Valuation Analysis
Usage
To generate a valuation model for a company, execute the Python script in this directory.
Command:
python skills/valuation/run_valuation.py --ticker <TICKER> --horizon "<TIME_HORIZON>"
Parameters:
ticker: The stock symbol (e.g., AAPL).horizon: Time period (e.g., "1 year", "18 months"). Default: "1 year".
Output
Returns a JSON object with:
valuation_range: Low, Base, and High price targets with methodology.assumptions: Key assumptions driving the model (e.g., revenue growth rates, discount rates) with evidence citations.sensitivity: A sensitivity matrix showing how price targets change under different scenarios.
Example
python skills/valuation/run_valuation.py --ticker AAPL --horizon "1 year"
Environment Requirements
- Python 3.12+
- Access to
research.dbSQLite database - Neo4j GraphRAG instance running (default: bolt://localhost:7687)
- Environment variables: NEO4J_URI, NEO4J_USER, NEO4J_PASSWORD (optional, defaults provided)