OneQuant A-Share Buy Sell Signal Backtesting Workbench

Problem Addressed

A-share entry and exit decisions often stall between scattered market data, ad-hoc indicator formulas, and backtest loops that mix paper trading with live account assumptions. OneQuant turns this into a local Flask service for validating rule-based MA and RSI signals without assembling a custom pipeline.

How It Works

• Data and signals: market indices, industry rankings, northbound flow, stock K-lines, five-level order book, and MACD/KDJ/RSI indicators with buy/sell signals.
• Strategy backtesting: templates for MA golden cross and RSI overbought/oversold, with adjustable periods, thresholds, initial capital, and date ranges, plus equity curves and performance records.
• Programmatic access: 14 modules and 102 REST endpoints for Python requests, CLI, curl, or batch scripts, with export of trade logs and backtest reports.
• Paper trading and risk controls: local simulated orders, positions, P&L, order management, stop-loss/take-profit, and max position limits.

Boundaries and Caveats

It is for local A-share research, not low-latency trading or direct broker execution. Free AkShare has roughly 3-5 second latency and limited intraday history; backtests omit commissions, slippage, and stamp duty, so signals are best confirmed after close. Paid JQData/RQData improve latency and historical depth, but outputs should still be treated as research references.