Skip to content

Quant101

A structured quantitative research laboratory for US equity alpha.

Built with Polars, backtested with a custom engine, and documented as a 59-page learning journal in LaTeX.

What This Project Is
Data Polygon.io flat files → Parquet, split-adjusted, 50-stock US large-cap universe
Alpha Factor signals → IC/IR evaluation → preprocessing → combination
Risk VaR/CVaR, drawdown, position sizing (EW, inverse-vol, vol-target, signal-weighted)
Execution Transaction cost models (fixed, spread, sqrt-impact), breakeven analysis
Validation Walk-forward, bootstrap Sharpe CI, PSR/DSR, multiple-testing corrections
Pipeline AlphaConfig → factor registry → sizing → rebalancing → backtest, all in one call

201 tests · 64 source files · 12.8K lines · 7 research notebooks

  • :material-rocket-launch: Quick Start

    Get up and running in 5 minutes

  • :material-file-tree: Architecture

    How the modules fit together

  • :material-flask: Research Guide

    End-to-end factor research workflow

  • :material-lightbulb: Lessons Learned

    Key insights from building this lab

Current Status

v1.0.0 — All research infrastructure complete (Phases 0–7). The lab is built; now it needs experiments that produce real alpha signal.

See the Changelog for version history, or the LaTeX encyclopedia for the full learning journal.