The COT Report Has No "EUR/USD Net Short" — And Why That Matters

Short answer: the CFTC's Commitments of Traders report is published per futures contract, not per currency pair. There is a Euro FX contract, a British Pound contract and a Japanese Yen contract. There is no EUR/USD contract, no GBP/JPY contract and no EUR/GBP contract that most tools would need to justify the labels they use. When a free COT tool shows you "EUR/USD: net short 60,010 contracts", it is showing you the Euro FX contract on its own and silently discarding the dollar side of the trade.

That omission matters most in exactly the situations where COT is most useful. As of the report dated 11 August 2026, large speculators were net short every single non-dollar major. If you read those numbers as pair positioning, you would conclude the market was bearish on the euro, the pound, the yen, the franc, the loonie, the aussie and the kiwi simultaneously — which is not seven views. It is one view, expressed seven times: long dollars.

This article explains how the report is actually structured, how to compare both legs of a pair correctly, what the trader categories mean, and where COT stops being useful. The CFTC COT report analysis screen does the two-leg comparison automatically, free and without an account.

What the COT report actually is

Every Friday at 15:30 ET, the Commodity Futures Trading Commission publishes positioning data for US futures markets as of the preceding Tuesday. Two things follow immediately from that sentence and both are routinely forgotten:

1. The data is three days old on arrival. By the time you read Tuesday's positioning on Friday afternoon, three sessions have happened. In a fast market that is a lifetime.
2. It covers exchange-traded futures only. The FX futures market at the CME is a rounding error next to the roughly $7.5 trillion daily spot and forward market. COT is a sample of speculative positioning, not a census of it.

Within each contract, positions are split into three groups:

The most common misreading in retail COT commentary is treating commercials as "smart money" whose opposition to speculators is a warning. A commercial position that mirrors the speculative one is arithmetic, not information. Futures are a zero-sum contract: if speculators are net short 60,010 Euro FX contracts, somebody is net long 60,010 of them. Look at the table below and notice that the commercial column is close to the exact inverse of the non-commercial column in every row. That is not seven independent confirmations. It is the same number written twice.

The current picture, and why it proves the point

Latest report at time of writing — 11 August 2026. Net is non-commercial long minus short; Net % OI normalises it against open interest so contracts of different sizes are comparable.

Three readings fall straight out of this table that you cannot get from single-leg pair labels.

First: the raw contract counts are misleading and the percentage column fixes it. The Canadian dollar shows a net short of 173,362 contracts against the kiwi's 39,586 — more than four times larger. But as a share of open interest, CAD is −47.6% and NZD is −40.9%. They are comparably crowded. Raw contract counts compare contract sizes, not conviction. Any cross-contract comparison must use net % of open interest or a normalised COT index; never raw contracts.

Second: the dollar is the trade. With every non-USD major net short and the dollar index net long 43.2% of its open interest, this is a single macro position. A trader reading seven separate "pair" shorts might diversify across all seven and believe they have spread their risk. They have concentrated it.

Third: some of these are genuinely extreme and some are not. The euro at −7.5% of open interest is barely leaning. The Canadian dollar at −47.6% and the kiwi at −40.9% are heavily one-sided. Yet in raw contract terms the euro's −60,010 looks bigger than the kiwi's −39,586. This is the single most common error in COT commentary.

How to read a pair: both legs, separately

For a currency pair, the honest procedure has three steps.

Step 1: Resolve each leg to its contract

EUR/USD has a base leg (Euro FX contract) and a quote leg (the US dollar, proxied by the dollar index contract). GBP/JPY has a British Pound leg and a Japanese Yen leg — no dollar involved at all.

One subtlety that trips people up: for USD pairs, the non-USD contract effectively is the pair, because the futures contract is already quoted against the dollar. Euro FX futures are euros priced in dollars, which is EUR/USD. But the direction inverts depending on whether the dollar is the base or the quote. Net long Euro FX is bullish EUR/USD; net long Japanese Yen is bullish the yen, which is bearish USD/JPY. Getting this sign wrong flips your conclusion, and it is easy to get wrong when moving between EUR/USD and USD/JPY in the same session.

Step 2: Convert each leg into a comparable strength score

Raw net positioning is not comparable across contracts. Two normalisations are standard:

• COT index — where the current net position sits within its own range over a lookback period, expressed 0–100. A 52-week COT index of 90 means positioning is near the top of its own last-year range, whatever the absolute contract count.
• Positioning momentum — the recent change, which often matters more than the level. A net short that is rapidly shrinking is a different signal from a net short sitting still.

FXTerminal blends them at 70% 52-week COT index and 30% 4-week momentum to produce a per-leg strength score. The specific weights are a judgement, not a fitted result, and the screen shows both inputs so you can weigh them yourself.

Step 3: Subtract

pair bias = base leg strength − quote leg strength

With interpretation thresholds:

Why the subtraction is the whole point. Take GBP/JPY today. The pound is net short 21.9% of open interest and the yen is net short 10.7%. Read separately, both look bearish. Read as a pair, the pound is more heavily shorted than the yen, so the positioning skew on GBP/JPY leans bearish — but far less than the pound's standalone number implies, because the yen is crowded on the same side. Single-leg tools cannot express this at all.

When a leg has no contract

Not every currency has a CFTC-reported futures contract. The Swedish krona, the Norwegian krone and the Turkish lira do not. Gold and silver have their own contracts but they behave differently from currency legs.

When one leg is untracked, the honest answer is to fall back to a single-leg read and say so, rather than pretending a two-leg analysis happened. A pair analysis with one missing leg is a currency analysis wearing a pair's name — which is exactly the problem this article is about.

What COT is good for, and what it is not

It is good for: identifying crowded trades

The strongest, most repeatedly documented use of COT is spotting positioning extremes. When speculative positioning reaches the top or bottom of its multi-year range, the pool of traders left to push the move further is small, and the pool who must eventually unwind is large. That asymmetry does not tell you when, but it tells you which direction carries more fuel for a violent unwind.

The kiwi at −40.9% of open interest and the loonie at −47.6% are the current candidates. Not a signal to buy them. A signal that a dollar-negative catalyst would find a lot of stops in a hurry.

It is good for: sanity-checking a thesis

If your macro view is "the dollar rallies from here", the table above tells you the market already agrees with you, comprehensively. That does not make you wrong. It does mean you are being paid less for the view than you would have been three months ago, and that the downside if you are wrong is amplified by everyone else's exit.

It is not good for: timing

Positioning extremes can persist for months and get more extreme. Every cycle produces traders who shorted a crowded trade six weeks early and were carried out. COT tells you about fuel, never about the match.

It is not good for: intraday or short-swing decisions

Data as of Tuesday, published Friday, updated weekly. If your holding period is measured in hours, COT is noise with a delay.

It is not: the whole market

CME FX futures are a small slice of global FX. COT is a well-measured sample of one specific participant set. Treat it as a survey with excellent methodology and limited coverage.

COT versus retail sentiment: two different crowds

COT and retail long/short data are frequently conflated. They measure opposite ends of the market and often disagree, which is the useful part.

When professional positioning and retail positioning sit on the same side, that is genuinely unusual and worth investigating. When they oppose each other — which is the normal state — the interesting question is which one the price has been rewarding lately. You can put the two side by side using retail long/short positioning, which tracks 16 symbols including gold and silver at roughly 15-minute resolution.

A practical weekly routine

1. Friday evening (after 15:30 ET): pull the new report. Look at the net % of open interest column, not contract counts.
2. Rank the eight majors by how extreme each currency's own positioning is against its 52-week range. Note anything above 80 or below 20 on the COT index.
3. For each pair you actually trade, subtract the quote leg's strength from the base leg's. That is your positioning bias, and it is often much weaker than either leg alone suggested.
4. Check the direction of the 4-week change, not just the level. Positioning that is unwinding tells a different story from positioning that is building.
5. Cross-check against the week ahead. A crowded position going into a central bank decision or an inflation print is a different risk from a crowded position going into a quiet week. The forex economic calendar tells you which one you are in.
6. Ignore the commercial column unless you are specifically studying hedging flows. It is the mirror of the speculative column by construction.

Frequently asked questions

Is there a COT report for EUR/USD? No. The CFTC publishes positioning for the Euro FX futures contract and, separately, the US Dollar Index contract. There is no EUR/USD contract in the report. Tools that display "EUR/USD net short" are showing the Euro FX contract alone and dropping the dollar leg. For a true pair reading you need to compare both legs and subtract.

What does "non-commercial net short" mean? That large speculators — hedge funds, CTAs and managed money — collectively hold more short contracts than long ones in that currency's futures. As of 11 August 2026, non-commercials were net short 60,010 Euro FX contracts, equal to 7.5% of that contract's open interest.

Are commercial traders the smart money in the COT report? Not in any usable sense for FX. Commercials are hedgers with a business need in the underlying, and because futures are zero-sum, their net position is close to the arithmetic inverse of the speculative position. A commercial net long that "opposes" a speculative net short is not a contrarian signal — it is the same fact restated. Commercials are worth studying for hedging-flow analysis, not for direction.

Why should I use net percentage of open interest instead of contract counts? Because contract sizes and market depths differ enormously. The Canadian dollar's −173,362 net short looks four times larger than the kiwi's −39,586, but as a share of open interest they are 47.6% and 40.9% — comparably crowded. Raw counts compare contract sizes; normalised figures compare conviction.

What is a COT index and how do I read it? The COT index places the current net position within its own range over a lookback window, usually 52 weeks, scaled 0–100. A reading of 95 means positioning is close to its most bullish level of the past year; a reading of 5 means the most bearish. It makes different contracts directly comparable and is the standard way to identify extremes.

When is the COT report released? Every Friday at 15:30 US Eastern, reflecting positions as of the previous Tuesday's close. The three-day gap is structural and is the main reason COT is a positioning-context tool rather than a timing tool.

Can COT data predict currency reversals? It identifies conditions under which a reversal would be violent, not when one will happen. Extreme positioning means few traders remain to extend the move and many must eventually unwind. Extremes routinely persist for months and become more extreme first. Use it for risk context and position sizing, not entry timing.

How do I analyse a cross like GBP/JPY that has no dollar leg? Exactly the same way: resolve each leg to its own contract (British Pound and Japanese Yen), score each independently, and subtract. Currently both are net short — the pound at 21.9% of open interest and the yen at 10.7% — so the positioning skew leans bearish on the pair, but much less than the pound's standalone figure suggests. The CFTC COT report analysis screen resolves both legs automatically for any pair whose currencies have reported contracts.

Conclusion

The COT report is one of the few genuinely free, genuinely high-quality datasets in retail FX. It is also one of the most consistently mislabelled, and the mislabelling is not cosmetic — it changes conclusions.

Read as pairs, this week's data says the market is bearish on seven different currencies. Read correctly, it says the market is long dollars, that the position is most crowded in the loonie and the kiwi and barely present in the euro, and that the trade is now sufficiently one-sided that a dollar-negative surprise would have a lot of exits to clear.

Those are different pieces of information, and only the second one is in the data.

Both legs, normalised, subtracted, with the commercial column left alone. The full two-leg breakdown for every major pair is free at fxterminal.app/cot-analysis, including the per-leg trend charts and the all-currency ranking.