sjain7117/quant-backtest-engine

How sjain7117/quant-backtest-engine works

This repository is a quantitative strategy research project built around an event-driven backtesting engine. Its main workflow loads adjusted prices, lets pairs or momentum strategies produce signals using cursor-limited history, executes orders on the next bar with modeled costs, accounts for fills and equity, then reports performance. Separate research branches screen cointegration and sizing, and explore market regimes with feature construction, detectors, and a strategy overlay. Engine wiring and regime components not directly excerpted are represented only where the README or source index supports their relationships.

Main entry points: backtest.py Event Loop · events.py Event Messages · loader.py Price Loader · features.py Regime Features

Components 22

Market Data

Price Loader data/loader.py

Backtest Engine

Pair Data Handler engine/data_handler.py
Multi-Asset Handler engine/data_handler.py
Event Messages engine/events.py
Event Loop engine/backtest.py
Strategy Interface engine/strategy.py
Simulated Execution engine/execution.py
Portfolio Accounting engine/portfolio.py

Trading Strategies

Pairs Strategy strategies/pairs_trading.py
Momentum Strategy strategies/momentum.py
Baseline Strategies strategies

Quant Research

Pair Screening analysis/cointegration.py
Kelly Sizing analysis/kelly.py
Performance Metrics analysis/performance.py

Regime Analysis

Regime Features regime/features.py
HMM Detector regime/detector_hmm.py
Online Detector regime/detector_online.py
Rule Detector regime/detector_rule.py
Strategy Sleeves regime/sleeves.py
Regime Overlay regime/overlay.py

Other components

Researcher
Market Data Source

Connections 32

Researcher → Event Loop: runs README.md
Market Data Source → Price Loader: provides prices data/loader.py
Price Loader → Pair Data Handler: loads pairs engine/data_handler.py
Price Loader → Multi-Asset Handler: loads universe engine/data_handler.py
Pair Data Handler → Event Messages: emits bars engine/data_handler.py
Multi-Asset Handler → Event Messages: emits bars engine/data_handler.py
Event Loop → Pair Data Handler: advances bars
Event Loop → Event Messages: dispatches engine/backtest.py
Event Loop → Simulated Execution: executes pending engine/backtest.py
Event Loop → Strategy Interface: requests signals
Event Loop → Portfolio Accounting: routes events engine/backtest.py
Pairs Strategy → Strategy Interface: implements strategies/pairs_trading.py
Momentum Strategy → Strategy Interface: implements strategies/momentum.py
Baseline Strategies → Strategy Interface: implements strategies/buy_and_hold.py
Pairs Strategy → Event Messages: emits signals strategies/pairs_trading.py
Momentum Strategy → Event Messages: emits signals strategies/momentum.py
Portfolio Accounting → Event Messages: creates orders engine/portfolio.py
Simulated Execution → Event Messages: creates fills engine/execution.py
Event Loop → Performance Metrics: reports curve
Researcher → Pair Screening: screens pairs README.md
Pair Screening → Price Loader: analyzes prices
Researcher → Kelly Sizing: sizes positions README.md
Researcher → Regime Features: studies regimes README.md
Regime Features → Price Loader: loads market data regime/features.py
HMM Detector → Regime Features: builds inputs regime/detector_hmm.py
Online Detector → Regime Features: uses features regime/detector_online.py
Rule Detector → Regime Features: uses features regime/detector_rule.py
Strategy Sleeves → Pairs Strategy: runs strategy regime/sleeves.py
Strategy Sleeves → Momentum Strategy: runs strategy regime/sleeves.py
Regime Overlay → Online Detector: uses detector scripts/regime_overlay.py
Regime Overlay → Strategy Sleeves: combines sleeves scripts/regime_overlay.py
Regime Overlay → Performance Metrics: evaluates results scripts/regime_overlay.py

Drawn by GitDiagram on November 10, 2026 from the repository's file tree, README and source files. Read as Markdown ·