A Chinese quant framework that takes technical analysis very seriously

What it does

abu is a Python-based quantitative trading and investment framework covering Chinese A-shares, US equities, Hong Kong stocks, plus futures, options, and Bitcoin. It bundles backtesting, signal generation, and what appears to be a non-programming UI layer (abupy_ui). The repo also serves as companion code for two Chinese books on quantitative trading and machine learning.

The interesting bit

The project’s public site (abuquant.com) publishes live “AI research reports” on major indices—from the CSI 300 to the Nikkei 225—suggesting the authors run production signals, not just toy examples. The README is essentially a massive catalog of technical analysis strategies with emoji taxonomy: wave theory, harmonic patterns, candlestick combinations, and something called “chan theory” (缠论), a Chinese technical analysis method rarely seen in Western quant libraries.

Key highlights

• Covers multi-asset classes: stocks, options, futures, Bitcoin
• Includes non-programming UI for non-coders
• Companion code for published books on quant trading and ML
• Live signal examples for 20+ global indices and ETFs
• Heavy focus on classical/chartist technical analysis rather than factor models or statistical arbitrage

Caveats

• README is almost entirely in Chinese; English documentation appears minimal or absent
• The live report links demonstrate outputs but don’t explain methodology or performance metrics
• Actual code structure and API design are unclear from the README alone

Verdict

Worth a look if you’re a Mandarin-speaking quant interested in technical analysis-heavy strategies or need a starting point for Chinese market data. Western systematic traders or those seeking rigorous, documented alpha models will likely bounce off the language barrier and the chart-pattern mysticism.