Gold (XAU/USD)

State (current)

Monday was the cleanest confirmation this entity has produced, and it arrived on the exact distinction we wrote 24 hours earlier. Yesterday's call said the instruction was narrower than "watch the 10Y": watch WHERE on the curve the move happens, because a nominal rise driven by term premium, deficits and oil is not the same as one driven by hike expectations, and only the second reliably lifts real yields. Monday delivered the term-premium version in its purest available form. The 30-year Treasury yield rose to 5.31%, the highest since June 2007, explicitly attributed to persistent-inflation worry and government borrowing, while the September hike probability FELL to roughly 30-31%. Gold rose. Spot gold traded near $4,394.70 an ounce, up 0.44%, with spot silver at $65.430, up 1.32%. A 19-year high in the long bond and a higher gold price on the same session is not an anomaly, it is the mechanism: nominals rose without expected inflation falling, so real yields did not have to rise.

The intraday behaviour is worth keeping because it shows the equation working at higher frequency too. Gold slipped BELOW $4,380 after the Empire State manufacturing index surprised at 20.6, its highest in more than four years, with prices paid up 6 points to 58.6. A hot activity print is a rate-expectations input, and gold reacted to it the way the framework says it should, then recovered as the dollar stayed softer and hike pricing stayed low. One survey moved the front-end channel and gold flinched; the long end moved on deficits and gold did not care. Same day, both channels, correct signs.

The geopolitical layer changed character rather than intensity. The 60-day US-Iran deadline expired Monday with no compromise, the memorandum of understanding lapsed, Iran ruled out direct negotiations, and Brent hit $90 after Trump said he does not see the war ending soon, with tanker traffic through Hormuz slowing sharply over the weekend. Set against that, Iran said it is close to finalising an understanding with Oman on a transit route, though the US is not party to those talks. We have been calling the Hormuz component a rentable bid rather than an owned one, and that stays right, but an expired deadline is a different object from a headline: it removes the option value of near-term resolution instead of repricing its odds. Kitco's own framing keeps the setup two-sided, with softer Fed pricing and a softer dollar supporting the metal while any further oil spike limits the rally through the yields channel.

Positioning caution is now three runs old and unresolved: Kitco's weekly survey has Wall Street fully bullish, which argues against adding into strength. Technically the box has widened on the upside, resistance $4,448 then $4,518 then $4,596, support $4,333 then $4,262 then $4,205, and gold has still not reclaimed $4,448, which is the level the current invalidation is written against.

Timeline

• 2026-08-18: Thesis confirmed on its own stated test, one day after the test was specified. Spot gold ~$4,394.70/oz +0.44% in early US trading Monday, spot silver $65.430 +1.32%, ON A DAY the 30-YEAR Treasury yield rose to 5.31%, the HIGHEST SINCE JUNE 2007 (+0.04pp), cited as driven by persistent-inflation worry and GOVERNMENT BORROWING. September hike probability fell to ~30-31% (from ~33%). That is the term-premium channel, not the rate-expectations channel, and gold behaved as the 08-17 call predicted it would. Intraday cross-check: gold traded BELOW $4,380/oz after Empire State manufacturing surprised at 20.6, highest in 4+ years, new orders 17.3, shipments 11.7, PRICES PAID +6 to 58.6, then recovered on a softer dollar index. Geopolitics: 60-day US-Iran peace deadline EXPIRED Monday with no compromise, MoU lapsed, Iran ruled out direct negotiations, BRENT HIT $90/bbl after Trump said he does not see the war ending soon (WTI ~$82.95, Brent ~$89.44 in the US morning); Hormuz tanker traffic slowed sharply over the weekend; offsetting, Iran said it is close to finalising an Oman transit understanding, with the US not party to those talks. Kitco framing: softer dollar and weaker Fed-hike pricing support gold, any further oil spike limits the rally through the YIELDS channel. Technicals UPDATED: resistance $4,448.00 then $4,518.00 then $4,596.00; support $4,333.00 then $4,262.00 then $4,205.00. Silver: resistance $66.78 / $68.24 / $69.63, support $65.15 / $63.57 / $62.20. No verified Kitco PM close for 17 Aug at capture time, the AM mark is what is banked (http://www.kitco.com/news/article/2026-08-17/gold-silver-rise-softer-data-cuts-fed-hike-odds-kitco-am-report ; https://www.kitco.com/news/article/2026-08-17/spot-gold-trades-below-4380oz-after-new-york-manufacturing-index-surprises ; https://www.cnbc.com/2026/08/17/treasury-yields-federal-reserve-fomc-minutes.html)
• 2026-08-17: Still no new session (weekend), but Wednesday's test is now specified with more precision than we had on 08-16. The Fed held July at 3.50-3.75% with three hike dissents and no forward guidance, September hold priced ~67%, and the July meeting produced a curve STEEPENING driven by TERM PREMIUM rather than by rate expectations. That distinction matters for gold. A nominal yield rise driven by term premium, deficits and the balance-sheet question is not the same as one driven by hike expectations, because the first can occur without breakevens falling and therefore without real yields rising. The 30Y at 5.216% and the $432B July budget deficit are term-premium inputs, not rate-expectation inputs. So the configuration that hurt gold on 14 August, breakevens falling faster than nominals, is a hawkish-minutes outcome, while a long-end selloff on deficits is not. On the other side, the BoJ is around 62% priced for a September hike with July core CPI expected at 1.8% from 1.6%, which is a competing source of global duration supply and a yen variable to watch. Support and resistance unchanged from the 08-15 Kitco marks; gold has still not reclaimed $4,448 (https://global-view.com/fomc-newsquawk-week-ahead-in-focus-17th-21st-august-2026/ ; https://www.fxstreet.com/news/japan-gdp-steady-inflation-near-boj-target-ing-202608141420)
• 2026-08-16: No new gold session (weekend), but two inputs to the real-rate equation moved and they pull in opposite directions. On the nominal side, market pricing for the first Fed hike moved fully into 2027 with only 9bp priced for September and 23bp by year-end, which caps how far nominals can rise on rate expectations alone. On the expected-inflation side, the Hormuz escalation continued with no diplomatic offset: a third ADNOC vessel attacked Friday evening, UKMTO notified Saturday of a projectile striking a bulk carrier, Trump vowing to declare the strait US territory, Iran saying it opens only under Iranian command, and Bessent promising unprecedented economic isolation with measures due this week. US gasoline sits at $4.07/gal vs $3.16 a year ago and WTI near $81. Lower priced-in nominals plus a live upside tail to expected inflation is the configuration under which real yields fall. Offsetting honestly: Iraq is exporting 2.0 mb/d, the highest daily rate since the crisis began, so physical supply is adapting and price has not broken out. IG's weekly mark had gold marginally lower on the week at $4,324 as at 14 Aug 05:46 BST, which is a pre-Friday-close snapshot and sits below the $4,373.48 equity-close mark already banked; use the Kitco close, not the IG weekly mark (https://www.ig.com/en-ch/news-and-trade-ideas/week-ahead--17-august-2026-260814 ; https://www.cbsnews.com/live-updates/iran-war-donald-trump-strait-of-hormuz-uae-accuses-attacks-vance-oil-gas/ ; https://gasprices.aaa.com/)
• 2026-08-15: The real-rate equation printed with the opposite sign to Thursday and gold behaved correctly again. Spot gold rallied to a session high $4,387/oz after July retail sales -0.6% (vs +0.1% cons, ex autos -0.3% vs +0.2%), last $4,380.88 +0.68% per Kitco AM, marked +0.24% at $4,373.48 at the US equity close; futures quoted 4,432.00 +0.26%. Silver $64.740 +0.66%. Mechanism: Sept hike odds fell to ~32-33% from ~50% a week earlier pushing nominal yields down, WHILE UMich year-ahead inflation expectations rose to 4.3% from 4.2%, so implied real yields fell. Cap on the move: oil firmed on Hormuz, WTI ~$82.80, Brent ~$88.50, after two UAE tankers were attacked and Washington signalled new economic-isolation measures on Iran; Kitco frames gold's setup as two-sided (shipping risk supports haven demand, higher crude keeps inflation risk alive and limits the downside in yields). DXY below 100.00 and softer. Kitco weekly survey: Wall Street fully bullish as hike bets recede, Main Street bullish majority. Technicals unchanged: resistance $4,448 then $4,575 then $4,666; support $4,330 then $4,322 then $4,263 and $4,205 (https://www.kitco.com/news/article/2026-08-14/gold-rebounds-retail-sales-miss-offsets-oil-driven-inflation-risk-kitco-am ; https://www.kitco.com/news/article/2026-08-14/gold-rallies-session-high-4387oz-after-retail-sales-drop-06-july ; https://www.kitco.com/news/article/2026-08-14/wall-street-goes-full-bull-gold-rate-hike-bets-recede-main-street-maintains ; https://finance.yahoo.com/markets/stocks/articles/stock-market-today-aug-14-213018651.html)
• 2026-08-14: The real-rate thesis got a corrective test and taught the sharper version of itself. Gold -1.31% to ~$4,349.80 (spot ~$4,387 right after the PPI release), silver -1.40% to $64.290, snapping a four-day advance, on a day the 10Y FELL to 4.648% from 4.686% and Sept hike odds fell to 34.6% from 40.6%. Cause per Kitco: cooler wholesale inflation cut yields and hike expectations but also cut demand for metals as near-term inflation hedges, i.e. breakevens fell with nominals so real yields did not ease. Simultaneously the Iran premium unwound: Brent -2.1% to $87.07, WTI ~$81.50 on lower demand forecasts, despite Hormuz traffic still restricted. Kitco technicals: support $4,332 then $4,262; resistance $4,448 then $4,575 (https://www.kitco.com/news/article/2026-08-13/gold-snaps-four-day-advance-yields-ease-oil-drops-kitco-pm-report ; https://www.kitco.com/news/article/2026-08-13/spot-gold-4387oz-after-july-ppi-posts-flat-00-core-ppi-02 ; https://www.kitco.com/news/article/2026-08-13/gold-silver-slip-flat-ppi-masks-firm-core-pressure-kitco-am-report)
• 2026-08-13: The real-rate test passed. On CPI day gold rose ~0.66% to ~$4,399 (spot near $4,406, +0.90% intraday), testing two-month/10-week highs, as July CPI printed in line/soft and the 10Y held flat at ~4.69%. Cooler CPI trimmed Sept hike odds and eased implied real yields; no VIX spike, so real-rate not fear. Oil stayed firm (WTI ~$83.20) keeping a modest safe-haven layer on top (https://www.kitco.com/news/article/2026-08-12/gold-prices-test-two-month-highs-cpi-trims-yields-fed-hike-odds-kitco-pm ; https://www.kitco.com/news/article/2026-08-12/gold-silver-rally-cpi-cools-oil-keeps-fed-risk-alive-kitco-am-report ; https://tradingeconomics.com/commodity/gold)
• 2026-08-12: Gold held >$4,400 (spot ~$4,396.53 9am ET Tue, Dec futures $4,446.90 +0.6%) as US-Iran peace optimism faded and WTI jumped to ~$82; a fresh geopolitical bid on top of the real-rate story, while the 10Y eased to 4.69% and VIX only ticked to ~15.8 (not a panic bid) (https://finance.yahoo.com/personal-finance/investing/article/gold-prices-today-tuesday-august-11-2026-gold-remains-over-4400-as-iran-situation-worsens-115212228.html ; https://www.forbes.com/advisor/investing/oil-prices-today/)
• 2026-08-11: Gold +1.14% to $4,390.85 (spot) as the 10Y rose to 4.71% on firmer oil into CPI week; holding a bid through a higher nominal yield = the real-rate/Fed-on-hold read, not fear (VIX 14.9); author framing = money repositioning for a Fed that stays on hold (https://www.indexbox.io/blog/gold-and-silver-mixed-as-markets-await-key-us-inflation-data/ ; https://www.fool.com/coverage/stock-market-today/2026/08/10/stock-market-today-aug-10-markets-slip-as-oil-gains-fuel-inflation-fears/)
• 2026-08-10: Gold holds the >$4,400 breakout (futures $4,402.00) into CPI week; VIX 14.90 and near-flat equity futures = calm-but-hedged tape; author framing = breakout could force under-allocated portfolios back into real assets (https://finance.yahoo.com/markets/article/golds-breakout-could-force-investors-back-in-chart-of-the-day-113004349.html)
• 2026-08-08: Gold $4,399.70/oz +2.33%, +8% over 5 sessions; UBS forecasts toward $5,000 in H1 2027; drivers = weak jobs -> lower real yields, China physical demand, ETF inflows, central-bank buying, yen-support capping yields (https://finance.yahoo.com/markets/article/golds-rally-has-support-with-prices-set-to-rise-toward-5000-next-year-ubs-says-100000563.html)
• 2026-08-05: Gold +2.2% to $4,245 as 10Y eased and oil sank (investopedia)
• Context: gold rose more than 65% in 2025, then traded roughly flat YTD in 2026 before this leg (yahoo/UBS)

Our calls (journaled)

• 2026-08-08: Gold's leg is a real-yield story, not a panic story; it belongs in the Cash/hedge sleeve as optionality against a yield or geopolitical shock. Invalidation: a decisive 10Y breakout with gold failing to hold a bid. Outcome: open (holding into 10-08; gold still bid at $4,400 with VIX sub-15)

• 2026-08-10: Gold holding a record while VIX is sub-15 is the cleanest confirmation yet that this is real-yield, not fear. Wednesday's CPI is the live test: a hot print that lifts real yields is the invalidation, an in-line one lets the structural bid run. Outcome: open (08-11 stress test passed: gold rose while the 10Y rose)

• 2026-08-11: The 08-10 tape was a live stress test and the thesis held. Gold up on a day the 10Y also rose separates the real-rate buyers from nominal-yield logic. The invalidation stays the same: a CPI that lifts real (not just nominal) yields with gold failing to hold. Outcome: open (08-12: thesis still holding; gold >$4,400 while 10Y eased to 4.69%).

• 2026-08-12: A geopolitical component (Iran) re-entered the gold bid alongside the real-rate story. For an allocator that does not change the sleeve logic - gold stays Cash/hedge optionality - but it does muddy the signal: some of this bid is now safe-haven, not pure real-rate, so read gold together with the 10Y and VIX, not alone. Cleanest confirmation of the core thesis would be gold holding after a soft CPI with yields steady. Invalidation unchanged: a hot CPI that lifts real yields with gold failing to hold. Outcome: VALIDATED 08-13 - the exact "cleanest confirmation" set-up occurred: soft CPI, 10Y steady at ~4.69%, gold up ~0.66% to two-month highs, no VIX spike. Real-rate driver confirmed, invalidation not triggered.

• 2026-08-13: The real-rate thesis is now confirmed by its own stated test, not just holding. Gold rallying on a soft CPI with flat nominal yields is the mechanism working: lower expected inflation with steady nominal = lower real yields = gold bid. Keep gold in the Cash/hedge sleeve as real-yield optionality; the Iran premium is a bonus, not the base case. Invalidation unchanged: a print (PPI/retail) that lifts real yields with gold failing to hold. Outcome: PARTIAL MISS 08-14 - the PPI print did lift implied real yields (nominals fell but expected inflation fell faster) and gold did fail to hold, dropping 1.31% and snapping its advance. Score this as a miss on the near-term call and a hit on the framework: we had written the invalidation in real-yield terms, not nominal, so the mechanism behaved exactly as specified even though the direction went against us. The error was assuming a soft print automatically means lower real yields.

• 2026-08-14: Restating the thesis with the ambiguity removed. Gold follows REAL yields, and the last two sessions gave both signs of the same equation: Wednesday's CPI lowered expected inflation less than it lowered rate risk, so real yields eased and gold rose; Thursday's flat PPI cut inflation expectations faster than nominal yields fell, so real yields firmed and gold fell 1.31% even as the 10Y dropped to 4.648%. Anyone reading gold off the nominal 10Y alone got Thursday exactly backwards. Second lesson: the Iran premium is a borrowed bid, and it left the same day oil did. Sleeve logic is unchanged, gold stays Cash/hedge optionality against a real-yield or geopolitical shock, and a 1.3% down day inside a two-month uptrend is not a thesis break. Invalidation, stated precisely this time: sustained disinflation that pulls breakevens down faster than nominals, with gold making lower highs against the $4,448 area. Watch retail sales and UMich Friday. Outcome: VALIDATED 08-15, and validated on the precise wording. We restated the invalidation in real-yield terms and specified that it required breakevens falling faster than nominals. Friday delivered the opposite configuration: nominals fell on the retail miss while UMich inflation expectations rose to 4.3%, so real yields fell and gold rose. The rewrite of the thesis was tested within one session and held. Note the discipline point: this is a hit on the framework and only a small one on price, gold closed +0.24% and is still below $4,448.

• 2026-08-15: With two clean sessions in opposite directions we can now state the rule without hedging: gold tracks real yields, and the only way to read it is nominal yields MINUS expected inflation, never nominal alone. Thursday and Friday are the matched pair that proves it. The second observation worth carrying is that the Iran premium is a fast-moving borrowed bid, it left with oil on Thursday and returned with oil on Friday, so any part of gold's level attributable to Hormuz should be treated as rentable, not owned. Sleeve logic unchanged: gold sits in Cash/hedge as real-yield optionality, sized as a hedge and not as a directional bet. One new caution: Kitco's survey has Wall Street fully bullish, which historically is a positioning risk rather than a signal, and it argues against adding into strength. Invalidation unchanged and still unmet: sustained disinflation pulling breakevens down faster than nominals, with gold making lower highs against the $4,448 area. Watch the Fed minutes and whether UMich inflation expectations keep rising. Outcome: open

• 2026-08-16: The thesis is unchanged and the week gives it two clean, separable tests. Test one is FOMC minutes on Thursday: if the committee reads as more hawkish than a market pricing the first hike in 2027, nominals rise without expected inflation rising and real yields firm, which is the configuration that beat gold on 14 August. Test two is the retail block plus flash PMIs: a weak consumer alongside a live energy tail is the falling-real-yield configuration that worked on 15 August. Because these can fire in the same week, do not read a single day in isolation, read which of the two inputs moved. The positioning caution from 08-15 stands and is now the more actionable point: Kitco's survey has Wall Street fully bullish, which argues against adding into strength, and gold has still not reclaimed $4,448. The Hormuz layer is escalating, but we are formally reclassifying it in the new energy-hormuz entity as rentable day to day and increasingly structural quarter to quarter, so a Hormuz-driven gold bid remains something to hold rather than to chase. Sleeve logic unchanged: Cash/hedge, real-yield optionality, sized as a hedge. Invalidation unchanged and still unmet: sustained disinflation pulling breakevens down faster than nominals, with gold making lower highs against the $4,448 area. Outcome: open

• 2026-08-17: Refining Wednesday's test rather than restating it. The minutes can move nominal yields through two different channels and gold cares which one. If hawkish minutes lift the FRONT end because the committee looks closer to a hike than the market's 2027 pricing, breakevens are unlikely to rise with them, real yields firm, and that is the 14 August configuration that cost gold 1.31%. If instead the long end cheapens further on term premium, deficits and balance-sheet speculation, with the 30Y already at 5.216% and the July deficit at $432B, nominals rise without expected inflation falling, real yields need not rise, and gold can hold. So the instruction is narrower than "watch the 10Y": watch WHERE on the curve the move happens. Positioning caution from 08-15 stands and is now two runs old without resolution, Kitco's survey has Wall Street fully bullish, which argues against adding into strength. Sleeve logic unchanged: Cash/hedge, real-yield optionality, sized as a hedge and not a directional bet. Invalidation unchanged and still unmet: sustained disinflation pulling breakevens down faster than nominals, with gold making lower highs against the $4,448 area. Outcome: open

• 2026-08-18: Yesterday's refinement was tested inside one session and passed on the precise wording, which is the second time in five days this entity has been validated on the exact clause we wrote rather than on direction. We said the minutes could lift nominals through two different channels and that gold cares which one, and that a long-end cheapening on term premium, deficits and balance-sheet speculation would let gold hold. The long end cheapened to a 19-year high on inflation worry and government borrowing, and gold rose 0.44%. Score it a HIT, with one honesty caveat: the test came from oil and the fiscal side rather than from the minutes, which have not landed yet, so Wednesday's front-end test is still ahead and still live. The framework now has a usable rule an allocator can carry: gold is short the real rate, so a nominal move that arrives WITHOUT a matching move in expected inflation is not a headwind, and the practical shortcut is to ask what caused the yield move before deciding whether it matters. The new nuance from Monday is that gold reads the two channels at different speeds, flinching intraday on a hot regional activity survey (a rate-expectations input) while ignoring a 19-year high in the long bond (a term-premium input). Two things to resist. First, do not let a validated framework become a bullish position: the Kitco survey still shows Wall Street fully bullish, three runs running, and gold has still not reclaimed $4,448, so the metal is being right without being strong. Second, the Hormuz component just changed from a headline bid to an expired-deadline bid, which is more durable but also more expensive, and our standing classification of it as rentable rather than owned is the discipline that stops us paying for it twice. Sleeve logic unchanged: Cash/hedge, real-yield optionality, sized as a hedge and not a directional bet. Invalidation unchanged and still unmet: sustained disinflation pulling breakevens down faster than nominals, with gold making lower highs against the $4,448 area. The specific way that could arrive Wednesday is hawkish minutes lifting the FRONT end. Outcome: open

Hit-rate

• 3 hits / 1 partial miss (4 closed). Latest: 08-17 "watch WHERE on the curve, a term-premium-driven nominal rise need not lift real yields" VALIDATED 08-18 when the 30Y hit a 19-year high on deficits and inflation worry and gold rose anyway. Prior: 2 hits / 1 partial miss (3 closed). Latest: 08-14 restatement of the thesis in real-yield terms VALIDATED 08-15 when falling nominals plus rising inflation expectations lowered real yields and gold rose. Prior: 08-12 "soft CPI + steady yields + gold up = real-rate confirmed" validated 08-13. 08-13 "soft print lets the bid run" partly missed 08-14 - gold fell 1.31% on a soft PPI because breakevens fell faster than nominals. Framework survived, the directional call did not.