skills.rest
tradermonty
tradermontyCOMMUNITY
Explore all skills by this author →
GitHub

news-reaction-failure-analyzer

Tests whether crowded markets failed to react to favorable news using drift-significance statistics.

2.8k|632|Updated Oct 19, 2025
Search Tags:#market sentiment#fmp api#contrarian trading#futures analysis#news reaction#cot report#drift significance test
One-click install
npx skills add https://github.com/tradermonty/claude-trading-skills --skill news-reaction-failure-analyzer
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: news-reaction-failure-analyzer
Source: https://github.com/tradermonty/claude-trading-skills/tree/main/skills/news-reaction-failure-analyzer
Command: npx skills add https://github.com/tradermonty/claude-trading-skills --skill news-reaction-failure-analyzer

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires requests, and includes scripts (resource) and references (resource) components.

What problem does it solve?

Crowded speculative positioning alone is not a contrarian signal — you need evidence the market has stopped rewarding the crowd's direction. This Skill automates step 2 of Jason Shapiro's COT contrarian process by statistically testing whether a market failed to react to news that should have favored the crowded side, avoiding naive failure-ratio rules that false-confirm on pure noise.

Core Features & Use Cases

  • Statistically validated verdicts: Produces a fail-closed CONFIRMED / NOT_CONFIRMED / INSUFFICIENT_EVIDENCE verdict using a drift-significance test with Monte-Carlo-verified false-positive bounds, plus event clustering to guard independence.
  • Robust price sourcing: Fetches price series from the FMP API with a documented per-symbol fallback chain (futures first, ETF proxy on 402 or empty responses) and explicit proxy-use caveats.
  • Fail-closed guardrails: Refuses to produce verdicts on stale or NEUTRAL detector reports, missing events, or markets with no viable price source — never crashes or fabricates data.
  • Use Case: After cot-contrarian-detector flags GBP futures (B6) as CROWDED_LONG, curate recent bullish news events via web search, then run the analyzer to check whether the market shrugged them off before considering a contrarian short setup.

Quick Start

Run a news-failure check on symbol B6 using my cot-contrarian-detector report and a curated events JSON to see if the crowded-long market failed to rally on bullish news.

Frequently Asked Questions about news-reaction-failure-analyzer

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I check if a crowded market failed to react to news?▼

Curate relevant news events into an events JSON with timestamps, source URLs, and BULLISH/BEARISH expected impact, then run analyze_news_reaction.py with the symbol, a cot-contrarian-detector report or explicit direction, and the events file. The script fetches prices and returns a CONFIRMED, NOT_CONFIRMED, or INSUFFICIENT_EVIDENCE verdict.

What is Jason Shapiro's COT contrarian news-failure test?▼

It is step 2 of a 5-step contrarian process: after identifying a crowded market via COT positioning, check whether it failed to rally on bullish news (crowded long) or sell off on bearish news (crowded short). This skill automates only step 2; price-action confirmation, entry, and exit remain manual.

Does the analyzer work without a cot-contrarian-detector report?▼

Yes, you can pass --direction CROWDED_LONG or CROWDED_SHORT explicitly instead of --detector-json. Without a detector report or explicit direction the skill refuses to run, and a NEUTRAL classification fails closed unless overridden.

Why does the analyzer return INSUFFICIENT_EVIDENCE for some markets?▼

INSUFFICIENT_EVIDENCE occurs when fewer than three usable event clusters exist, the detector report is missing or stale, or no price source works. Markets like VX, ZQ, HO, and agricultural futures have no viable FMP price feed on the tested key, so they always fail closed.

What API and Python dependencies does the news reaction analyzer need?▼

It requires Python 3.9+ with the requests library and an FMP API key set via FMP_API_KEY or --api-key. Prices come from FMP's stable/historical-price-eod/light endpoint, with ETF proxies as documented fallbacks.

Is a CONFIRMED verdict a trade signal?▼

No. CONFIRMED only completes step 2 of 5 in the Shapiro process, meaning the market significantly failed to reward the crowd's direction. Price-action confirmation, entry, and exit steps remain manual, and the output is for research and education, not investment advice.

Related Skills

Explore More Finance & Accounting Skills→
EternisAIEternisAI
1

cftc

Analyze CFTC COT data to compute net positions and open interest.

Official
Advanced
datagridSolutiondatagridSolution

Trailing Buy/Sell Bot

Execute trailing buy and sell orders after trigger and reversal callback confirmation.

Official
Intermediate
duolongworldduolongworld
59

crowding-state2x2

Diagnose stock crowding states via 2x2 quadrant analysis of inflow metrics.

Community
Intermediate
cubexchcubexch
13

gcr

Identify crowded crypto trades and generate contrarian theses from sentiment extremes.

Official
Advanced