cot-contrarian-detector

---
name: cot-contrarian-detector
description: Use when the user asks about COT report analysis, crowded futures positioning, who is "trapped" in a market, or a Jason Shapiro-style speculative-positioning screen across CFTC futures.
category: Market & Macro
version: 1.0.0
tools: []
---

# COT Contrarian Detector

Detect futures markets where large speculators are crowded onto one side, using the first step of Jason Shapiro's COT (Commitment of Traders) contrarian process. Crowded positioning is a PRECONDITION for a contrarian trade, not a signal by itself — a market only becomes tradable once crowding is confirmed by a news failure and a price-action reversal (steps 2-3 below).

**Core thesis (Shapiro):** Large speculators (hedge funds, CTAs, momentum traders) tend to be maximally positioned at trend exhaustion, not at trend inception. Once they are already crowded onto one side, the next big move is statistically more likely to run them over than reward them further. Fade the speculators, not the commercials — commercials hedge for structural reasons and are not a crowd-psychology signal.

## When to Use
- "What markets are speculators crowded into right now?"
- "Run a COT report analysis" / "Show me COT positioning extremes"
- "Is anyone trapped in gold / the dollar / bonds?"
- User wants a Jason Shapiro-style contrarian futures screen

Not for: a trade signal on its own (crowding alone is not actionable — see Guardrails) or individual equities (COT reports cover CFTC futures markets only — indices, rates, FX, metals, energy, agri, crypto — not single stocks).

## Phase 1: Screen for Crowding
Use web_search / knowledge_base_search to pull each market's weekly legacy COT report (large-speculator long/short positions) for a curated core futures universe (roughly 23 liquid, representative markets), the full ~65-market universe, or specific symbols the user names.

For each market, compute a COT Index over a 156-week (3-year) lookback and a 26-week lookback:
COT Index = (current net position − minimum net position over the lookback) / (maximum net position over the lookback − minimum net position over the lookback) × 100

Classify:
- CROWDED_LONG: COT Index ≥ 90 (near the 3-year net-long high)
- CROWDED_SHORT: COT Index ≤ 10 (near the 3-year net-short high)
- NEUTRAL: everything in between

Never silently drop a market with insufficient history — flag it explicitly (e.g. "insufficient history: 40/156 weeks") instead of guessing.

## Phase 2: Present the Crowding Report
Highlight: which markets are CROWDED_LONG / CROWDED_SHORT and by how much; the 26-week index for context (is the crowding fresh or aging?); week-over-week net-position swings (fast-moving crowds are more fragile); and a clear note that crowding alone is not a trade signal.

## Phase 3: Guide the Remaining Steps Manually
For any crowded market the user wants to pursue, walk through the rest of Shapiro's 5-step process — these stay manual, not automated:

1. Crowding detection — done in Phase 1.
2. News failure — use web_search to check whether recent news favorable to the crowd's direction failed to move price the way the crowd would expect (e.g., a crowded-long market that doesn't rally on bullish news). This is the core edge and the most important confirmation.
3. Price-action confirmation — check the weekly chart for a reversal pattern or a failure at a new high/low.
4. Entry — against the crowd, with a stop at the recent swing extreme and small, fixed-risk position sizing.
5. Exit — when positioning normalizes back toward neutral (COT Index near 50) or the stop is hit.

Never recommend an entry from crowding alone — steps 2 and 3 must both confirm first.

## Cadence
CFTC publishes the COT report Fridays around 3:30pm ET, with positions as of the prior Tuesday — the data is always 3+ days old when published, and up to 9 days old by the following Friday. Run this weekly after Friday's publication, or ad hoc when the user asks about a specific market (the underlying data is still from the most recent Friday release).

## Guardrails
- Crowdedness alone is NOT a trade signal — it is a precondition. Never suggest an entry without news failure and price-action confirmation.
- Data is lagged 3-9 days — never treat it as real-time.
- Fade speculators (non-commercial / large-speculator positioning), not commercials — commercial hedging flows are structurally different.
- All output is for research/educational purposes, not investment advice.