How does a copula help calculate portfolio VaR?

Question

AcadiFi · Accepted Answer

A copula helps build the joint distribution of the underlying risk drivers. VaR itself is still taken from the one-dimensional distribution of total portfolio loss. The multivariate step comes before that: you model each marginal distribution, connect the marginals with a copula, simulate or integrate joint outcomes, convert those outcomes into portfolio losses, and then read the loss quantile.

Suppose Harborline Credit Portfolio has two risky exposures with different marginal loss distributions. A Gaussian copula or t copula can impose a dependence structure between those losses. Once joint loss pairs are generated, the portfolio loss is simply the sum, such as L = L1 + L2. The VaR is then the selected quantile of that summed-loss distribution.

```mermaid
flowchart LR
  A[Marginal loss 1] --> C[Copula dependence]
  B[Marginal loss 2] --> C
  C --> D[Joint scenarios]
  D --> E[Portfolio loss distribution]
  E --> F[VaR quantile]
```

The FRM takeaway is that copulas do not give VaR by themselves. They give dependence. You still need portfolio aggregation and a final one-dimensional loss distribution. On the exam, also state the horizon, confidence level, data source, modeling assumption, and validation step, because those choices usually explain why two reasonable VaR answers differ materially. Practice more in our FRM question bank.

How does a copula help calculate portfolio VaR?

Master Part II with our FRM Course

Related Questions

Practice Questions