risk-metrics-calculation

Calculate VaR and CVaR from portfolio return time series by computing historical and parametric tail quantiles, with an optional Cornish-Fisher adjustment for non-normality, to estimate potential tail losses across daily or rolling windows.

Compute Sharpe and Sortino ratios for portfolio monitoring by comparing returns against volatility and downside deviation, producing risk-adjusted performance metrics suitable for evaluating portfolio exposures and setting risk limits.

Calculate maximum drawdown and drawdown duration by deriving a drawdown series from cumulative returns, then summarizing the maximum and average drawdowns alongside duration statistics for downside risk monitoring.

Yes, you can use numpy and pandas to compute rolling window portfolio risk metrics, generating rolling risk regime insights and stress sensitivity analysis across multiple time horizons from your return series.

Yes, this approach supports regulatory-style reporting by consolidating distributional outputs, tail risk metrics, and drawdown statistics into a summary that captures volatility, tail losses, and capital drawdowns for risk teams.

Use a Cornish-Fisher adjustment when estimating expected shortfall to correct for non-normality in return distributions, refining parametric tail quantiles so CVaR estimates better reflect skewed or fat-tailed portfolio outcomes.