AgentSkillsCN

risk-measurement

适用于估算各交易账簿VaR、预期缺口、回撤风险以及风险敞口分布的风险计量工作流。当任务涉及量化风险指标的测算，而非治理流程的管理时，可选用此类工作流。

SKILL.md
--- frontmatter
name: risk-measurement
description: "Risk measurement workflows for estimating VaR, expected shortfall, drawdown risk, and exposure distributions across trading books. use when tasks involve quantifying risk metrics rather than managing governance processes."

Risk Measurement

objective

Measure portfolio and strategy risk with calibrated statistical and scenario-based metrics.

workflow

  1. •define horizon, confidence levels, and risk-factor mapping.
  2. •estimate historical and parametric VaR/ES metrics.
  3. •run stress scenarios and tail-loss decomposition.
  4. •backtest risk forecasts against realized pnl outcomes.
  5. •publish metric sets only after calibration and exception checks.

required diagnostics

  • •VaR and ES calibration error by book and horizon.
  • •exception rate and clustering diagnostics.
  • •tail dependence and correlation-break analysis.
  • •stress-loss attribution by factor family.
  • •drawdown distribution and recovery-time statistics.

risk controls

  • •enforce model versioning and reproducible risk runs.
  • •enforce exception-triggered recalibration rules.
  • •enforce data completeness checks before measurement runs.

outputs

  • •run python scripts/risk_measurement_diagnostics.py input.csv --output diagnostics.json and keep the json artifact.
  • •write an implementation memo using references/risk-measurement-playbook.md with assumptions, tests, limits, and rollout plan.

resources

  • •use scripts/risk_measurement_diagnostics.py for deterministic diagnostics.
  • •use references/risk-measurement-playbook.md for the domain checklist and delivery structure.