What it does
Pyfolio Reloaded is a Python library for portfolio performance and risk analysis. It produces tear sheets that combine charts with return, volatility, drawdown and other statistics. The free Apache-licensed package analyzes existing results and can be used without Zipline.
Pyfolio turns strategy results into a tear sheet: a set of plots and summary tables that show both overall performance and how it changes through time. It is useful after you have a return series to examine.
The introductory example converts closing prices to percentage returns, uses a date index with UTC timezone information, and calls create_returns_tear_sheet. Its live_start_date argument marks the boundary between the earlier sample and the later evaluation period. The resulting tables separate those periods and identify drawdown peaks, troughs and recoveries. Set that date to the actual boundary in your research.
Install the local package with pip install pyfolio-reloaded, then open an example notebook in Jupyter. No Pyfolio account or subscription is needed. Zipline is optional, but if both packages share an environment, Zipline's pandas constraints still apply. The maintainers also distinguish current pip releases from the older releases kept on the Conda channel.
Reloaded receives bug, security and compatibility fixes. New feature development is directed toward ML4T Diagnostic, making Pyfolio especially useful for existing Python research workflows and reproducible analyses.
Strengths
- Combines portfolio summary statistics with time-series plots
- Separates earlier and later evaluation periods in reports
- Works independently of the Zipline backtesting engine
Watch out for
- The project focuses on maintenance and compatibility fixes
- Older Conda artifacts and current pip releases have different requirements
Best for
- Researchers analyzing an existing portfolio return series
- Python users adding drawdown and risk reports to a backtest
Sources
- Pyfolio Reloaded package installation and maintenance (Oct 27, 2026)
- Pyfolio report scope and environment compatibility (Oct 27, 2026)
- Pyfolio single-stock returns tear-sheet example (Oct 27, 2026)
- Pyfolio Reloaded Apache 2.0 license (Oct 27, 2026)