7.6k stars · Apache-2.0 · v0.0.81 (2026-01-13)
Portfolio tearsheets from a returns series — the monthly-returns heatmap that tests the 2%-a-month claim, under a licence he can ship.
▶Repo detailsthe review · specs · pros & cons · install
What it is
A portfolio analytics library. Hand it a series of returns and it produces the full professional statistics set — Sharpe and Sortino ratios, maximum drawdown, win rate, rolling volatility, a monthly-returns heatmap — and writes the lot into a single self-contained HTML report.
Why it matters
Your target is stated as a monthly number, and quantstats's headline output is a monthly-returns table: the exact grid that shows whether two percent a month is a real pattern or one lucky quarter dragging eleven mediocre ones behind it. But the sharper reason it belongs here is the licence. Edition 2 flagged that backtesting.py is AGPL, and its own correction flagged that vectorbt carries a Commons Clause — which together mean neither can safely sit inside a paid ScalpingMate feature. QuantStats is Apache-2.0. It is the reporting layer you can actually put in front of customers. Build the tearsheet your users will see on this, and keep the AGPL backtester strictly for your own private research.
- Apache-2.0 — unlike both of your backtesters, it is safe inside a commercial product
- The monthly-returns heatmap is your stated goal rendered as a picture you can look at
- One call produces a complete HTML report you can hand straight to a user or serve in a browser
- It reports; it does not backtest. Something else still has to produce the returns series
- Version 0.0.81 after years of releases — a low-ceremony project, and some plotting dependencies are fussy about versions
- Beautiful statistics on an overfitted strategy are still an overfitted strategy. It will flatter you convincingly
source venv/bin/activate pip install quantstats