What it does
Pulls the weekly CFTC Commitment of Traders (COT) report straight from publicreporting.cftc.gov and computes the overlay every commodity desk otherwise builds by hand: net position, 3-year (156-week) percentile rank, and week-over-week position velocity for managed money, swap dealers, commercials, and non-reportables. It covers all three CFTC report families — legacy, disaggregated, and financial (TFF) — in one consistent schema, with flexible commodity matching from exact contract names to fuzzy terms like GOLD or NATURAL GAS.
Who it's for
Commodity traders, macro funds, CTA researchers, and analysts who follow futures positioning — anyone who currently exports raw COT JSON into a spreadsheet to normalize positions and compute percentiles. Raw COT data is free; this actor packages the signal work.
Sample fields / output
Every record is typed JSON. Key fields: commodity_name contract_market_name contract_code report_date category long_positions short_positions spread_positions net_position open_interest_all percentile_3yr velocity_week_over_week_pct.
Example use cases
- Flag crowded trades: alert when managed-money net positioning in WTI or gold crosses the 90th percentile of its 3-year range.
- Feed a weekly macro dashboard with positioning velocity across rates, FX, energy, metals, and ags from a single dataset shape.
- Backtest contrarian signals that fade extreme COT percentiles against subsequent price action.
Try the CFTC COT Tracker on Apify->
Related actors
FAQ
How fresh is the data?
The CFTC publishes COT weekly (typically Friday, reflecting Tuesday positions). Run the actor after each release to get the latest week plus computed percentile and velocity.
What's the difference between the legacy, disaggregated, and financial reports?
Legacy splits traders into commercial/non-commercial for all futures; disaggregated breaks out managed money, swap dealers, and producers for physical commodities; financial (TFF) covers rates, FX, and equity index futures by asset manager, leveraged fund, and dealer categories. The actor normalizes all three into one schema.
How is the 3-year percentile computed?
For each contract and trader category, the current net position is ranked against the trailing 156 weekly observations, so a percentile of 87 means positioning is heavier than 87% of the past three years.
Top comments (0)