--- name: alpha-monitor description: > Factor monitoring. Check active factors for IC decay and health issues. 因子监控。检查活跃因子的IC衰减和健康状态。 Triggers: "health check", "monitor", "检查因子健康", "因子状态" --- # alpha-monitor — Factor Health Monitoring / 因子健康监控 你是一个因子监控系统。检查因子库中所有活跃因子的当前健康状态。 You are a factor monitoring system. Check health status of all active factors in the library. ## Bilingual Terms / 双语术语 | English | 中文 | |---------|------| | Factor | 因子 | | IC (Information Coefficient) | 信息系数 | | ICIR (IC Information Ratio) | IC信息比率 | | Quintile | 五分位/分组 | | Long-Short | 多空 | | Sharpe Ratio | 夏普比率 | | Max Drawdown | 最大回撤 | | Monotonicity | 单调性 | | Robustness | 鲁棒性 | | Holding Period | 持有期 | | Factor Registry | 因子注册表 | | Backtest | 回测 | | Gate Check | 门控检查 | ## 项目定位 / Project Context **Multi-Market Support / 多市场支持**: Alpha Skills support A-share (default), HK, and US stocks via data adapters: Alpha Skills 通过数据适配器支持A股（默认）、港股和美股： ```markdown # .claude/alpha-agent.config.md MARKET: A-share # or "HK" or "US" DATA_MODULE: (leave empty for A-share Tushare default) # or "examples.us_data_yfinance" # or "examples.hk_data_yfinance" ``` When a custom DATA_MODULE is set, the skill loads MARKET_CONFIG from that module to determine benchmark, cost rate, and trading rules. 设置自定义DATA_MODULE时，skill从该模块加载MARKET_CONFIG来确定基准、成本和交易规则。 **Language Rule / 语言规则**: - If the user speaks English, output in English - If the user speaks Chinese, output in Chinese - Table headers always show both languages: "IC Mean IC均值" ## 执行流程 / Execution Pipeline ### Step 1: 读取因子库 / Read Factor Library ```python import sys, os, sqlite3, json, uuid from datetime import datetime PROJECT_DIR = '<当前工作目录 current working directory>' # ── 因子注册表（自包含，无外部依赖）/ Factor Registry (self-contained, no external deps) ── class FactorRegistry: def __init__(self, db_path="alpha_skills.db"): self.db_path = db_path with sqlite3.connect(db_path) as conn: conn.execute("""CREATE TABLE IF NOT EXISTS factors ( id TEXT PRIMARY KEY, name TEXT UNIQUE NOT NULL, expression TEXT NOT NULL, category TEXT, description TEXT, status TEXT DEFAULT 'active', market TEXT DEFAULT 'A-share', ic_mean REAL, icir REAL, best_holding_period INTEGER, quality TEXT, eval_date TEXT, created_at TEXT DEFAULT CURRENT_TIMESTAMP, metadata TEXT)""") def register(self, name, expression, **kwargs): fid = str(uuid.uuid4())[:8] with sqlite3.connect(self.db_path) as conn: conn.execute( "INSERT OR REPLACE INTO factors (id,name,expression,category,description,status,market,ic_mean,icir,best_holding_period,quality,eval_date,metadata) VALUES (?,?,?,?,?,?,?,?,?,?,?,?,?)", (fid, name, expression, kwargs.get("category",""), kwargs.get("description",""), "active", kwargs.get("market","A-share"), kwargs.get("ic_mean"), kwargs.get("icir"), kwargs.get("best_holding_period"), kwargs.get("quality"), datetime.now().isoformat(), json.dumps(kwargs.get("metadata",{})))) return fid def list_all(self, status=None): with sqlite3.connect(self.db_path) as conn: conn.row_factory = sqlite3.Row if status: rows = conn.execute("SELECT * FROM factors WHERE status=? ORDER BY icir DESC", (status,)).fetchall() else: rows = conn.execute("SELECT * FROM factors ORDER BY icir DESC").fetchall() return [dict(r) for r in rows] def get(self, name): with sqlite3.connect(self.db_path) as conn: conn.row_factory = sqlite3.Row row = conn.execute("SELECT * FROM factors WHERE name=?", (name,)).fetchone() return dict(row) if row else None def update_status(self, name, status): with sqlite3.connect(self.db_path) as conn: conn.execute("UPDATE factors SET status=? WHERE name=?", (status, name)) def delete(self, name): with sqlite3.connect(self.db_path) as conn: conn.execute("DELETE FROM factors WHERE name=?", (name,)) reg = FactorRegistry(os.path.join(PROJECT_DIR, "alpha_skills.db")) active_factors = reg.list_all(status='active') ``` 如果因子库为空，提示用户先注册因子。 If library is empty, prompt user to register factors first. ### Step 2: 加载近期数据 / Load Recent Data 加载最近1年的数据用于计算滚动指标（同alpha-evaluate的数据加载流程）。 Load last 1 year of data for rolling metrics (same as alpha-evaluate data loading pipeline). ### Step 3: 对每个活跃因子计算健康指标 / Compute Health Metrics for Each Active Factor 对每个因子 / For each factor： 1. 重新计算因子值 Recompute factor values（使用注册时的expression/函数调用 using registered expression/function call） 2. 计算近60日滚动IC均值 Compute 60-day rolling IC mean 3. 计算近60日滚动ICIR Compute 60-day rolling ICIR 4. 与注册时的ICIR对比，计算衰减比例 Compare with registered ICIR, compute decay ratio 5. 计算近3个月IC>0的占比 Compute IC>0 ratio for last 3 months 6. 判定健康状态 Determine health status ### Step 4: 健康状态判定 / Health Status Determination ``` 🟢 HEALTHY: Rolling ICIR 滚动ICIR ≥ Registered ICIR 注册ICIR × 50% 🟡 WARNING: Rolling ICIR 滚动ICIR < Registered ICIR 注册ICIR × 50% or 或 2 consecutive months IC negative 连续2月IC为负 🔴 ALERT: Rolling ICIR 滚动ICIR < 0 or 或 4 consecutive months IC negative 连续4月IC为负 or 或 quintile monotonicity collapse 分组单调性崩溃 ``` ### Step 5: 对WARNING/ALERT因子生成诊断 / Generate Diagnosis for WARNING/ALERT Factors 分析可能的原因 Analyze possible causes： - 市场环境变化 Market regime change（趋势→震荡 trending→range-bound，或反之 or vice versa） - 因子拥挤 Factor crowding（与主流指数相关性升高 increased correlation with major indices） - 数据问题 Data issues（缺失率升高 higher missing rate） - 季节性效应 Seasonal effects 给出建议 Provide suggestions： - 🟡 WARNING: "Reduce weight, continue monitoring / 降低权重，继续观察" - 🔴 ALERT: "Suspend usage, consider retiring or finding replacement / 暂停使用，考虑退役或寻找替代" ### Step 6: 输出报告 / Output Report ``` 🏥 Factor Health Report / 因子健康报告 (YYYY-MM-DD) 🟢 pv_diverge Rolling ICIR 滚动ICIR=0.62 (Registered 注册 0.70) Healthy 健康 🟢 turnover_20 Rolling ICIR 滚动ICIR=0.48 (Registered 注册 0.52) Healthy 健康 🟡 volatility_20 Rolling ICIR 滚动ICIR=0.19 (Registered 注册 0.43) ← Decay 衰减 56% Diagnosis 诊断: Suggestion 建议: Reduce weight, observe 1 month / 降低权重，观察1个月 Summary 总览: N🟢 N🟡 N🔴 / Total 共 N active factors 个活跃因子 ``` ### Step 7: 更新注册表状态 / Update Registry Status ```python reg.update_status("factor_name", "warning") # or 或 "alert" ``` ## 注意事项 / Notes 1. 如果因子库只有少量因子，监控仍然运行（哪怕只有1个）/ Monitoring still runs even with just 1 factor 2. 滚动窗口60日，不够60日时用全部可用数据 / Rolling window 60 days, use all available data if less than 60 days 3. 诊断信息由AI推理生成，基于市场环境和因子特性 / Diagnosis is AI-generated based on market environment and factor characteristics 4. 建议更新状态但不自动退役，退役需要用户确认 / Suggest status update but no auto-retire, retirement needs user confirmation