COT Report: The Advanced Read
The advanced COT read begins where the first COT lesson ended: past the headline net positioning, into the disaggregated report that separates the market's participants by what they actually do with their positions, and into the financial futures data that most traders never open at all. The commitments of traders report publishes every Friday with each week's positioning through Tuesday, and its advanced use is not reading one number but reading a cast list: producers who hedge inventory, managed money that rents direction, swap dealers who absorb flow and hedge it, and the difference between what each group did is where the market's real information hides. The first lesson taught what the report is. This lesson teaches what it is for: separating the participants whose positioning reflects physical reality from those whose positioning reflects opinion, and reading the tension between them.
Three Reports, One Ledger
The CFTC publishes three versions, and the advanced reader uses them for different questions. The legacy report splits the market into commercial and non-commercial, the original split that made the report famous but blurs modern markets badly. The disaggregated report, the workhorse for commodity futures, divides participants into four functioning groups: producer, merchant, processor, and user on one side; swap dealers; managed money; and other reportables, so the hedger and the speculator finally appear under their own names. The traders in financial futures report applies the same disaggregation to the contracts that matter most to macro traders: equity indexes, Treasuries, and currencies, revealing how the fast-money funds and asset managers are positioned in the very markets the rest of this academy trades. The advanced read is the habit of asking all three: the disaggregated for commodities, the financial report for index and rate positioning, and the legacy only for continuity with the decades of history that predate the better versions.
| Report | Groups it separates | Best used for |
|---|---|---|
| Legacy | Commercial vs non-commercial | Long history, continuity |
| Disaggregated | Producers, swaps, managed money, other | Commodity futures |
| Financial futures | Dealer, asset manager, hedge-fund category, other | Indexes, rates, currencies |
The Disaggregated Split
The disaggregated structure exists because the groups' positions mean opposite things. Producers and merchants are the physical market: a gold miner's short position is not a prediction but a hedge on ore in the ground, and its size tracks planned production, not conviction. Managed money is the rent-a-direction crowd: hedge funds and other fast-money accounts whose longs and shorts are explicit bets on price, sized by conviction and reversed in weeks. Swap dealers sit between, absorbing client flow and hedging it, which makes their book a mirror of what the public is doing. When the groups agree, price is running with everyone's interest at once; when they split, with funds adding longs while producers add shorts, the market is showing you its own tension, and that tension is the report's real signal: a price being pushed by opinion against a supply that reality keeps selling into.
A Worked Example: Gold, Six Weeks
Follow one advance through the disaggregated gold data, week by week. Week 0: gold trades at 2,310, managed money holds a net long of 187,000 contracts, and producers are short 289,000. Over six weeks the price grinds to 2,392, and the ledger moves in a revealing pattern: managed money adds 27,000 contracts to reach a 214,000 net long, a classic momentum build; but producer shorts grow by 18,000 contracts over the same weeks, reaching 307,000, the physical side selling into the rally at an accelerating pace; open interest across the market expands 9 percent as both sides build. Two groups, one price, opposite behavior: the opinion side pressing the trend, the physical side pricing it as a gift to be hedged.
The tension resolves in the read, not the tape. Funds adding longs while producers add shorts is the signature of a rally being bought by traders who must chase and sold by participants who know their costs, and the historical record of that configuration is cautious for the trend's durability: the marginal buyer is renting, the marginal seller is hedging real supply. The confirmation arrives from the percentile math: the 214,000 contract net long sits at the 92nd percentile of five years of history, a level from which the crowd has rarely had room to add, and the price stalls accordingly, holding a 2,395 to 2,401 range for the next three weeks while managed money's position plateaus and then bleeds. The advance did not reverse on bad news; it stalled when the only group still able to buy had already bought, and the producer shorts, patient the whole way up, were positioned for exactly that exhaustion.
The third lesson of the example is the calendar discipline the report imposes. The COT data is Tuesday's positioning published Friday, stale before it arrives, and useless as a timing tool in isolation: the funds' 27,000 contract build was visible in the price on any chart without the report. The report's value is structural: it tells the trader who is on each side, how crowded the side they are considering joining has become, and how much fuel, fresh buying or short covering, remains. Read weekly, it converts the chart's "what happened" into the ledger's "who did it, and how much room is left."
Extremes and the Contrarian Turn
Positioning extremes are the advanced report's most cited export, and they deserve their careful definition: an extreme is a percentile, not a number. A 214,000 contract net long means nothing alone and everything at the 92nd percentile, because percentiles answer the only question that matters: how much positioning room does the crowd have left in the direction it is already leaning. Crowded longs mean the marginal buyer is gone and the covering fuel is large; crowded shorts mirror the logic downward. The report's history provides the percentiles, five years is the working baseline, and the extremes mark exhaustion zones where trends stall, not reversal signals that fire on their own: the crowd can stay crowded far longer than the impatient trader can stay short a rally, which is why the extreme is a condition to trade around, with tighter stops and smaller size, rather than a signal to trade against.
The financial futures report extends all of it to the markets most traders actually trade, and its weekly read has become a standing part of professional macro desks' panels: the hedge-fund category's net S&P positioning as the speculative temperature, asset managers' Treasury positioning as the real-money rate view, and the currency futures ledger as the crowd's dollar vote. The instrument changes, the method does not: separate the groups, watch the split, and let the percentiles define the room.
The COT ledger reads positioning in the futures market once a week. The next lesson moves to a faster market ledger: the options market's put/call ratio and open interest, published daily, and what the crowd's insurance buying reveals before price moves.
Advanced COT Questions
Why does the disaggregated report beat the legacy commercial/non-commercial split?
Because the legacy split lumps together participants whose positions mean opposite things: swap dealers hedging client flow are filed with speculators, and the signal drowns. The disaggregated report separates producers, swap dealers, and managed money by function, so the reader can finally distinguish a hedged producer's short from a fund's directional bet, which is the entire informational content of the ledger.
What does it mean when funds add longs while producers add shorts?
The market's opinion is running against its physical reality. Funds rent direction and reverse quickly; producers hedge real production and sell into strength because their costs tell them the price is generous. The worked example's gold rally carried exactly this split for six weeks, and it stalled at the 92nd percentile net-long extreme once the funds ran out of room, with the producer shorts positioned for the exhaustion.
How is a positioning extreme defined, and what does it predict?
A percentile of history, not a fixed number: the extreme is where net positioning sits within its own five-year range. Extremes mark exhaustion zones, levels where the crowd has little room to add and large fuel to cover, at which trends stall and risk-reward deteriorates for joining. They are conditions to trade around, not reversal signals, because crowded positioning can persist far longer than a counter-trader's patience.
The COT data is days old by the time it publishes. How can stale data be useful?
Because its value is structural, not chronological. The report answers who is on each side of the market and how crowded their side has become, questions that barely change in a week and cannot be answered by any price chart. Timing comes from the market; the ledger supplies the positioning context that tells the trader whether the timing signal is riding fresh flow or joining an exhausted crowd.