intermediate 7 min read
Finance & economics · Topic
Value at Risk
probability theory · gaussian distribution · monte carlo methods
Value at Risk (VaR) summarises portfolio risk as a single number: the loss level that will not be exceeded over a given horizon with a given confidence level. Despite its ubiquity in regulation, VaR has important limitations that motivate Expected Shortfall (CVaR) as a superior measure.

Definition

\[\text{VaR}_\alpha(L) = \inf\{l : P(L > l) \leq 1-\alpha\} = F_L^{-1}(\alpha)\]

The $\alpha = 99\%$ VaR is the 99th percentile of the loss distribution — losses exceed this level on 1% of days.

Calculation methods

Parametric (variance-covariance): assume losses are normally distributed:

\[\text{VaR}_\alpha = \mu_L + z_\alpha\,\sigma_L\]

Fast but misses fat tails. For a portfolio: $\sigma_L^2 = \mathbf{w}^\top\Sigma\mathbf{w}$ (in return space).

Historical simulation: sort the last $T$ daily P&L observations; the $\alpha$-VaR is the $(1-\alpha)T$-th worst loss. Non-parametric, captures fat tails, but limited by history.

Monte Carlo: simulate thousands of scenarios; compute the empirical quantile. Flexible but computationally expensive.

Limitations and Expected Shortfall

VaR fails the subadditivity axiom: $\text{VaR}(A+B) > \text{VaR}(A) + \text{VaR}(B)$ is possible — diversification can appear to increase risk.

Expected Shortfall (CVaR) at level $\alpha$:

\[\text{ES}_\alpha = \mathbb{E}[L \mid L > \text{VaR}_\alpha]\]

ES is subadditive and coherent. The Basel III framework replaced VaR with ES at $\alpha = 97.5\%$ for trading book capital requirements.