Gold funds added longs before Friday’s selloff. The CFTC record stops on Tuesday.
Managed money increased its net long position in standard COMEX gold futures in the week to 27 October. The report reached readers on Friday. Anyone using it to identify Friday’s sellers needs to move the evidence date back three days.
- Managed-money net length rose by 1,356 contracts in the week to Tuesday, even as open interest fell.
- Outright shorts increased alongside longs, so short covering does not explain this weekly net increase.
- The report supplies positioning context going into Friday and leaves Friday’s sellers unidentified.
What changed in the managed-money book
In the CFTC’s disaggregated futures-only report, managed money held 122,508 outright longs and 16,946 outright shorts in COMEX gold on Tuesday 27 October. The subtraction gives 105,562 contracts net long. The report separately lists 40,502 spreading positions. The market row is standard Gold, contract code 088691, with 100 troy ounces per contract. Micro Gold has a different row.
From the prior week, outright longs increased by 1,538 contracts and outright shorts increased by 182. Net length therefore grew by 1,356. That decomposition matters. A rising net long can come from covering shorts or from adding longs. In this report both sides grew, and the larger addition was on the long side. The figures concern outstanding positions in the category, so they do not identify which individual managers changed direction.
Source notes: [1] Commodity Futures Trading Commission — COMEX Gold, futures only · [2] CFTC — Disaggregated explanatory notes
Use open interest to put the increase in proportion
Total open interest fell by 5,469 contracts to 396,810. The managed-money net position represented approximately 26.6% of that number. Reversing the reported weekly changes gives 104,206 net long against 402,279 open interest in the previous report, or about 25.9%. The rise in the share reflects both increased net length and a smaller market denominator.
Consider a counterfactual in which the previous net position had stayed constant while open interest still declined to 396,810. Its share would have risen to about 26.3% anyway. This calculation separates the sensitivity to the denominator from the outright change. Calling the actual 26.6% a share of total gold ownership would be a category error. A futures contract has counterparties on both sides, while physical inventories and other venues sit outside this table.
Source notes: [1] Commodity Futures Trading Commission — COMEX Gold, futures only
Friday’s headline belongs to a different information window
The CFTC timetable applicable to this report uses Tuesday’s positions and a scheduled Friday release at 15:30 Eastern. Reuters was already reporting sharply lower spot gold at 14:13 Eastern on Friday. There are two clocks here. Some of the price move preceded the public release, and all of Friday’s activity came after the report’s measurement cutoff.
The temptation is to construct a sequence in which those same funds bought early and then sold into the slide. The first snapshot supplies neither Friday’s position nor transaction timestamps. Even if the next report shows a decline in net length, that later comparison spans other sessions as well. I would describe Tuesday’s positioning as a possible vulnerability and leave the seller’s identity unresolved.
Source notes: [1] Commodity Futures Trading Commission — COMEX Gold, futures only · [3] CFTC — Commitments of Traders release schedule · [4] Reuters — Friday precious-metals market report
An upcoming schedule change will alter the comparison
On 27 October the CFTC announced a plan to move to two reports per week by year-end. Under the announced design, Monday observations would appear on Wednesday and Wednesday observations on Friday. The gold table examined here explicitly carries the Tuesday 27 October date. Researchers should use the observation and release dates on each actual report as the transition arrives.
In my worksheet I would keep the market code, report type, observation date and first public release time beside the gross long, gross short, spreading and open-interest columns. I would compare Tuesday with the previous Tuesday for this edition. When the frequency changes, a two-day difference and a seven-day difference deserve separate labels. Otherwise a faster publication schedule could be mistaken for a change in fund behaviour.
Source notes: [5] CFTC — Commission announces plan for semiweekly COT reports
Sources, dates & evidence limits
Sources checked 1 November 2026 at 04:09:10 UTC / 1 November 2026 at 08:09:10 Dubai. Source observation and release dates are separate from this article’s publication date. Linked data may later be revised.
- Commodity Futures Trading Commission — COMEX Gold, futures only ↗Positions as of Tuesday 27 October 2026; Friday 30 October release cycle
Rolling report. Locate the historical 27 October disaggregated futures-only Gold row under code 088691 when the current report advances.
- CFTC — Disaggregated explanatory notes ↗Methodology reviewed 1 November 2026; page time unspecified
Defines reportable trader groups and outright versus spreading positions.
- CFTC — Commitments of Traders release schedule ↗Schedule applicable to the week ended 30 October 2026, 15:30 Eastern regular release time
Scheduled release time, distinct from the position observation date.
- Reuters — Friday precious-metals market report ↗30 October 2026, spot observations at 18:13 UTC
Intraday US-dollar spot quotes and reported percentage moves. The prices are rounded as published.
- CFTC — Commission announces plan for semiweekly COT reports ↗27 October 2026; publication time not specified
Announced timetable for a planned frequency change by year-end. The report examined here retains its Tuesday observation date.
Follow the reading path
The Gold Trading Journal: Measure Decisions, Not Just Profit →Gold Trading & Research →Research Lab: test the evidence →CyGuru trading books: continue your study →Follow new Gold Intelligence articles →All 1 November 2026 articles →Gold Intelligence library →Original educational analysis. Illustrative calculations are not historical performance. Interpretation is identified in the text; no trade outcome is established by this article. Editorial standards →