Changelog

v0.1.0 — 2026-03-11

Initial release of PortfolioBench, a multi-asset portfolio benchmarking framework built on top of freqtrade.

Multi-asset support — backtest across cryptocurrencies, US equities (119 instruments), global indices, and Polymarket prediction markets in a single framework
16 built-in strategies — 8 trading strategies and 8 portfolio allocation algorithms ready to use out of the box
5 alpha factors — pluggable alpha-factor interface with EMA, RSI, MACD, Bollinger Bands, and Polymarket implementations
Full benchmarking suite — run all strategies across multiple asset classes and timeframes with one command, with parallel execution support

Custom exchange (Portfoliobench) — clean exchange subclass extending Binance with offline-tolerant market loading, synthetic market injection for stocks/indices, and USD/USDT normalization
Trading strategies: EmaCross, MacdAdx, IchimokuCloud, RsiBollinger, StochasticCci, MlpSpeculative, PolymarketMeanReversion, PolymarketMomentum
Portfolio strategies: ONS (Online Newton Step), InverseVol, MinVar, BestSingleAsset, ExpGradient, MaxSharpe, RiskParity, PolymarketPortfolio
Alpha factors: EmaAlpha, RsiAlpha, MacdAlpha, BollingerAlpha, PolymarketAlpha with IAlpha interface
portbench CLI — unified command-line interface with backtesting, portfolio, benchmark, and generate-data subcommands
Benchmark report — HTML report generation with detailed metrics per strategy
Parallel backtest execution — --workers flag for concurrent benchmark runs
Google Drive data integration — automatic download of OHLCV data (119 instruments × 3 timeframes)
GitHub Actions CI — unit tests, strategy import validation, benchmark report generation, and GitHub Pages deployment
Polymarket integration — backtest strategies on binary outcome prediction market contracts
Standalone portfolio pipeline — run portfolio construction outside of freqtrade's backtest loop
Synthetic data generation — portbench generate-data for testing without real market data