M2Quant

A Claude Code Skill for Global Liquidity Quantitative Analysis.

Overview

M2Quant is a quantitative analysis system that tracks global liquidity indicators and generates trading signals for core assets. It analyzes Federal Reserve balance sheet data, Treasury General Account (TGA), and Overnight Reverse Repo (RRP) to quantify money liquidity, then compares against Gold, Nasdaq, and Bitcoin using a "scissors factor" methodology.

Features

• Real-time Liquidity Analysis: Fetches and processes Fed liquidity indicators from FRED
• Multi-Asset Coverage: Tracks Gold, Nasdaq, and Bitcoin against liquidity
• Scissors Factor: Proprietary metric comparing liquidity growth vs asset price growth
• Trading Signals: Generates Buy/Hold/Sell recommendations based on quadrant analysis
• Professional Reports: Clean, formatted output for analysis review
• Standalone Runner: Execute analysis without Claude Code CLI

What's New in v1.2.0

• npm Distribution: Install via npm install @m2quant/claude-skill
• skills.sh Integration: One-click installation at skills.sh
• Cross-Platform Support: Auto-installs to Claude Code, Cursor, and Codex agents

See CHANGELOG.md for full details.

Installation

Option 1: One-Click Install (Recommended)

Visit skills.sh and search for m2quant.

Option 2: npm Install

npm install @m2quant/claude-skill

The skill will be automatically installed to your AI agent's skill directory.

Option 3: Manual Setup

Prerequisites:

• Python 3.10+
• Claude Code CLI (optional, for skill usage)
• FRED API key (free from FRED)

Steps:

1. Clone the repository:

git clone https://github.com/Kampter/M2Quant.git
cd M2Quant

2. Create and activate virtual environment:

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies:

pip install -r requirements.txt

4. Configure API key:

cp config/api_keys.example.json config/api_keys.json
# Edit config/api_keys.json with your FRED API key

Usage

Via Claude Code Skills

# Full analysis with detailed report
claude /m2quant

# View liquidity conditions only
claude /m2quant liquidity

# View trading signals only
claude /m2quant signal

# Generate complete report
claude /m2quant report

Via Standalone Script

Run analysis directly without Claude Code:

# Full analysis (default)
python scripts/run_analysis.py

# Liquidity data only
python scripts/run_analysis.py liquidity

# Trading signals only
python scripts/run_analysis.py signal

Data Sources

Liquidity Indicators (FRED API)

Series	Description	Frequency
WALCL	Federal Reserve Total Assets	Weekly
WTREGEN	Treasury General Account	Weekly
RRPONTSYD	Overnight Reverse Repo	Daily
M2SL	M2 Money Supply	Monthly

Asset Prices (Yahoo Finance)

Symbol	Description
GC=F	Gold Futures
^IXIC	Nasdaq Composite
BTC-USD	Bitcoin

Core Methodology

Net Liquidity Formula

Net_Liquidity = WALCL - TGA - RRP

Scissors Factor

Scissors = Liquidity_YoY% - Asset_Price_YoY%

The scissors factor measures the divergence between liquidity growth and asset price appreciation, helping identify over/undervalued conditions.

Signal Generation

Signals are generated using a weighted composite:

• Regime Signal (50%): Based on liquidity regime and scissors factor
• Threshold Signal (30%): Based on deviation from liquidity-implied value
• Momentum Signal (20%): Based on liquidity vs price momentum divergence

Project Structure

M2Quant/
├── .claude/              # Claude Code configuration
├── skills/               # Claude Code Skills
│   ├── m2quant/          # Main orchestration skill
│   │   ├── SKILL.md      # Skill definition with frontmatter
│   │   ├── reference.md  # Methodology documentation
│   │   └── examples/     # Example outputs
│   ├── fetch-liquidity/  # FRED data fetching
│   ├── fetch-prices/     # Yahoo Finance fetching
│   ├── calculate-factors/
│   ├── generate-signals/
│   └── generate-report/
├── src/                  # Python modules
│   ├── __init__.py       # Version: 1.2.0
│   ├── data_fetcher.py
│   ├── factor_engine.py
│   ├── signal_generator.py
│   └── report_generator.py
├── scripts/              # Standalone scripts
│   └── run_analysis.py   # Direct execution runner
├── config/               # Configuration files
├── data/                 # Data cache (gitignored)
└── reports/              # Generated reports (gitignored)

Configuration

API Keys (config/api_keys.json)

{
  "fred_api_key": "YOUR_FRED_API_KEY"
}

Signal Thresholds (config/thresholds.json)

{
  "buy_threshold": -1.5,
  "sell_threshold": 1.5,
  "strong_buy_threshold": -2.0,
  "strong_sell_threshold": 2.0,
  "zscore_window": 60,
  "momentum_window": 20
}

Skill Context Costs

Command	Context Cost	Use When
/m2quant	High	Full analysis needed
/m2quant liquidity	Low	Quick liquidity check
/m2quant signal	Medium	Just need trading signals

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

MIT License - see LICENSE for details.

Disclaimer

This analysis is for informational purposes only. Not financial advice. Past performance does not guarantee future results.

Author

Created as a Claude Code Skill for quantitative analysis.