Customization Principle
By modifying the belief_list.json configuration file, the user can define the intelligence's:
- Underlying preference (market capitalization/industry)
- Risk threshold (maximum retracement/volatility)
- Position cycle (intraday/weekly)
Specific methods of implementation
- Conservative Configuration Example::
"聚焦市盈率＜15、近一月波动率＜20%的标的，仅交易业绩预告超预期+高管增持的复合事件" - Example of an aggressive configuration::
"主攻市值＜50亿、涨停板突破年线的题材股，配合融资余额攀升因子" - mixed strategy: Simultaneous deployment of 3-5 intelligences of different risk levels to achieve portfolio balancing
Tuning Tips
It is recommended to start withpython -m cli.main backtestBacktesting of different belief portfolios is then finely controlled by adjusting the descriptive statements in the JSON (e.g., adding constraints such as "exclude ST stocks").
This answer comes from the articleContestTrade: an AI multi-intelligence trading framework for event-driven investingThe
May not be reproduced without permission:AI productivity tools " How can we optimize the investment style of research intelligences to match individual risk preferences?
English 
简体中文 
日本語 
Deutsch 
Português do Brasil