CFTC Gold Positioning vs Gold Price

The CFTC publishes COMEX Gold (GC contract, 100 troy ounces) Disaggregated Commitments of Traders data every Tuesday at 3:30pm ET. The managed-money category covers CTAs, CPOs, and hedge funds engaged in directional speculation. For most of the 2006 to 2020 era managed-money net positioning had a 0.7 to 0.85 correlation with the COMEX gold front-month price, with positioning typically leading price by 1 to 4 weeks at cycle inflection points. The series produced the canonical pattern at the 2011 peak: net long positioning expanded toward 290,000 contracts in early September 2011 as gold rallied from $1,500 in May to $1,920 on September 5, the previous all-time intraday high.

The September 2011 episode produced the textbook unwind. Gold fell from $1,920 to $1,535 by late December 2011 (a 20 percent decline in 14 weeks), with managed-money positioning collapsing from approximately 290,000 to 110,000 contracts (62 percent reduction). The 2011 to 2015 bear market that followed saw positioning fall to a record net-short reading in July 2015 (approximately minus 35,000 contracts), the only sustained net-short episode in the post-2006 record, alongside gold's $1,053 cycle low in December 2015. The 2011 to 2015 cycle is the cleanest demonstration of positioning's predictive power: extreme readings at both ends marked the price extremes within 4 to 8 weeks. Gold-fund AUM also followed the positioning cycle, with SPDR Gold Shares (GLD) holdings peaking at 1,353 tonnes in December 2012 before falling to 633 tonnes by year-end 2015.

Gold's rally from $2,000 in early 2024 to $4,746 in April 2026 (a 137 percent gain in 28 months) is the second-largest 28-month gold move on record after the 1979 to 1980 episode. The Lawrence Williams series at GoldChartsRUs and Bloomberg COT data both show that managed-money net long positioning never approached the September 2011 record of 290,000 contracts during this rally. The April 2026 reading sits in a 175,000 to 225,000 range, well below the 290,000 peak and not at the kind of extreme that would historically signal an imminent unwind.

The break has a specific structural cause: central-bank buying. The World Gold Council documented official-sector purchases of more than 1,000 tonnes per year in 2022, 2023, and 2024, the first three-year run above that threshold in WGC records. The National Bank of Poland alone added 102 tonnes in 2025, taking reserves to 550 tonnes. Central banks acquire physical gold in over-the-counter London markets and through the Bank for International Settlements, completely outside the COMEX futures complex. The mechanical effect is that the price-discovery channel that historically routed through speculative positioning has been augmented by a structural buyer that does not appear in CFTC data. The April 2026 configuration with gold at $4,637 and managed-money positioning moderate is the visible signature of this structural shift. The 2022 acceleration in central-bank buying coincided directly with the post-Russian-invasion sanctions on the Russian central bank, which froze approximately $300 billion of reserves and pushed several emerging-market central banks to diversify away from Treasury holdings into physical gold.

August 2018 produced the cleanest historical demonstration of how CFTC data can mislead at structural breakpoints. Swap dealers (the producer-hedging category) reached a record net-short position of approximately minus 195,000 contracts in early August 2018 after gold fell from $1,365 in April to $1,167 by August 16. Wolf Street and several gold-market analysts documented the swap-dealer short as the single largest net-short reading in the COMEX series since 2006. The conventional read was that swap-dealer record shorts marked an imminent price reversal because the producer-hedging book is structurally short, and an extreme net-short signaled commercial exhaustion.

That read was correct. Gold rallied from $1,167 in August 2018 to $1,560 by August 2019, a 34 percent gain in 12 months, with swap-dealer positioning normalizing toward minus 100,000 contracts by mid-2019 as gold rose. The 2018 episode is the canonical example of using swap-dealer rather than managed-money positioning to identify cycle troughs. The translation to the April 2026 setup is informative: managed-money net long positioning at moderate levels would historically suggest gold has further upside before any speculative-driven peak, but the structural-buyer framework means traditional positioning signals are less reliable at this point in the cycle than at any time in the post-2006 series. Cross-checking against silver positioning is the disciplined adjacent test: silver typically tracks gold's positioning cycle with a 1 to 3 week lag and amplifies the moves through higher implied volatility, so silver's April 2026 managed-money reading and the silver-gold ratio (running near 75 in late April 2026) provide important corroborating signals.

Through 2006 to 2024 spot gold's price was best explained by three macro factors: 10-year TIPS real yields, the trade-weighted dollar, and managed-money positioning. Real yields and gold ran a near-perfect minus 0.85 correlation: when 10-year TIPS yields fell from 0 to minus 1.10 percent in mid-2020, gold rallied from $1,500 to $2,070; when TIPS yields rose to plus 2.50 percent in late 2023, gold drew down to $1,820 from the May 2023 high. Each 100 basis-point move in the 10-year TIPS yield produced approximately a $200 to $300 move in gold in the inverse direction.

From early 2024 forward the relationship inverted. The 10-year TIPS yield held in a 1.80 to 2.20 percent band through 2024 to 2026 (well above the historical levels that would support gold rallies) while gold rose from $2,000 to $4,746. The dollar (DXY) ran 100 to 105 through the same window, neutral to mildly bullish, providing no obvious tailwind. Real yields and the dollar both predicted gold should be flat to lower over 24 months; gold doubled instead. The mechanical residual is the central-bank buying channel. The pre-2024 framework (real yields + dollar + positioning) explained 75 to 80 percent of gold's variance; the post-2024 framework requires adding a fourth variable for official-sector demand, which the World Gold Council Quarterly Gold Demand Trends report tracks with a one-quarter lag.

Gold rallied from $1,500 in early 2020 to $2,070 in August 2020 (a 38 percent gain in 7 months) on COVID-driven real-yield collapse, dollar weakness, and ETF inflows. Total gold ETF holdings reached 3,929 tonnes in early November 2020, a record at that time. Managed-money net long positioning reached approximately 250,000 contracts in February 2020, retreated during the March 2020 risk-off liquidation event, then rebuilt toward 230,000 by late summer 2020. The COVID rally was the textbook script: positioning, real yields, and ETF flows all moved in the same direction with gold price, producing a high-conviction macro trade that could be sized off CFTC data with reliable signal.

The 2024 to 2026 rally has produced no comparable pattern. ETF holdings actually declined through much of 2024 (Western flows turned net negative as 10-year real yields stayed elevated) before reversing in 2025. Managed-money positioning never reached the 2020 levels despite gold doubling. The structural buyer (central banks) is the dominant marginal participant for the first time in the post-2006 record. For the cross-asset desk this means traditional positioning analytics must be supplemented with WGC central-bank purchase data, IMF International Financial Statistics gold-reserve series, and the Bank for International Settlements Triennial Survey for the complete picture. The April 2026 configuration cannot be interpreted using the 2011 or 2020 templates. WGC October 2025 data shows GLD ETF holdings rising back to 3,200 tonnes from a 2024 trough near 2,750 tonnes.

Three observations frame the operational read. First, the 2025 deceleration in central-bank buying matters: WGC reported 863 tonnes purchased in 2025, below the 1,000-plus tonne threshold of 2022 to 2024. The 2025 figure represents a roughly 15 percent year-over-year decline that the WGC attributed to official sectors becoming more cautious at elevated valuations. If the 2026 first-half data confirms continued deceleration, the structural-buyer support fades and the price is more vulnerable to managed-money-driven correction than the 175,000 to 225,000 contract net long would suggest.

Second, the WGC 2025 Central Bank Gold Reserves Survey showed 95 percent of respondents expecting global official gold reserves to increase over the next 12 months, the highest reading in the survey's eight-year history, and a record 43 percent of central banks indicated plans to increase their own holdings (up from 29 percent in 2024). This suggests structural demand continues but with more dispersion across buyers. Third, the April 2026 spot price at $4,637 is roughly 130 percent above the long-run real-yield-implied fair value of $2,000 to $2,200, meaning the structural-premium component is the largest in series history. Reading the pair correctly requires watching the WGC quarterly release alongside the Tuesday CFTC update, with the Bank of Poland and Kazakh National Bank monthly reserve disclosures as the highest-value individual data points. Long gold versus short equivalent-duration TIPS captures the structural-premium thesis cleanly; the trade has worked through 2024 to 2026 with low drawdown despite extreme valuation.