A pure Python backtesting engine using NumPy to evaluate futures and stock trading strategies with core performance metrics.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install backtesting-framework
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install backtesting-framework using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Backtesting Framework?
The Backtesting Framework is a lightweight, dependency-free Python tool designed for algorithmic traders to validate quantitative trading strategies. Operating entirely with standard Python structures and NumPy, it eliminates the bloat of traditional heavy trading libraries while preserving performance.
By feeding the engine historic close-price arrays and signal matrices, developers can quickly simulate asset management strategies for both stocks and futures markets. The framework abstracts mathematical calculations for essential risk metrics, helping users refine their logic before deploying real capital.
Backtesting Framework Use Cases
- Strategy Validation: Test historical viability of standard indicators like Moving Average Crossovers, Relative Strength Index (RSI), and Bollinger Bands.
- Risk Management Profiling: Quantify worst-case drawdown scenarios and evaluate volatility-adjusted asset returns via the Sharpe Ratio.
- Custom Logic Testing: Prototype experimental quantitative signals by converting vector-based market observations into transactional holds, longs, or shorts.
How Backtesting Framework Works
- Data Ingestion: Market close prices for stock or futures instruments are loaded into a continuous one-dimensional NumPy array.
- Signal Generation: A chosen technical indicator function evaluates the price array to construct an aligned array mapping actions: long (1), hold (0), or short (-1).
- Return Vectorization: The backtesting engine shifts historical returns and multiplies them against active position signals to calculate strategy performance per interval.
- Metrics Formulation: The engine processes aggregated mathematical gains to compute overall returns, Sharpe Ratio, win rate, and peak-to-trough maximum drawdown.
- Report Compilation: Performance statistics are aggregated into a cleanly formatted technical analysis summary showing final simulated capital.
Backtesting Framework Setup
Prerequisites
Ensure you have python3 and NumPy installed on your local execution machine.
Installation
To use this capability within Openclaw Skills, set up your Python environment and verify your installation:
# Verify python3 installation
python3 --version
# Ensure numpy package is present
pip install numpy
Backtesting Framework Data Schema & Taxonomy
Input Requirements
The framework relies on synchronized vector dimensions for execution calculations:
| Data Object | Format | Values / Types | Description |
|---|---|---|---|
prices |
NumPy Array | Float / Int | Historical close prices over a continuous timeline. |
signals |
NumPy Array | 1 (Long), 0 (Hold), -1 (Short) |
Generated target exposure positions matching price steps. |
Output Schema
When a simulation run finishes, the engine outputs a dictionary mapping performance elements:
total_return: The absolute percentage growth/decline generated over the time series.sharpe_ratio: Volatility-adjusted performance metric relative to risk.win_rate: Percentage of periods that yielded positive financial returns.max_drawdown: Peak-to-trough decline value tracking historical capital risk.final_capital: Ending balance calculation based on the initial funding seed.
Backtesting Framework Advanced Features
- Zero External Dependencies: Built entirely with pure Python and vector-optimized NumPy, ensuring low resource footprints and rapid iteration speeds.
- Mathematical Vector Convolutions: Utilizes fast NumPy matrix operations (
np.convolve,np.where,np.cumprod) rather than slow iterative tracking loops. - Extensible Indicator Blueprint: Offers clear base templates for constructing highly complex proprietary algorithmic strategies beyond standard default models.
SKILL.md
Loading
Related Openclaw Skills
China Futures Quotes
An AI agent skill to monitor real-time prices, market volumes, and contract details across major Chinese futures exchanges.
China Options Pricing
An analytical toolkit for calculating European and American option pricing, Greeks, and implied volatility tailored to the Chinese derivatives market.
China Technical Analysis
Generate high-precision technical indicators, buy/sell signals, and chart trend descriptions for Chinese commodity and financial futures.
Google Maps Scraper (Apify)
An AI-agent skill utilizing Apify to scrape rich local business data, reviews, images, and website contacts directly from Google Maps.
China Stock Quotes
An AI coding agent skill to fetch and analyze real-time A-share stock prices, market indices, and sector data from Shanghai, Shenzhen, and Beijing exchanges.
Agent OS
Agent OS is a lightweight, file-based collaboration layer that enables seamless, auditable, and structured multi-agent coordination without relying on bloated chat histories.
Featured*
