Analyzing Positions

Analyzing Positions

from cftc_cot import COTClient, COTAnalysis

client = COTClient()
df = client.legacy().market("Crude Oil").last_n_weeks(156).execute()

analysis = COTAnalysis(df, classification="legacy")

# 1. Net Positions (long - short) per category
df_net = analysis.net_positions()

# 2. Z-Scores (rolling, standardized)
df_z = analysis.z_scores(window=52)

# 3. COT Index (classic 0-100 normalized positioning)
df_index = analysis.cot_index(window=156)

# 4. Extremes (bullish/bearish flags off the COT Index)
df_extremes = analysis.extremes(threshold=0.9)

# 5. Long/Short ratios
df_ratios = analysis.long_short_ratios()

# 6. Week-over-week change in net positions
df_wow = analysis.wow_change()

# 7. Percentile rank of the latest non-commercial net position
pct = analysis.percentile_rank("noncomm_net")
print(f"Current non-commercial net is in the {pct:.0%} percentile of its history")

Each enrichment method returns the same internal DataFrame, so you can build up the columns you need. See COTAnalysis for the full method reference and the columns each one adds.