In four weeks during late summer 2026, managed money in CBOT corn went from a net long of 166,770 contracts to 431,062. Over the same four weeks, commercial hedgers went from 481,831 short to 764,734 short.
Both sides more than doubled down. And then something worth noticing happened to the price.
In the week ending September 1, funds bought another 54,549 contracts — a substantial add by any normal measure — and September corn finished the week exactly where it started. Three weeks earlier, a comparable wave of buying had moved the board sharply higher. Same direction, same kind of money, much less result.
A crowded position does not predict direction. Peer-reviewed research on ten agricultural futures markets found very little evidence that trader positions forecast returns. What a crowded position does change is the shape of the risk: how far a market can move on a surprise, how quickly, and which direction has more fuel behind it. That matters most when a scheduled USDA report is on the calendar, because the date is known in advance and the positioning is already in place when it arrives.
What “crowded” actually means
The Commitments of Traders report breaks down each Tuesday’s open interest in markets where 20 or more traders hold positions at or above CFTC reporting levels. It is released the following Friday, usually at 3:30 p.m. Eastern.
The disaggregated version splits reportable positions into four groups: Producer/Merchant/Processor/User, Swap Dealers, Managed Money, and Other Reportables, with nonreportable positions shown separately. A feedlot, grain elevator, or merchandiser may fall into the Producer/Merchant category if its predominant business and futures use meet the CFTC’s criteria — which is why we call that column “commercials” and treat it as the physical trade.
Iowa State Extension makes a caution worth carrying: each category holds a variety of traders, and the classifications should not be read as a clean statement of every trader’s intent. The columns are useful in aggregate, not as a roll call.
The CFTC does not publish a “52-week rank” or a “COT Index.” Both are third-party constructions, and providers do not agree on how to build them.
The most common published convention is a min-max index: current net position minus the period minimum, divided by the range, then multiplied by 100. Under that scale, 100 is the largest net long in the lookback, and 0 is the largest net short. Several data providers describe it that way.
Ordinal ranks are a different animal, and the publisher has to tell you which end is which. On this page and in our weekly reports, rank 1 is the largest net long of the past 52 weeks and rank 52 is the smallest. A soybean managed-money rank of 1 means funds are more net long than at any point in the year. A live cattle rank of 52 means the opposite.
If you compare our numbers against another provider’s, check their convention first. The label is not self-explanatory, and the two scales run in opposite directions.
Start with what the research says, because it is not flattering
Any honest guide to positioning data has to begin here.
Sanders, Irwin and Merrin examined the forecasting content of COT positions across ten agricultural futures markets. Their bivariate Granger-causality tests found very little evidence that trader positions were useful in forecasting or leading returns. They found substantial evidence of the reverse: traders respond to price changes, with non-commercial traders displaying trend-following behavior. Their conclusion was that the results generally did not support using COT data to predict agricultural futures price movements.
In their tests, they rejected the hypothesis that non-commercial positions lead returns at the 5% level in only one of the ten markets. Commercial positions showed some forecasting usefulness in three of ten.
An earlier study reached the same place from a different angle: it found statistically significant contemporaneous correlations between net positions and returns, but positions did not Granger-cause returns. Returns led positions. Positive returns were associated with increased non-commercial and decreased commercial positions the following week.
Read that plainly. The best available research says funds mostly follow price rather than lead it, and that COT positioning is not a forecasting tool. Anyone selling you a COT-based buy or sell signal is arguing against the literature. We publish positioning weekly and we still say the same thing every week: positioning describes the market; it does not predict it.
So what is positioning good for?
Description. Which is more useful than it sounds, because it tells you how the market is loaded when news arrives.
Think of it as the difference between a weather forecast and a river gauge. The forecast tries to tell you what happens next. The gauge tells you how full the channel already is. Neither predicts the storm, but only one tells you what a given rainfall would do.
A market where funds hold their largest net long of the year has a specific property: the buyers are already in. New buying has to come from someone who has not bought yet. And if the news disappoints, the selling pressure comes not just from new sellers but from existing longs deciding to leave. That is not a forecast. It is arithmetic about who is standing where.
The fading-response tell
The most useful signal in a crowding situation is not the size of the position. It is how much price the position is still buying.
Here is the corn sequence from our own weekly reports in late summer 2026:
| COT week | Managed Money net | Contracts bought | Commercials net |
|---|---|---|---|
| Aug 11 | +166,770 | — | -481,831 |
| Aug 18 | +250,505 | +83,735 | -540,774 |
| Aug 25 | +376,513 | +126,008 | -677,164 |
| Sep 1 | +431,062 | +54,549 | -764,734 |
Through the middle of that run, heavy fund buying coincided with a market that rose sharply — December corn added more than 25 cents in the week ending August 21. By the week ending September 4, funds had bought another 54,549 contracts and September corn closed the week unchanged.
In our opinion, that is the moment worth flagging in any crowded market. When fresh buying stops producing fresh price, the position stops being fuel and starts being weight. It does not tell you the top is in. It tells you the marginal buyer is getting expensive, and that the next surprise has less friendly money left to absorb it.
Why the other column matters more when specs are extreme
Notice what the commercial column did across those same four weeks. As funds bought 264,000 contracts, the physical trade sold roughly 283,000 into them.
Those are the people who own the bushels. When they hedge that aggressively into a rally, they are making a statement about the level. The research is kinder to this column than to the other one — commercial positions showed some forecasting usefulness in three of the ten markets Sanders and colleagues tested, versus one for non-commercials. Not a signal. But when specs are at a yearly extreme and the physical trade is leaning hard the other way, the disagreement itself is information about the price you are being offered.
What the funds themselves do before a scheduled report
This is where the calendar comes in — and where the historical record is genuinely instructive, provided you read it for behavior rather than for outcomes.
Some of the largest documented positioning swings in corn have happened in the days immediately before a USDA release. Over several years, they fall into three recognizable patterns. None of them predicts what price did afterward. All of them tell a producer something about how the market arranges itself as a known date approaches.
Why a WASDE interacts with positioning at all. A USDA report does not move price because of what it says. It moves price because of the gap between what it says and what the market had already assumed — and positioning is the visible record of that assumption. A large net long is a large bet that supply will tighten, or demand will hold. When the report confirms it, there is less new buying left to do. When it contradicts it, many contracts have to change hands quickly. The report supplies the surprise; the positioning determines how much has to move because of it.
What that adds up to
Big money treats a USDA report date as something to get ready for. It will move an enormous number of contracts in a single week to do it — cut a quarter of a position, or switch sides entirely, in seven days.
What you should not take from these examples is a rule about what price did afterward. We looked for that and couldn’t find it documented, and we would be careful with anyone who claims otherwise. Four examples are not a pattern; the reports were different, and the broader research says positioning does not forecast returns. The useful lesson is simpler: the position you read on Friday afternoon may not be the position that meets the next report. Funds have four trading days to change their minds before you ever see the number.
What a report day actually looks like
Agricultural economists have spent thirty years measuring what USDA reports do to grain markets. The findings come out as plain numbers a producer can plan around.
This one costs money if you have it backward. In a 2025 study of weekly grain options, corn implied volatility in weeks with a major report pending averaged 25.6% while the volatility that actually happened averaged 16.8% — a gap of 880 basis points, roughly double the 420-point gap in ordinary weeks. Separately, implied volatility fell an average of 0.7 percentage points in corn and 0.8 in soybeans right after a WASDE, and one 2025 study found it stayed low for up to five days afterward.
Put plainly: the market charges you extra to own a floor across a report, and hands some of it back once the report is out. If the report isn’t what you are worried about, the days after are cheaper than the days before. If the report is what you are worried about, you are paying up for a reason — just go in knowing you are paying up.
And here is the honest gap. We went looking for research connecting how crowded the funds are before a report to how big the move is after it. We could not find any — no peer-reviewed paper, no extension publication with a validated model.
It stands to reason that a market where more contracts have to change hands would move further. But that is our reasoning, not a measured finding, and we would rather say so than dress it up. What is measured is that a report day runs about seven times as long as a normal day either way. That is enough to plan around without anyone guessing at the rest.
The mechanical risk: price limits and locked markets
A crowded position and a scheduled report create one practical hazard that has nothing to do with forecasting: you may not be able to trade when you want to.
CME price limits cap how far a futures price can move from the prior day’s settlement. A contract cannot trade above or below that limit. For grain and oilseeds, expanded limits are generally triggered when a listed non-spot contract month within the first five to eight months settles at the initial limit, and the expanded limit runs roughly 50% higher than the daily limit, staying in place until no contract settles at limit. For live cattle and feeder cattle, CME has maintained expanded limits at 150% of the initial limit.
Limit values are contract-specific and change through exchange notices — CME reset live cattle and feeder cattle limits effective June 1, 2026, for example. Check the current price-limit page for the date that matters to you rather than relying on a number you remember.
What a locked market means for an open position. A price limit restricts the prices at which new trades can occur. It does not, by itself, close your position. If the market is at the limit and there are no executable orders on the other side, an offsetting trade may simply not be available — and the position stays open and exposed until it can be traded, offset, or otherwise disposed of under your brokerage and exchange arrangements.
Margin requirements, risk controls, and liquidation rights vary by brokerage agreement and clearing arrangement. If you carry futures positions through report days, that conversation belongs with your broker before the report, not during it.
Four questions to run when positioning looks one-sided. None of them produce a price forecast. Together they describe the risk you are actually carrying.
If the answer to all four points the same way, that is not a signal to trade. It is a reason to have your own numbers decided before the date arrives.
What a producer actually does with this
Nothing in a COT table tells you to sell bushels. Your break-even does that, along with your basis and cash-flow calendar. Crowded positioning changes when you should make the decision, not what the decision is.
Decide before the date, not during it. A report drops at a fixed time. If your marketing plan calls for pricing a tier at a given level, that order can be resting beforehand. A decision made in the twenty minutes after a release is a decision made in the worst available conditions — fastest price movement, thinnest liquidity, and the possibility of a limit move.
Separate the futures decision from the basis decision. Crowded speculative positioning is a futures phenomenon. Your cash bid is futures plus local basis, and basis answers to harvest flow, elevator space, processor demand, and freight. A crowded fund long can unwind without your basis moving at all — or basis can weaken while the board holds. Our guide to cash versus futures covers both.
Size positions against what you can defend. If you carry futures or options through a report, the question is not what you think the number will be. It is what happens to your margin and your ability to exit if the market moves the limit against you and stays there. Discuss that with your broker in advance.
Watch the reaction, not just the number. A market that has been described as crowded and then fails to fall on unfriendly news — or fails to rally on friendly news — has told you something the position size alone could not. That is the same fading-response idea, applied to the report itself.
Frequently asked questions
Does a crowded fund position mean the market is about to reverse?
No. Research across ten agricultural futures markets found very little evidence that trader positions forecast returns, and substantial evidence that traders respond to price rather than lead it. A crowded position describes how the market is loaded, not where it is going. Crowded markets can stay crowded and keep trending for weeks.
What is a 52-week rank in COT data, and who calculates it?
The CFTC does not publish one. Fifty-two-week ranks and COT Index measures are third-party constructions. The most common published version is a min-max index where 100 is the largest net long in the lookback, and 0 is the largest net short. Ordinal ranks vary by provider, and the publisher has to disclose which end is which. On our pages, rank 1 is the largest net long of the past 52 weeks and rank 52 is the smallest.
Why does the timing gap in the COT report matter?
The report captures each Tuesday’s open interest and is released the following Friday, usually at 3:30 p.m. Eastern. Anything that happens between Tuesday and Friday is not in the numbers. Documented cases show large speculative money repositioning sharply around scheduled report dates — in one week of January 2023, managed money cut its corn net long by roughly a quarter ahead of USDA data. The position you read on Friday may not be the position that meets the next report.
How much does a USDA report actually move the corn market?
A twenty-one-year study found that WASDE releases carrying NASS crop production estimates produced corn price variance about 7.38 times normal and soybean variance about 6.87 times normal. When several reports land together, such as the January Grain Stocks, Crop Production Annual Summary, and WASDE cluster, corn variance ran roughly 7.7 times normal. Across 74 major report dates in a separate study, the average daily high-low range in nearby corn futures was about 14 cents. A report day is a different kind of trading day, and the date is known months in advance.
Are grain options more expensive before a USDA report?
Yes. In a 2025 study of weekly grain options, corn implied volatility in weeks with a major report pending averaged 25.6% while realized volatility averaged 16.8% — a gap of 880 basis points, roughly double the 420-point gap in ordinary weeks. Implied volatility then fell an average of 0.7 percentage points in corn and 0.8 in soybeans immediately after a WASDE, and one study found it stayed low for up to five days. In practical terms, the market charges a premium to own a floor across a report and gives some of it back once the report is out.
Which is more useful, the commercial column or managed money?
They answer different questions. Managed money shows where speculative risk is concentrated. Commercials — producers, merchants, processors and users — show what the physical trade is doing with the price on offer. In the Sanders, Irwin, and Merrin tests, commercial positions showed some forecasting usefulness in three of ten markets, versus one of ten for non-commercials. Neither is a signal, but the physical trade’s behavior is worth respecting when it disagrees with the funds.
What happens to my hedge if the market locks limit after a report?
A price limit restricts the prices at which new trades can occur; it does not close your position. If there are no executable orders on the other side at the limit, you may not be able to offset until the market trades again. Grain and oilseed expanded limits run about 50% above the daily limit; live cattle and feeder cattle expanded limits have been maintained at 150% of the initial limit. Margin and liquidation terms vary by brokerage agreement, so have that conversation with your broker before a report week.
If positioning does not predict prices, why follow it at all?
Because it describes the conditions your marketing decisions are made in. Knowing that funds hold their largest long of the year, that the physical trade is selling into it, and that a WASDE lands Friday does not tell you what price will do. It tells you that the move could be larger than usual in either direction, that liquidity may be worse than usual afterward, and that a decision made in advance is likely to be a better decision than one made in the minutes after the release.
Sources
- CFTC, About the Commitments of Traders Reports. Tuesday open interest, the 20-trader reporting condition, and report structure. cftc.gov
- CFTC, Commitments of Traders Release Schedule. Friday 3:30 p.m. Eastern release timing. cftc.gov
- CFTC, Disaggregated Explanatory Notes. Definitions of Producer/Merchant/Processor/User, Swap Dealers, Managed Money, and Other Reportables. cftc.gov
- Sanders, Irwin & Merrin, “Smart Money: The Forecasting Ability of CFTC Large Traders in Agricultural Futures Markets,” Journal of Agricultural and Resource Economics, August 2009. Granger-causality tests across ten agricultural markets. AgEcon Search
- “Does the CFTC Commitments of Traders Report Contain Useful Information?” February 2000. Finding that returns lead positions rather than the reverse. RePEc
- Iowa State University Extension, “Watching ‘smart money’ commodity trading could pay,” June 2023. Caution on trader-category composition and report timing. extension.iastate.edu
- Reuters, January 17, 2023. Managed money reducing corn and soybean risk ahead of January USDA data. reuters.com
- Reuters, August 14, 2023. Managed money corn position reversal ahead of August USDA data. reuters.com
- Reuters, July 15, 2024. Record managed-money net shorts in CBOT corn and soybeans. reuters.com
- Isengildina-Massa, Irwin, Good & Gomez, “The Impact of Situation and Outlook Information in Corn and Soybean Futures Markets: Evidence from WASDE Reports,” Journal of Agricultural and Applied Economics 40(1), 2008. Source of the 7.38x corn and 6.87x soybean report-day variance figures, 1985–2006. AgEcon Search
- Isengildina-Massa et al., “When does USDA information have the most impact on crop and livestock markets?” Journal of Commodity Markets, 2021. Corn, soybeans, wheat, cotton, live cattle and lean hogs, 1985–2018; source of the ~7.7x January report-cluster figure. ScienceDirect
- Isengildina-Massa, Irwin, Good & Gomez, “Impact of WASDE Reports on Implied Volatility in Corn and Soybean Markets,” Agribusiness 24(4), 2008. Source of the 0.7 and 0.8 percentage-point implied-volatility declines. Wiley
- Diersen, “Weekly Options on Grain Futures,” Journal of Agricultural and Applied Economics, February 2025. Source of the 25.6% implied versus 16.8% realized report-week figures and the 14-cent average range across 74 major report dates. Cambridge Core
- Ying, Chen & Dorfman, “Flexible Tests for USDA Report Announcement Effects in Futures Markets,” American Journal of Agricultural Economics, 2019. Source of the finding that report impact has grown over time while Crop Progress impact has declined. Wiley
- Virginia Tech Department of Agricultural and Applied Economics, “Do USDA Reports Move the Markets?” Plain-language synthesis of the report-impact literature, including the Garcia, Irwin & Leuthold (1997) finding that unanticipated Crop Production information explained roughly 31–45% of the price change that followed. aaec.vt.edu
- Bunek, “Characterizing the Effect of USDA Report Announcements,” 2015, and Bunek et al., “Does Public Information Facilitate Price Consensus?” 2024. Findings that volatility rose rather than fell after USDA wheat reports, and that releases did not meaningfully move volatility expectations. RePEc
- CME Group, Grain and Oilseed Price Limit FAQ, May 3, 2021. Daily and expanded limit mechanics. cmegroup.com
- CME Group, Price Limits: Ags, Energy, Metals, Equity Index, August 9, 2026. Current contract-by-contract limit values. cmegroup.com
- CME Group, Livestock Market Enhancements and Resetting of Price Limits for Live Cattle and Feeder Cattle Futures, notice dated May 14, 2026, effective June 1, 2026. cmegroup.com
Methodology: Positioning figures in the corn example are from our weekly Grain & Livestock COT Report and reflect Producer/Merchant net as commercials and Managed Money net as non-commercials, futures only. Contract months are labeled where price is referenced because September and December corn are separate contracts and their prices are not interchangeable. Historical episodes are cited for the pre-report positioning they document; we do not attribute subsequent price moves to those positions, because the research does not support that causal claim. Price limits are contract-specific and change through exchange notices—verify current values against CME for the applicable date.
Disclosure: This material is for educational and informational purposes only. It is not a recommendation to buy or sell any commodity, futures contract, option, cash contract, or other financial product, and it is not a trading signal. Futures and options involve substantial risk and are not suitable for every producer or investor. Past performance is not necessarily indicative of future results.
Optimus Futures does not maintain a research department as defined in CFTC Rule 1.71. Ag Optimus is a registered DBA of Optimus Futures LLC [NFA ID 0481133]. All trading decisions remain yours.
