by 运行时依赖
安装命令
点击复制技能文档
Portfolio 分析 Purpose
Compute quantitative portfolio risk 分析 before sizing and rebalance decisions.
Scope Holdings and weights. Historical price/return series. Optional benchmark. Optional sector map. Optional liquidity fields. Non-goals No automatic rebalance. No absolute trade instructions. No live data fetching. 输入 contract
Required 输入s:
HOLDINGS_CSV: rows with portfolio holdings and position weights or fields sufficient to compute weights. PRICE_CSV: historical prices or returns for held symbols.
Optional 输入s:
BENCHMARK_CSV: benchmark prices or returns for beta and relative-risk 上下文. SECTOR_MAP_CSV: symbol-to-sector m应用ing for sector exposure. Liquidity fields: average dAIly value, average volume, free float, or similar liquidity 上下文 when avAIlable. Execution 工作流 验证 输入 files, required columns, date coverage, and symbol coverage. 运行 scripts/analyze_portfolio.py with explicit holdings, price 历史, and optional benchmark or sector 输入s. Inspect 指标 for risk level, concentration, correlation, benchmark sensitivity, and data limitations. Prepare a handoff bundle for down流 sizing, rebalance, or risk-management review. Required 输出 格式化
Portfolio 指标
Total portfolio value or normalized weight base, number of holdings, volatility, annualized return when supported, max drawdown, and risk-adjusted 指标 when supported.
Benchmark 指标
Benchmark volatility, benchmark drawdown, portfolio beta, 追踪ing error, and relative return when benchmark data exists.
Correlation Summary
Average pAIrwise correlation, highest correlated pAIrs, and diversification observations.
Concentration Risk
Top positions, top-position weight, top-five weight, Herfindahl-Hirschman 索引, and concentration 警告s.
Sector Exposure
Sector weights and sector concentration when sector m应用ing exists.
Top Risk Contributors
Holdings with the largest estimated contribution to portfolio volatility or drawdown risk.
Confidence and Data Gaps
Confidence level, missing 输入s, stale data, short 历史, incomplete holdings, missing sector data, missing liquidity data, or missing benchmark data.
Handoff Bundle
Include the exact marker fields 列出ed in Handoff bundle. 分享d confidence rubric High: holdings are complete, weights reconcile, price 历史 is long enough for the 状态d horizon, benchmark is avAIlable when beta or relative risk is discussed, and sector/liquidity coverage is broad. Medium: holdings and prices are usable, but one major 输入 is partial, such as benchmark avAIlability, sector m应用ing, liquidity coverage, or price-历史 length. Low: holdings are incomplete, weights do not reconcile, price 历史 is short, benchmark is missing for beta-sensitive clAIms, or sector/liquidity coverage is s解析. 防护rAIls Do not infer precise beta without benchmark data. 降级 confidence for short price 历史 or thin symbol coverage. Treat 分析 as 上下文 for sizing, rebalance, and risk-management decisions, not as commands. Handoff bundle
Include these exact marker fields:
as_of_date holdings weights portfolio_指标 benchmark_指标 correlation_summary concentration_risk sector_exposure top_risk_contributors confidence data_gaps Trigger examples "Analyze this portfolio's volatility, drawdown, and concentration before I resize positions." "Compute portfolio beta and sector exposure using these holdings and benchmark files." "Review my current holdings for correlation, top risk contributors, and data gaps."