OneQuant 4.0 A-Share Quantitative Backtesting & Paper Trading System

What Problem Does It Solve?

For individual A-share investors and quantitative strategy learners, building a complete local backtesting and paper trading environment is often complex. It requires separate handling of data acquisition, strategy coding, a backtesting engine, trade simulation, and risk control. OneQuant 4.0 addresses this by providing a locally deployed, integrated environment that consolidates these functions into a single web interface, while also exposing a full set of REST APIs for programmatic use and secondary development.

How Does It Work?

OneQuant 4.0 is a single-page application built on Flask. Its core workflow is as follows:

1. Data Layer: It integrates data sources like AkShare (free), JQData, and RQData (paid) to fetch A-share historical quotes, real-time snapshots (with ~3-5 second delay), financial data, and sector information. All data is fetched via actual API calls—fabricating data is strictly prohibited.
2. Strategy & Backtesting Engine: It includes templates for common strategies like MA Golden Cross/Death Cross and RSI Overbought/Oversold. Users can run backtests on specified stocks and timeframes via the web UI or API endpoints such as /api/backtest/ma, generating equity curves and performance metrics. Backtest results exclude transaction costs (commissions, slippage, stamp duty), which must be assessed separately.
3. Paper Trading & Risk Control: The system provides a local paper trading account for limit/market orders, position management, and trade history. It integrates risk control modules to configure stop-loss/take-profit levels and monitor maximum single-stock position ratios, with risk overviews queryable via APIs like /api/risk/overview.
4. Invocation & Integration: It offers 5 invocation methods, with the REST API being core. For example, you can use curl or Python requests to call /api/market/kline for candlestick data or /api/selection/list for conditional stock screening (e.g., limiting up stocks). All API responses are in JSON format, and the success field must be checked upon invocation.

Applicable Boundaries and Caveats

This tool is not a universal solution; understanding its boundaries is crucial:

• Applicable Scenarios: It is ideal for strategy logic validation, technical indicator learning, paper trading practice, and technical analysis of individual A-shares. All trade records and strategy parameters are stored in local JSON files, ensuring data privacy.
• Non-Applicable Scenarios: Due to its REST API polling (non-event-driven), it is unsuitable for high-frequency trading or scenarios requiring millisecond-level latency. The default mock_broker is a simulation environment and must never be used to operate real brokerage accounts. Data source coverage is primarily A-shares, with limited support for futures, options, or in-depth analysis of Hong Kong/US stocks.
• Key Limitations: The free data source AkShare's minute-level K-lines are aggregated from 5-minute data, and real-time quotes have a multi-second delay. Backtesting performance depends on Python computation; restarting the service is recommended for long-running sessions (e.g., >7 days). The web server uses Flask's development server, suitable for single-user local use; multi-user concurrency requires deployment to a production environment.

Important Disclaimer: Backtest results are for reference only—past performance does not guarantee future returns. All system functions run locally, and no profitability of any strategy is assured.