Problem / Opportunity
Currently, the trading agents lack institutional positioning data, which limits their ability to gauge long-term market sentiment and trend reversals in Forex pairs.
Proposed Feature
I would like to integrate the Commitment of Traders (COT) data—specifically focusing on the Commercials Report and a calculated COT Index—into the repository. This will allow the models to factor in where major banks and institutions are hedging their positions, significantly enhancing the structural reasoning and predictive power of the trading agents.
Proposed Implementation Plan
- Data Sourcing: Fetch historical and weekly COT data (likely via the CFTC API or a wrapper library like
chuback).
- Feature Engineering: Calculate a normalized COT Index (e.g., a 3-year or 26-week lookback percentile) to transform raw net-positioning numbers into actionable technical signals.
- Agent Integration: Expose these features to the trading environment/agents so they can use it as a high-timeframe macro filter.
Additional Context
I am happy to work on this feature myself! Please let me know if this aligns with the project's architecture or if you have any preferences on data pipelines.
Problem / Opportunity
Currently, the trading agents lack institutional positioning data, which limits their ability to gauge long-term market sentiment and trend reversals in Forex pairs.
Proposed Feature
I would like to integrate the Commitment of Traders (COT) data—specifically focusing on the Commercials Report and a calculated COT Index—into the repository. This will allow the models to factor in where major banks and institutions are hedging their positions, significantly enhancing the structural reasoning and predictive power of the trading agents.
Proposed Implementation Plan
chuback).Additional Context
I am happy to work on this feature myself! Please let me know if this aligns with the project's architecture or if you have any preferences on data pipelines.