How dhritij27/quant-signal-backtest works
This is a quant research pipeline for evaluating a value/quality investment signal, not a live trading system. The principal workflow selects synthetic or real market data, scores point-in-time fundamentals, constructs and evaluates a rebalanced portfolio, applies turnover costs, and reports gross/net performance and statistics. The graph separates the engine’s snapshot, portfolio, and return responsibilities. Source excerpts show the engine, signal, loaders, and CLI wiring; README statements support the broader documented capabilities. Real-data survivorship and filing-date limitations remain as documented.
Main entry points: run_backtest.py Backtest CLI · engine.py PIT snapshot · loader.py Real data loader · quality_value.py Value-quality composite
Components 15
Run orchestration
- Backtest CLI
scripts/run_backtest.py - Run configuration
config.py
Market data
- Synthetic generator
data/synthetic.py - Real data loader
data/loader.py
Signal research
- Value-quality composite
signals/quality_value.py - Signal transforms
signals/base.py
Portfolio evaluation
- PIT snapshot
backtest/engine.py - Portfolio construction
backtest/engine.py - Period returns
backtest/engine.py - Turnover costs
backtest/costs.py - Performance statistics
backtest/metrics.py - Return series and summaries
scripts/run_backtest.py
Other components
- Researcher
- Wikipedia constituents
- Yahoo Finance
Connections 15
- Researcher → Backtest CLI: runs mode
README.md - Backtest CLI → Run configuration: loads defaults
scripts/run_backtest.py - Backtest CLI → Synthetic generator: generates data
scripts/run_backtest.py - Backtest CLI → Real data loader: loads real data
scripts/run_backtest.py - Real data loader → Wikipedia constituents: fetches constituents
data/loader.py - Real data loader → Yahoo Finance: fetches prices and fundamentals
data/loader.py - Backtest CLI → Value-quality composite: creates signal
scripts/run_backtest.py - Backtest CLI → PIT snapshot: runs engine
scripts/run_backtest.py - PIT snapshot → Value-quality composite: scores snapshot
backtest/engine.py - Value-quality composite → Signal transforms: uses transforms
signals/quality_value.py - PIT snapshot → Portfolio construction: constructs weights
backtest/engine.py - PIT snapshot → Period returns: calculates returns
backtest/engine.py - PIT snapshot → Turnover costs: computes turnover and costs
backtest/engine.py - PIT snapshot → Performance statistics: summarizes performance
backtest/engine.py - Backtest CLI → Return series and summaries: prints and saves
scripts/run_backtest.py