cobweb-py

Cobweb-py is a Python-based backtesting API designed for evaluating trading strategies with a focus on realistic execution modeling. It targets users who need to simulate and analyze trading performance, offering features that address execution details such as slippage, spread, market impact using a square root model, participation caps, and multi-bar order slicing. The tool supports three execution horizon presets: intraday, swing, and long-term, allowing for flexible scenario testing.

The platform provides access to 71 technical features, including returns, momentum, MACD, RSI, Bollinger Bands, ATR, stochastic indicators, volume, risk metrics, beta, and regime detection. These features are computed server-side and are lazily evaluated. Cobweb-py includes a full backtest engine that handles signal-to-trade conversion, cash and margin constraints, various rebalance modes, and generates reinforcement learning-ready reward series with turnover and drawdown penalties.

For visualization and analysis, cobweb-py offers 27 chart types such as equity curves, drawdown plots, rolling Sharpe and Sortino ratios, regime performance, trade analysis, volume diagnostics, and correlation heatmaps. Charting is facilitated through Plotly helpers, and the platform can save plots as HTML files. The tool also provides a built-in Yahoo Finance data helper for data acquisition and supports input in CSV, DataFrame, or JSON formats.

Cobweb-py is accessible via a Python SDK, which can be installed with pip, and through a REST API with three endpoints (/features, /backtest, /plots) that accept and return JSON, making it usable from any programming language. The service requires no infrastructure management and can be used for free without an API key. This combination of technical analysis, backtesting, and execution realism positions cobweb-py as a comprehensive solution for quantitative developers and researchers working on trading strategy evaluation.