Every week the US Commodity Futures Trading Commission publishes the Commitments of Traders report, usually called the COT. Commentators quote it constantly: "speculators are the most short euro in years". It is a genuine, free source of positioning data, but it has important limits. This guide explains what it contains and how to use it with realistic expectations.
What the report is
The COT summarises open positions in US-listed futures, including currency futures traded on the Chicago Mercantile Exchange such as euro, yen, sterling, Canadian dollar, Swiss franc, Australian dollar and others. The data is a snapshot taken on Tuesday and released on Friday afternoon US Eastern time. So by the time you read it, the numbers are at least three days old.
The categories
The report comes in several formats. The most common for currencies are:
- Legacy: splits traders into "commercial" and "non-commercial" (large speculators), plus a small "non-reportable" group.
- Disaggregated and Traders in Financial Futures (TFF): a finer split. For currencies, the TFF breaks out dealer/intermediary, asset managers, leveraged funds and other reportables.
Leveraged funds (hedge funds and similar) are the group most often watched as the "speculative" position. Asset managers tend to hold longer-term exposure, and dealers are mostly on the other side of client trades.
Net positions
For each group the report gives long and short contracts. The net position is longs minus shorts. A net long of 50,000 contracts in euro futures means that group holds that many more long contracts than short ones. The change from the previous week is often more informative than the level, since it shows whether positioning is building or unwinding.
Why it is useful
- Crowding. When one side is very large relative to its own history, a surprise in the opposite direction can force quick position reductions, which can exaggerate moves. This is the "squeeze" idea.
- Context for news. The same data surprise can produce different reactions depending on how the market was positioned beforehand (interest rate expectations).
- Regime awareness. Sustained positioning shifts often accompany changes in rate outlook (central banks and FX).
Why it is not a timing tool
It is lagged. A lot can happen between Tuesday and Friday, and again before you act on it.
It covers futures only. Spot FX is largely over-the-counter and is not in the report. Futures positioning is a sample, not the whole market. Retail brokers' spot positions and bank flows do not appear.
Extremes can persist. A crowded trade can get more crowded for months. "Extreme" has no fixed meaning, and calling a reversal from positioning alone is a classic error.
Futures contracts have rolls. Open interest shifts around quarterly expiries, which can distort week-to-week changes.
Net numbers hide hedges. A fund long euro futures might be hedging a cash exposure, not expressing a directional view. The report cannot tell the difference.
No intent. Positions say nothing about a participant's time horizon, stop level or reason.
Relative positioning
Raw contract counts are hard to compare across currencies because contract sizes and market depth differ. Many analysts express positioning as a percentile of its multi-year range or as a share of open interest. That gives context, though the result still depends on the window chosen, so be wary of any single "percentile" quoted without its lookback period.
How to read it in practice
- Pick one format (TFF is a sensible start) and stay with it.
- Look at the net position and the weekly change for the group you follow.
- Compare to a multi-year range for context.
- Ask whether price has already moved in the direction of the positioning. Positioning that follows a big move is a description of the past.
- Combine with other information, such as trend, real yields and the event calendar, never use it alone.
Common misuses
- Treating "speculators are short" as a buy signal. There is no such rule; positioning is information about risk, not direction.
- Ignoring the Tuesday date and assuming it describes Friday.
- Extrapolating from a single week.
- Applying futures positioning to exotic pairs that have no futures contract.
Where it fits in a process
The COT is best treated as sentiment context, one input beside price and macro drivers (technical vs fundamental). A trader who understands crowding may size down into an event where the market is heavily one-sided, since the reaction to a surprise can be sharper (position sizing). That is a risk-management use, and a far more defensible one than prediction.
The report is published free on the CFTC website. Read it as a lagging map of who is exposed, not a forecast. This article is general education, not advice.