bist_quant

Quantitative research and backtesting library for Borsa Istanbul (BIST). Version 0.3.3.

bist_quant is a Python library for factor research, signal construction, backtesting, portfolio analytics, and BIST-focused data workflows. It is designed to be usable from notebooks, scripts, dashboards, and external applications without requiring a repo checkout.

Install

pip install bist-quant

Quick Start

from bist_quant import DataLoader, DataPaths, PortfolioEngine

paths = DataPaths()
loader = DataLoader(data_paths=paths)
engine = PortfolioEngine(
    data_loader=loader,
    options={"use_regime_filter": False},
)

result = engine.run_factor("momentum")
print(f"Sharpe: {result['sharpe']:.2f}")
print(f"CAGR:   {result['cagr']:.1%}")

By default the library uses user-scoped directories:

• data: ~/.local/share/bist-quant/data
• regime outputs: ~/.local/share/bist-quant/regime/simple_regime
• cache: ~/.cache/bist-quant

Override them with BIST_DATA_DIR, BIST_REGIME_DIR, and BIST_CACHE_DIR when needed. If you are working from local parquet/CSV files, set BIST_DATA_SOURCE=local.

What Ships in the Library

• portfolio.py - PortfolioEngine and backtest entry points
• signals/ - factor and signal builders
• analytics/ - performance analytics and metrics
• clients/ - built-in data-provider integrations
• common/ - data loading, backtester, risk, and shared helpers
• configs/ - strategy registry and configuration loaders
• regime/ - regime classification helpers
• realtime/ - quote and market snapshot helpers

Data Setup

The package does not ship datasets. Point it at an existing dataset directory or seed data into the default user-scoped location.

export BIST_DATA_DIR="$HOME/.local/share/bist-quant/data"
python -m bist_quant.fetchers.fetch_gold_prices
python -m bist_quant.fetchers.fetch_indices
python -m bist_quant.clients.update_prices

Documentation

• Installation
• Quick Start
• Configuration
• Signals Guide
• Backtesting Guide
• Portfolio & Risk
• Multi-Asset Guide

Development

Contributor setup lives in CONTRIBUTING.md. App-stack deployment notes remain in deploy/README.md, but they are out of scope for the published library package.

License

Apache License 2.0 - see LICENSE.