You read the COT report (the CFTC's Commitments of Traders, a weekly count of open futures positions at Tuesday's close) by subtracting each trader category's short contracts from its long contracts, then watching how those nets change from week to week. In the E-mini S&P 500 row as of October 13, 2026, Asset Managers were net long 915,835 contracts and Leveraged Funds net short 408,726. The catch: the report lands on Friday at 3:30 p.m. Eastern time, three days after the snapshot, and a day trader who is flat at the close is not in it.
For ES and NQ, read the CFTC's Traders in Financial Futures table and subtract each category's short contracts from its long contracts. In the ES row as of October 13, 2026, Asset Managers were net long 915,835 contracts and Leveraged Funds net short 408,726. The data is 3 days old at release and 10 days old when replaced, a trader flat at the close never appears, and Chen and Maher (2013) found that signals from a popular position-based sentiment index produced no significant average returns in S&P 500 futures.
What is the COT report?
The COT report is the Commodity Futures Trading Commission's weekly census of open futures positions, sorted by type of trader. According to the CFTC's About the COT Reports page, it breaks down "each Tuesday's open interest for markets in which 20 or more traders hold positions equal to or above the reporting levels established by the CFTC."
Open interest (contracts opened and not yet closed out, delivered or exercised) is the base, and every number in a row is a slice of it. The first modern version appeared in June 1962, according to the CFTC's 2026 Federal Register notice reviewing the program. It now comes in three families: the legacy report, the Disaggregated report for physical commodities, and the Traders in Financial Futures report for financial contracts such as stock indexes, currencies and Treasuries.
When is the COT report released, and how old is the data?
The CFTC releases the report every Friday at 3:30 p.m. Eastern time with positions from the previous Tuesday, so the data is 3 days old on arrival and 10 days old when the next report replaces it. In some federal holiday weeks the release moves to the following Monday, and the data is then 6 days old on arrival.
That is Eastern time, not EST. From March to November the US runs on daylight time, so guides that print 3:30 p.m. EST are an hour out for most of the year.
The snapshot is taken at the close on Tuesday, October 13, 2026. The sessions on Wednesday, October 14 and Thursday, October 15 trade in full, and most of Friday has passed when the report lands at 3:30 p.m. on October 16. The next snapshot is Tuesday, October 20, and its report, due Friday, October 23, retires this one at 10 days old.
The CFTC's release schedule warns that "Federal holidays may delay release by one or two days." For the rest of 2026 it lists Monday releases on November 16, November 30 and December 28, carrying data from November 10, November 24 and December 22 (our inference from its "previous Tuesday" wording, since the page prints no as-of dates). For exchange hours in those weeks, see the CME holiday trading calendar.
The format may change. On April 30, 2026 the CFTC asked for public comment on publishing the reports "on a more frequent basis" and changing their content, and Chairman Michael S. Selig said in the CFTC press release that "the Commission is examining the current structure and publication of our COT Reports" after outreach to the agricultural community and commercial end users. Comments closed on June 4, 2026, and as of October 16, 2026 we found no final decision.
Which COT report should you read for ES and NQ?
Read the Traders in Financial Futures report (TFF), futures only. Its four categories describe who holds index futures in more detail than the legacy split of commercial against non-commercial.
According to the CFTC's TFF explanatory notes, which date the report's announcement to July 22, 2010, it "separates large traders in the financial markets into the following four categories: Dealer/Intermediary; Asset Manager/Institutional; Leveraged Funds; and Other Reportables." Its history runs back to June 13, 2006.
Many guides still tell ES traders to compare commercials with non-commercials, which is only partly right. The notes say TFF is not a disaggregation of the legacy report: one TFF category can hold traders the legacy report calls commercial and others it calls non-commercial, so the two sets of labels do not map onto each other.
Gold is different. It sits in the Disaggregated report, whose categories are Producer/Merchant/Processor/User, Swap Dealers, Managed Money and Other Reportables, and its reporting level under 17 CFR 15.03 is 200 contracts.
What does each column in the COT report mean?
Each row splits one market's open interest five ways, four reportable categories plus Nonreportable, with long, short and spreading positions for each; below that come the week's changes, each figure as a percent of open interest, and the number of traders. The TFF notes define the categories by who the trader is:
- Dealer/Intermediary: the sell side, large banks and dealers in securities, swaps and other derivatives, which tend to run matched books.
- Asset Manager/Institutional: pension funds, endowments, insurance companies, mutual funds and managers with mostly institutional clients.
- Leveraged Funds: typically hedge funds and money managers, including registered commodity trading advisors (CTAs) and commodity pool operators.
- Other Reportables: large traders outside those three, mostly hedging business risk.
- Nonreportable: everyone below the reporting level, computed as a residual.
Spreading is a computed amount: the part of a trader's position that is long one contract month and short another, or futures against options, offset against itself. It sits on both the long and the short side, so it cancels out of any net. In the October 13, 2026 ES row, spreading totalled 192,432 contracts, 10.0% of open interest.
The trader counts never add up to the total. The ES row lists 444 total traders, but the category counts sum higher because a spreader also appears in the long or short count. A dot replaces any count below four traders, which is why Other Reportables spreading shows "." in ES.
The live TFF page shows only the latest week, so here is the row from the CFTC's archived TFF futures-only report as of October 13, 2026, published October 16. The next section nets each box.
How do you read the COT report step by step?
Find your market's row in the TFF futures-only report, subtract short from long for each category, divide by open interest, and compare with last week using the changes line. The six steps, worked on the October 13, 2026 ES row:
- Find the row. E-mini S&P 500 is code 13874A, the E-mini Nasdaq-100 (listed as NASDAQ MINI) is 209742, and the Micro E-mini S&P 500 has its own row, 13874U.
- Check the date and open interest. This row is positions as of October 13, 2026, with open interest of 1,931,607.
- Net each category. Long minus short, ignoring spreading. Leveraged Funds: 121,404 - 530,130 = -408,726.
- Scale it. Divide by open interest: -408,726 / 1,931,607 = -21.2%.
- Take the weekly change. Change in net = change in long minus change in short: 8,181 - 1,820 = +6,361, so last week's net was -408,726 - 6,361 = -415,087.
- Check who is underneath. Note the trader count (444) and the Nonreportable share (12.6% of long open interest) before reading anything into a category.
The table runs the same arithmetic for all five groups.
| Category | Long | Short | Net | Net % of OI | Weekly change in net |
|---|---|---|---|---|---|
| Dealer/Intermediary | 200,960 | 803,318 | -602,358 | -31.2% | -1,517 |
| Asset Manager | 1,110,384 | 194,549 | +915,835 | +47.4% | +9,220 |
| Leveraged Funds | 121,404 | 530,130 | -408,726 | -21.2% | +6,361 |
| Other Reportables | 62,418 | 73,497 | -11,079 | -0.6% | -9,387 |
| Nonreportable | 244,009 | 137,681 | +106,328 | +5.5% | -4,677 |
| Sum of nets | 0 | 0.0% | 0 |
Computed by Phoenix Technologies from the CFTC TFF futures-only report as of October 13, 2026 (changes from October 6, 2026); method: net = long minus short, share = net / open interest, weekly change = change in long minus change in short. Long and short sides each sum to exactly 1,931,607.
In ES, asset managers carry the net long while dealers and leveraged funds carry the net short. Each side totals exactly 1,022,163 contracts, because every long has a matching short.
Why does the weekly change matter more than the level?
The weekly change matters more because the level barely moves: between October 6 and October 13, 2026, the Asset Manager net long grew 1.0% and the Leveraged Funds net short shrank 1.5%. The level describes the market's structure; the change is the only new information each Friday.
The Leveraged Funds net moved from -415,087 on October 6 to -408,726 on October 13: 8,181 contracts of new longs against 1,820 of new shorts. The Asset Manager net rose from +906,615 to +915,835, up 9,220, while open interest slipped from 1,932,227 to 1,931,607, which the report prints as a total change of -620.
Size the change before you build a story on it: 6,361 contracts is 0.33% of ES open interest. Some guides treat a big single-week shift as proof of conviction, but the TFF notes warn that "staff classifies traders, not their trading activity", so a hedge fund in Leveraged Funds may be hedging rather than betting.
That settles the most repeated misreading: Leveraged Funds net short ES does not show hedge funds betting on a fall. A short future can be one leg of a position against cash stocks, the kind priced off basis and fair value, and the report cannot tell the two apart. NQ has a similar layout: in the NASDAQ MINI row (open interest 279,515), Leveraged Funds were net short 11,593, a net that rose 7,619 on the week (less short), against Asset Managers net long 75,323 and Dealers net short 77,932.
Who does the COT report leave out?
The COT report leaves out anyone who is flat at the close, and it folds everyone below the reporting level into one unnamed Nonreportable residual. For ES that level is 1,000 E-mini contracts; for other broad-based stock indexes, which on our reading of the rule includes the Nasdaq-100, it is 200.
The legal test in 17 CFR 15.00 is a position held "at the close of the market on any business day" at or above that level. Like the CME daily settlement price, it is an end-of-day number, so trades opened and closed inside the session never reach it. A 2004 CFTC rule, effective January 20, 2005, raised the E-mini level from 300 to 1,000 so that the full-size and E-mini levels "will be the same". As of October 2026, 17 CFR 15.03 still lists 1,000 contracts for the S&P 500 Stock Price Index.
Nonreportable is not measured directly. The CFTC's explanatory notes say it is "derived by subtracting total long and short 'Reportable Positions' from the total open interest." The CFTC prints no count of nonreportable traders, so no public data says how many people sit inside it.
Calling it small retail is only partly right. At an illustrative S&P 500 level of 6,000, one E-mini carries $300,000 of exposure ($50 x 6,000), so a holder of 999 contracts, about $299.7 million, is still nonreportable. At Tuesday's close the report sorts every holder into three groups: at 0 contracts you are absent, which is where a flat day trader sits; 1 to 999 E-minis is Nonreportable, 12.6% of ES long open interest on October 13, 2026; 1,000 or more is reportable and classified.
Micros have their own row. In the Micro E-mini S&P 500 (13874U, open interest 164,490), Nonreportable held 47.8% of long and 45.1% of short open interest on October 13, 2026, against 12.6% and 7.1% in ES. The S&P 500 Consolidated row (13874+) adds micros at one tenth, 1,931,607 + 164,490 / 10 = 1,948,056, and Nasdaq has its own consolidated row, 20974+. We could not verify the micro reporting level, so we do not state one; for sizes, see how ES, NQ, MES and MNQ compare.
Is the COT report useful for day trading?
Not as a timing tool. It measures positions at Tuesday's close, arrives three days later, excludes anyone flat at the close, and the more recent of the two S&P 500 futures studies below found no reliable tradable signal in the weekly report, let alone an intraday one.
The evidence is mixed and dated. Wang (2003), using legacy categories for S&P 500 futures, found large speculator sentiment worked as a continuation indicator and large hedger sentiment as a contrary one, and that "Small trader sentiment hardly forecasts future market movements." With later data, Chen and Maher (2013) found the commercial link unstable, hedge funds' edge visible only at high frequency, and a weekly report with a 3-day delay too late to capture it.
Their verdict on the most common retail use: "trading signals generated by a popular, position-based sentiment index do not produce significant average returns." We found no public study showing that TFF positioning predicts intraday ES or NQ moves, and no public source splits intraday volume by COT category. Hold COT claims to the same standard as chart signals tested in research.
So the smart-money-right-now framing fails twice: the categories describe who traders are, not whether they are right, and now was 3 to 10 days ago. For same-session positioning, use market internals like TICK and ADD or footprint charts and cumulative delta, and keep the releases that move ES on your screen. Before you act on a COT reading, run this checklist:
- If you are flat by the close every day, the report describes a population you are not in. Do not use it for entries.
- If you hold for weeks, track the weekly change in Asset Manager and Leveraged Funds nets as a share of open interest, not the raw level. Many guides also rank the net within its own range over a lookback window, often as a 0 to 100 COT index; that is a position-based sentiment reading, the same family of signal the research above found unreliable, so treat an extreme as context rather than an entry.
- If one week's change is under 1% of open interest (the Leveraged Funds move above was 0.33%), do not build a thesis on it until it repeats.
- If you trade MES, read row 13874U, where nearly half of open interest is nonreportable.
- If the release lands on a Monday, the data is already 6 days old.
- If someone sells you a COT signal, ask for the dated row and the test behind it.
Go deeper
- Market Internals for Index Futures: Reading TICK, ADD and VOLD Without Fooling Yourself
- Footprint Charts and Cumulative Delta Explained
- Does Technical Analysis Work? What the Research Says
- Why ES Does Not Match the S&P 500: Basis, Fair Value and the Premium You Are Trading
Frequently asked questions
How far back does Traders in Financial Futures data go?
To June 13, 2006. The CFTC launched the TFF report on July 22, 2010 and rebuilt the earlier years by applying recent trader classifications backwards, which its notes call backcasting. The notes say accuracy diminishes the further back you go, though the data back to 2006 should be reasonably representative.
Why is my futures market missing from the COT report?
Most likely it has too few large traders. The CFTC only breaks down open interest for markets where 20 or more traders hold positions at or above the reporting levels, so thinly held contracts get no category split.
Does an inter-market spread count as spreading in the COT report?
No. The TFF notes state that inter-market spreads are not considered; spreading covers offsetting positions across contract months, or futures against options, in the same market. The consolidated reports are different: a CFTC announcement of June 18, 2010 says a trader long the full-size S&P 500 contract and short an equivalent amount of E-minis shows up as spreads in the S&P 500 Consolidated report, because it treats the two markets as one.
Does the COT report include options?
Yes, in a separate version. The CFTC releases a futures-only report and a futures-and-options-combined report at the same Friday 3:30 p.m. Eastern time slot. The worked ES numbers on this page come from the futures-only version.
Who sends the position data to the CFTC?
Reporting firms do. They file daily reports showing the positions of traders above the reporting levels, and the CFTC says the weekly figures are "based on Tuesday position data that is sent to the Commission by reporting firms on Wednesday."
Can I see older COT reports?
Yes. The live TFF page always shows the latest week and rolls over every release, so a figure you read there today will be gone next Friday. The CFTC keeps weekly archive copies, for example the October 6, 2026 report, which is the one to cite when you need a dated reading.
What does a negative number in the COT report mean?
Positions themselves are never negative. Negative figures appear in the changes line, where a total change of -620 for ES means 620 fewer contracts were open than on October 6, 2026, and in the nets you compute, where a negative net means the category is net short.
Sources
- CFTC (2026), About the COT Reports
- CFTC (2026), Release Schedule
- CFTC (2026), Traders in Financial Futures - Futures Only Positions as of October 13, 2026
- CFTC (2010), Traders in Financial Futures Explanatory Notes
- CFTC (2026), COT Explanatory Notes
- CFTC (2026), CFTC Seeks Public Comment Relating to Commitments of Traders Reports (Release 9223-26)
- Legal Information Institute, Cornell Law School (2026), 17 CFR 15.00 Definitions
- CFTC via Federal Register (2004), Reporting Levels and Recordkeeping (Federal Register Vol. 69, 04-27750)
- Journal of International Financial Markets, Institutions and Money (2013), On the predictive role of large futures trades for S&P500 index returns: An analysis of COT data as an informative trading signal
- Applied Financial Economics (2003), Investor sentiment, market timing, and futures returns
- cot-reports.com (2026), COT Index: Normalized 0-100 Positioning for Every CFTC Market