-
Notifications
You must be signed in to change notification settings - Fork 163
Home
youngcan edited this page Apr 3, 2026
·
22 revisions
欢迎来到 Wyckoff-Analysis 官方知识库。
本 Wiki 聚焦三件事:
- 策略逻辑:Wyckoff Funnel + 宏观水温 + OMS 风控
- 工程实现:GitHub Actions 调度、LLM 与 RAG 集成、数据快照回放
- 回测方法:当前可复现口径、已知偏差、参数对比方法
- 01. 威科夫理论:探寻主力的足迹
- 02. 量化选股因子:截面强弱与防守带
- 09. 板块轮动与大盘水温:A 股适配策略 (v2.1 新增)
-
完整术语速查 → 主仓库
GLOSSARY.md
-
hold_days=[5,6,7,8](Grid 搜索) exit_mode=sltp-
stop_loss=[-6,-7,-8,-9](Grid 搜索) -
take_profit=0(不设止盈) -
regime_filter=True(大盘水温仓位控制) - summary 强制输出:
最大回撤 / VaR95 / CVaR95 / 最长连亏
- 先选环境,再选个股:优先排除系统性风险,再讨论个股机会。
- AI 只做参谋,OMS 执行风控:模型建议不直接下单,执行权在风控引擎。
- 快照隔离计算与网络取数:支持离线重放与参数回测,提升可复现性。
- 新同学:按
03 -> 04 -> 05阅读,先看系统全貌,再看 AI,再看回测口径。 - 做策略迭代:先看
07再做回测,避免跨版本横向比较失真。
- Home
- 01_Product_Overview
- 02_Finance_Wyckoff_Method
- 03_Finance_Quantitative_Metrics
- 04_Finance_Sector_Rotation_Regime
- 05_Finance_Risk_Management
- 06_Backtest_Methodology
- 07_Backtest_Simple_vs_Compound
- 08_Research_Strategy_Decay
- 09_Tech_Architecture
- 10_Tech_LLM_RAG_Integration
- 12_Tech_Actions_Operations
- 13_Tech_Python_Engineering