Skip to main content
aifinhub
Risk & Portfolio Construction Formula

Parametric VaR Formula

Parametric Value-at-Risk assumes returns are normally distributed and computes the loss threshold from the mean, the volatility, and the z-score of the chosen confidence level. Also called the variance-covariance or delta-normal method, it is fast and needs only two moments, but it systematically understates tail risk because real returns have fatter tails than the normal it assumes.

By AI Fin Hub Research · AI Fin Hub Team
Best Next MovePlaygrounds

VaR Backtest — Kupiec & Christoffersen

Paste P&L + VaR series and run Kupiec POF, Christoffersen independence, and joint conditional-coverage tests. Likelihood-ratio χ² p-values.

CalculatorOpen ->

On This Page

Formula

Copy the exact expression or work through it step by step below.

VaR_alpha = - (mu - z_alpha x sigma) x V For a zero-mean assumption: VaR_alpha = z_alpha x sigma x V z_alpha = 1.645 (95%), 2.326 (99%)

Variables

mu

Mean return

Expected periodic return of the portfolio. Over short horizons it is often set to zero, since the volatility term dominates the tail at daily frequency.

sigma

Return volatility

Standard deviation of periodic returns. This is the single risk input that drives the VaR magnitude in the parametric method.

z_alpha

Normal z-score

The standard normal quantile for the confidence level: 1.645 for 95%, 2.326 for 99%. It is the number of standard deviations into the left tail at which the VaR cutoff sits.

V

Portfolio value

Current market value of the position or portfolio, used to convert the return-space VaR into a currency loss.

Step By Step

  1. 1

    Estimate the periodic mean and volatility of portfolio returns.

    Daily mean assumed 0, daily volatility 1.8%.

  2. 2

    Select the z-score for the desired confidence level.

    For 99% confidence use z = 2.326.

  3. 3

    Compute the VaR return as z times sigma (subtracting the mean if it is nonzero).

    2.326 x 0.018 = 0.04187, a 4.19% loss threshold.

  4. 4

    Multiply by portfolio value for the currency VaR.

    On a 2,000,000 portfolio, 0.04187 x 2,000,000 = 83,740.

Worked Example

One-day 99% parametric VaR on a 2,000,000 portfolio, zero mean

Daily volatility

1.8%

Mean return

0%

Confidence (z-score)

99% (z = 2.326)

Portfolio value

2,000,000

VaR return = z x sigma = 2.326 x 0.018 = 0.041868. VaR currency = 0.041868 x 2,000,000 = 83,736.

One-day 99% parametric VaR of about 83,700. Under the normal assumption there is a 1% chance of losing more than this in a day. Because true equity returns are leptokurtic, the real probability of breaching this threshold is higher than 1%, so parametric VaR should be backtested (Kupiec, Christoffersen) and supplemented with a fatter-tailed or historical method.

Common Variations

Student-t parametric VaR: replaces the normal z-score with a t quantile to account for fat tails, raising the estimate.
Cornish-Fisher VaR: adjusts the normal quantile for sample skewness and kurtosis, a middle ground between normal and fully empirical.
Historical VaR: drops the distributional assumption entirely and reads the quantile straight from sorted past returns.

Try These Tools

Run the numbers next

Sources & References

Related Content

Keep the topic connected

Planning estimates only — not financial, tax, or investment advice.