

What are commentators saying after the fall?

In CPM Group’s written commentary of 23 October, published on Kitco, Jeffrey Christian discusses gold and silver falling alongside other assets and notes that investors who have been waiting for lower prices may keep waiting until the market shows signs of stabilising. CPM says selling pressure could persist into the US congressional elections, and that a difficult post-election environment could encourage renewed buying. It states that the timing and price of a bottom remain uncertain.

Read it as an attributed outlook, not an explanation of every sale. Its discussion of economic risk also needs the counterweight of this week’s strong business survey.

A different method appears in Chris Vermeulen’s written update of 22 October for The Technical Traders. He writes that recent rebounds have struggled to gain traction and short-term trends have turned unfavourable, which is why his strategy moved to the sidelines. That is an external technical view, not a GoldSilverPortfolio forecast or an instruction for every investor to sell.

Both are written accounts accompanying presentations; we have not reproduced their charts or tested their strategies. The productive comparison is between methods and time horizons, not a tally of bullish and bearish commentators. Our two-horizon portfolio framework develops that distinction.

What is a precious-metals holding supposed to do in a week like this?

It depends on the purpose the investor gave it. GoldSilverPortfolio favours a deliberately sized precious-metals portion of a diversified portfolio; the allocation, metal mix, ownership method and strategy remain the reader’s decisions. Our aim is to explain the options and their trade-offs, not to prescribe the same position for everyone.

A long-term purchasing-power objective is different from expecting a gain after each inflation release. Judging it needs a chosen horizon, purchase costs and an appropriate inflation comparison. Our inflation calculator compares historical gold and silver prices with US CPI using annual averages; it does not reproduce an individual’s purchase dates, costs or household inflation.

An ownership objective is different again. Someone who wants to hold metal personally weighs different dependencies from someone who wants convenient exchange-traded exposure. Storage, insurance, custody terms and exit costs still matter, and changing the form of ownership does not remove price risk. The Academy’s physical-versus-paper section frames that comparison.

A tactical objective needs rules for entry, exit and the possibility of being wrong. A strategic holding also needs review: calling it long term is not permission to overlook concentration, changed circumstances or money needed soon. Mining shares and other speculative exposures should not quietly stand in for a bullion holding meant to serve a different purpose.

Our article on silver’s two peaks makes the time-horizon question concrete: on LBMA annual averages, 1980’s silver price was not exceeded for 31 years. That is one historical episode on an annual-average convention, not a forecast for today’s silver. The Long Shadow simulation in the Academy’s common mistakes section is a prompt to test expectations, not a prediction of a repeat.

Depending on their circumstances, investors might use allocation ranges, scheduled purchases or separate strategic and tactical positions. None removes the risk of loss. The allocation lessons help organise that decision; a lower price alone does not decide it.

Canadian readers also carry currency exposure. The Bank of Canada’s daily average moved from C$1.4154 per US dollar on 16 October to C$1.3993 on 23 October: the Canadian dollar strengthened by about 1.2%. All else equal, that lowers the Canadian-dollar value of a US-dollar metals holding. The daily averages and the spot readings are not taken at the same moment, so we have not combined them into a precise Canadian-dollar return.

A useful portfolio role does not require a gain every week. It does require a size and a rationale the investor can live with when the price falls.

What is scheduled next?

The Federal Reserve’s and the Bank of Canada’s rate decisions both come on 28 October, followed by US third-quarter GDP and September’s PCE inflation data on 29 October. The questions are whether policy communication changes the rate outlook, whether price pressure persists, whether energy risks ease in practice, and whether metals regain strength as new positioning data arrive. None of those outcomes is assumed here.

The next three days on the calendar

The bigger picture

Gold below US$4,000 and Brent above US$116 describe a difficult week, not a complete investment strategy. Lower real yields did not produce a metals recovery, strong business activity did not remove inflation concerns, and one commodity’s gain did not protect another’s price.

The lesson is to separate what happened, what might explain it and what a portfolio is designed to do. The reader decides what exposure fits, how it is held and which risks they are prepared to carry.

Questions readers ask

Did gold fall below $4,000 in October 2026?
On Investing.com’s daily spot history gold was US$3,995.80 an ounce on Friday 23 October 2026, down 4.07% from US$4,165.50 on 16 October. That is one provider’s spot observation; the December futures contract, the LBMA Gold Price and dealer quotes are separate prices and need not have closed at the same level.
What was the Macro Conditions Score on 23 October 2026?
The recorded Tactical Score was 45/100, the top of the Cautious band, against 46/100 on 16 October. The fiscal and Treasury stress block fell from 8 to 5 points and the momentum block rose from 13 to 15; the core levels stayed at 19 and the confirmation basket at 6. The direction indicator read Stable, with a direction index of −0.02. The Long-Term Structural Score was 61/100 in its 19 October update.
Why did gold fall when real yields fell?
In the week to 23 October 2026 the US 10-year real yield fell 8 basis points to 2.66%, which lowers one cost of holding gold, but it remained well above zero, and spot gold still fell 4.07% while December Brent rose 6.24%. A lower yield removes part of a headwind without setting the price, and the week’s other evidence, from strong business activity to rising price pressure, pointed in different directions.

Sources and Methodology

This review covers 19 to 23 October 2026 and compares Friday 23 October with Friday 16 October; every figure keeps its own reference date. Gold, silver and the US Dollar Index use Investing.com’s daily-history rows rather than the live banners on the same pages, read on 24 October, not dealer prices, an LBMA benchmark or a futures settlement. Brent settlements are secondary reporting, Reuters for 16 October and Bloomberg for 23 October, not an exchange file. Share-index changes are the Associated Press’s closing table. Percentage changes are (later ÷ earlier − 1) × 100 and basis-point changes are (later − earlier) × 100, both our arithmetic, as are the gold-silver ratios and the futures net positions. The high-yield spread is described from FRED’s ICE BofA series, with observations to 22 October, without reproducing its values. The flash PMI figures were read from S&P Global’s release as reported by ActionForex and the Seoul Economic Daily. Prices exclude transaction costs, taxes and individual currency arrangements.

The Macro Conditions Score readings are the dashboard’s recorded Engine V2 snapshots for each weekday run from 16 to 23 October 2026 and the Long-Term Structural Score update of 19 October, with their blocks, component signals, direction components and input dates, read from the record on 24 October; nothing was recalculated or adjusted for this article. India’s permits are Reuters’ reporting of what four dealers said, not an inspection of the licences. The views of CPM Group and Chris Vermeulen are taken from their written accounts, not independent transcriptions of their presentations. Proposed explanations and portfolio implications are editorial analysis, not a transaction-level causal audit, and the linked calculators’ interactive results were not re-tested for this article.

Conflicts of interest, stated plainly. CPM Group sells research and advisory services, and its commentary was published on Kitco, which also sells bullion. The Technical Traders markets trading services. GoldSilverPortfolio has no affiliate or commission arrangement with CPM Group, Kitco or The Technical Traders, and their work is included for discussion, not as an endorsement of any product or strategy. GoldSilverPortfolio may earn a commission through two links elsewhere on this site: an affiliate link to SilverGoldBull, a bullion dealer, and a referral link to Wealthsimple, a brokerage. It also sells TaxStack, a paid record-keeping tool, and publishes the Macro Conditions Score discussed here. No fund, dealer, trading signal or commentator’s forecast is presented here as a recommendation to buy or sell.

Primary sources