Einstein Research — Market Bubble Risk Detector

Evaluates market bubble risk through quantitative, data-driven analysis using a revised Minsky/Kindleberger framework. Prioritizes objective metrics over sub...

Security Scan

The skill's requirements, instructions, and included files are consistent with a market-bubble detector: it fetches public market/sentiment data, computes normalized indicators, and produces reports; nothing requested is disproportionate to that purpose.

Purpose & Capability

The name, description, and documentation all describe a data-driven bubble detector and the requested actions (fetch public market/sentiment data, compute scores, output JSON/Markdown) match that purpose. Minor inconsistency: documentation and examples refer to several entry points (bubble-detector CLI, scripts/bubble_detector.py, and the repository actually contains scripts/bubble_scorer.py). This looks like sloppy naming/version drift rather than a mismatch in required privileges or unrelated credentials.

Instruction Scope

SKILL.md and the implementation guide give a bounded, explicit workflow: collect specific public indicators (CBOE, VIX, FINRA, Google Trends, IPO stats), normalize, weight, and output a report. It instructs use of web_search/HTTP calls and public data APIs — appropriate for the stated task. There are no instructions to read unrelated user files, system credentials, or to exfiltrate private data. The skill requires network access to public sources (explicitly listed).

Install Mechanism

There is no install spec (instruction-only style) and included README lists normal Python dependencies (yfinance, requests, pandas). No downloads from untrusted URLs or archive extraction steps are present. The lack of an install step reduces risk, but the code is present and would run locally if invoked.

Credentials

The skill declares no required environment variables, no credentials, and no configuration paths. That is proportionate: the design uses public data sources and common Python libraries. The documentation explicitly states 'No API keys required.' If you plan to adapt the code to use paid APIs, additional credentials would be needed but are not requested here.

Persistence & Privilege

The skill is not marked always:true and uses normal autonomous invocation semantics. It does not request to modify other skills or system-wide settings. The skill includes a runnable script (local), which will run only when invoked — standard for this category.

Assessment

This package appears internally consistent and implements a public-data, quantitative bubble detector. Before you run it: 1) review the script file (scripts/bubble_scorer.py) to confirm there are no hidden network endpoints or unexpected behavior (the repository references multiple script/CLI names — bubble-detector, bubble_detector.py, bubble_scorer.py — so verify the actual runnable entrypoint). 2) Install dependencies in a sandbox or virtual environment (pip install yfinance requests pandas) and run the script with limited privileges. 3) Be aware the tool fetches public internet data (Google Trends, CBOE, FINRA, etc.) — confirm your environment permits that and that no private credentials are accidentally used. 4) If you plan to integrate with paid/private APIs, supply only the minimal credentials needed and audit any code changes. Overall: coherent and expected for its stated purpose, but do the routine code review and run in a controlled environment due to the presence of executable script files.

Market Bubble Risk Detector

Overview

This skill evaluates market bubble risk through a quantitative, data-driven analysis based on a revised Minsky/Kindleberger framework. It prioritizes objective metrics over subjective impressions to prevent confirmation bias and support practical investment decisions.

Core Principles:

• Data over Narrative: Relies on measurable data, not just "it feels frothy."
• Composite Score: Generates a score from 0-100 to quantify bubble risk.
• Multi-Factor Model: Incorporates sentiment, valuation, leverage, market structure, and new issuance data.
• Action-Oriented: Provides clear thresholds for tactical adjustments (e.g., raising cash, hedging).

---

When to Use This Skill

Explicit Triggers:

• "Are we in a stock market bubble?"
• "Analyze the risk of a market crash."
• "Is the market overvalued?"
• "Should I be taking profits?"
• User asks about "bubble risk," "market froth," "irrational exuberance," or "Minsky moment."

Implicit Triggers:

• User expresses anxiety about high valuations or a rapid market run-up.
• User is considering de-risking their portfolio.

---

Workflow

Step 1: Execute the Data Collection and Analysis Script

The bubble-detector CLI tool automates the entire process.

bubble-detector run

The script performs the following actions:

1. Fetches Data: Collects data for each of the 7 quantitative indicators.
   • Put/Call Ratio (CBOE)
   • VIX Index (CBOE)
   • Margin Debt (FINRA)
   • Market Breadth (% Stocks > 200d MA)
   • IPO Issuance (e.g., from a public data source)
   • Retail Volume as % of Total
   • Forward P/E Ratio vs. Historical Average
2. Normalizes Indicators: For each indicator, it calculates a percentile rank over the last 5 years. A rank of 100 means the indicator is at its most "bubbly" level in 5 years.
3. Calculates Composite Score: A weighted average of the normalized indicator scores.
   • Sentiment (Put/Call, VIX, Retail Volume): 40%
   • Leverage (Margin Debt): 20%
   • Market Structure (Breadth): 20%
   • Valuation & Issuance (P/E, IPOs): 20%
4. Generates Report: Outputs a JSON file and a Markdown summary.

Step 2: Analyze the Report

JSON Output (bubble_report_YYYY-MM-DD.json):

• Contains the raw data, normalized scores for each indicator, and the final composite score.

Markdown Report (bubble_report_YYYY-MM-DD.md):

• Overall Bubble Score: e.g., "78 / 100 (High Risk)"
• Indicator Dashboard: A table showing the current value and normalized score for each of the 7 indicators.
• Key Drivers: Highlights which indicators are contributing most to the high score.
• Historical Context: Compares the current score to levels seen before previous market corrections.
• Recommended Posture: Translates the score into a tactical recommendation.

Interpretation & Recommended Actions

The composite score maps to specific risk postures:

• 0-40 (Low Risk - "Accumulate"):

  • Characteristics: Fear is high, valuations are reasonable, leverage is low.
  • Action: A good time to be deploying capital and taking on risk.

• 41-60 (Moderate Risk - "Cautious Accumulation"):

  • Characteristics: Market is healthy but not cheap. Some signs of optimism are emerging.
  • Action: Continue to invest, but perhaps with a greater focus on quality.

• 61-80 (High Risk - "Hold & Hedge"):

  • Characteristics: Greed is prevalent, valuations are stretched, breadth may be narrowing.
  • Action: Hold existing positions, but stop new aggressive buying. Consider adding hedges (e.g., puts) or raising a small amount of cash.

• 81-100 (Very High Risk - "Distribute & Protect"):

  • Characteristics: Euphoria, extreme valuations, high leverage, widespread speculation.
  • Action: Systematically take profits from high-beta positions. Raise significant cash (e.g., 20-40%). Actively hedge the remaining portfolio. This is the time to be selling to the optimists.

Important Considerations

• Not a Timing Tool: This skill indicates when risk is high, not the exact top of the market. Bubbly conditions can persist for months.
• Context is Key: Always present the score in the context of the underlying indicators. A high score driven by stretched valuations is different from one driven by extreme sentiment.
• No Panicking: The goal is to make small, rational adjustments to risk exposure, not to sell everything in a panic.