Analysis of COT reports on forex: does this make sense?
The Commitments of Traders report is real data about real positions, published by a regulator. It is also three days stale, covers a fraction of the market, and is routinely read backwards.
Every Friday the Commodity Futures Trading Commission publishes a breakdown of open positions in US futures markets, including currency futures on the Chicago Mercantile Exchange. Traders use it as a proxy for positioning in spot forex.
The schedule, and the lag
The report is released Friday at 15:30 Eastern time. The data in it is a snapshot of positions as of the close on Tuesday.
So the freshest figure you ever see is three days old, and by the following Tuesday it is a week old. In a market that can reprice entirely on one central bank meeting, this is a serious constraint on any use that depends on timing.
The three categories
| Category | Who | Behaviour |
|---|---|---|
| Non-commercial | Funds, large speculators | Trend-following; positions build over weeks and reach extremes near turning points |
| Commercial | Corporates, banks hedging | Hedging, price-insensitive; typically the other side of speculators |
| Non-reportable | Small traders below reporting limits | Residual; traditionally read as the crowd |
The classic reading is that non-commercials are right in the middle of a trend and wrong at its extremes. When speculative net length reaches a multi-year high, everyone who intends to buy has bought, and the marginal buyer no longer exists.
Why forex is the weakest application
This is the honest objection and it deserves to be stated plainly.
- Coverage. Spot forex turns over trillions per day, essentially all of it over the counter. CME currency futures are a rounding error beside it. You are inferring the ocean from a bucket.
- No USD contract in the usual sense. Positioning is expressed against the dollar through each individual currency future, so a dollar view has to be assembled from six or seven separate series.
- Hedging distorts the categories. A bank's futures position may offset an OTC book you cannot see, so a large commercial short is not necessarily a bearish opinion.
In commodities the report is far stronger evidence, because futures are where the market actually trades. The further an instrument's real liquidity sits from the exchange, the less the report tells you.
Reading it usefully anyway
- Work with the net position of non-commercials, not gross long or short.
- Normalise it as a percentile of its own history over three years or so. The absolute contract count is meaningless without that context.
- Treat extremes as a condition, not a signal — an argument against adding to a crowded position rather than an argument for the opposite one.
- Watch the rate of change. Positioning unwinding quickly from an extreme is more informative than the extreme itself.
The data is free at cftc.gov under Market Reports → Commitments of Traders. The Legacy report is the familiar three-category format; the Traders in Financial Futures report splits participants more finely and is the better choice for currencies.
So does it make sense?
As a slow contextual input, yes. As a timing tool for a forex position, no — the lag alone rules it out, and the coverage problem compounds it.
It answers the question "is this trend crowded?" and answers it with real data. It does not answer "should I buy now", and no amount of processing will make it answer that.