Performance & Risk Metrics
Parametric Value-at-Risk
A statistical method that estimates maximum potential loss over a given time horizon at a specified confidence level by assuming asset returns follow a normal distribution and using volatility and correlation parameters.
Parametric VaR, also called analytical or delta-normal VaR, estimates risk from just two numbers: the mean and standard deviation of returns, assuming they follow a normal distribution. It is fast and easy to compute, which is its main appeal. The cost of that simplicity is the normality assumption, which understates how often extreme moves actually happen.
Because it reduces to a closed-form expression, parametric VaR can be recomputed instantly across many confidence levels (95%, 99%) and rolling windows. That speed is why risk desks use it for routine monitoring, while reserving heavier historical or Monte Carlo methods for portfolios with options or other non-linear payoffs.
Formula
- \mu
- Mean or expected return of the portfolio
- Z_\alpha
- Standard normal quantile at confidence level alpha (e.g., 1.645 for 95%, 2.326 for 99%)
- σ
- Standard deviation or volatility of portfolio returns