---
name: cftc-commitments-of-traders-ingestion
description: "How to get CFTC Commitments of Traders (COT) data into a pandas DataFrame and Parquet: access, the request, the format, limits, quirks and terms."
---

# Load CFTC Commitments of Traders into pandas and Parquet

Source entry: https://signal-lab-jet.vercel.app/datasets/cftc-commitments-of-traders
Provider: https://www.cftc.gov/MarketReports/CommitmentsofTraders/index.htm
Checked against the provider's documentation: 2026-10-29

## Access
- No account, key or app token. The CFTC asks you not to overuse the API.

## Request
```bash
curl 'https://publicreporting.cftc.gov/resource/6dca-aqww.json?$limit=5&$order=report_date_as_yyyy_mm_dd%20DESC'
```

## Format
JSON array of rows (Socrata). Dataset 6dca-aqww is the Legacy futures-only report; other report types use other dataset ids. Numbers arrive as strings.

## Load into pandas and Parquet
Needs Python 3 with requests, pandas and pyarrow.
```python
"""CFTC Commitments of Traders, Legacy futures-only (Socrata API) -> pandas -> Parquet. No key or token."""
import requests
import pandas as pd

URL = "https://publicreporting.cftc.gov/resource/6dca-aqww.json"  # Legacy - Futures Only
PAGE = 5000
rows, offset = [], 0
while True:
    r = requests.get(URL, params={
        "$where": "report_date_as_yyyy_mm_dd >= '2024-01-01'",
        "$order": "report_date_as_yyyy_mm_dd, id",   # a stable order makes $offset paging safe
        "$limit": PAGE, "$offset": offset,
    }, timeout=60)
    r.raise_for_status()
    batch = r.json()
    rows += batch
    if len(batch) < PAGE:
        break
    offset += PAGE

df = pd.DataFrame(rows)
df["report_date_as_yyyy_mm_dd"] = pd.to_datetime(df["report_date_as_yyyy_mm_dd"])  # the Tuesday "as of" date
num = [c for c in df.columns if c.endswith("_all") or c.endswith("_old") or c.endswith("_other")]
df[num] = df[num].apply(pd.to_numeric, errors="coerce")  # Socrata returns numbers as strings
df.to_parquet("cftc_cot_legacy_futures.parquet", index=False)
print(len(df), "rows;", df["market_and_exchange_names"].nunique(), "markets; latest", df["report_date_as_yyyy_mm_dd"].max().date())
```
Run by us on 2026-10-29 (python 3.13.5 | pandas 3.1.0 | pyarrow 26.0.0 | requests 2.34.2). Result: 49,251 rows for 512 markets from 2024-01-02, latest report date 2026-10-20.

## Limits
- Rate limit: Not stated; the CFTC says you can use the API without a token as long as you are not overusing it.

## Quirks
- report_date_as_yyyy_mm_dd is the Tuesday the positions are as of. The report is normally released the following Friday at 3:30 p.m. ET, and federal holidays can delay it, so date each row to its release day in backtests.
- Page with $limit and $offset over a stable $order (the snippet orders by report date and id).
- Only markets where 20 or more traders hold reportable positions are covered.

## Terms and attribution
- Public domain US government information; the CFTC requests appropriate acknowledgement in any use (no set wording).
- Terms page: https://www.cftc.gov/WebPolicy/index.htm

## Sources
- https://www.cftc.gov/MarketReports/CommitmentsofTraders/index.htm: "As long as you are not overusing the API, you should be able to use the API without a token."
- https://www.cftc.gov/MarketReports/CommitmentsofTraders/ReleaseSchedule/index.htm: "The Commitments of Traders reports are released at 3:30 p.m. Eastern time."
- https://www.cftc.gov/WebPolicy/index.htm: "Public domain information may be freely distributed and copied, but it is requested that in any subsequent use the CFTC be given appropriate acknowledgement."

## Help improve this skill
Missing a source, or see something to improve? Call the MCP tool suggest_source or feedback (or POST /api/suggest or /api/feedback). Reviewed ideas: https://signal-lab-jet.vercel.app/agent-suggestions

Information for educational purposes only. Free tiers, limits and licenses belong to the providers and may change; check the provider's own terms before using the data.