First, the key terms.

Horizon
The period over which loss is measured.
Tail
The extreme end of the modeled loss distribution.

The idea, explained

A risk estimate needs a horizon, probability level and model or historical sample. VaR does not describe the worst possible loss. Expected shortfall provides more information about the modeled tail, but remains sensitive to assumptions and scarce extreme observations. Stress tests examine chosen adverse scenarios, including events unlike recent history. Use several views of risk and inspect liquidity and correlations under stress rather than treating one number as a safety certificate.

STEP 1

Read the full statement

A VaR value without horizon and confidence level is incomplete. A one-day threshold differs from a monthly one. The calculation depends on data and assumptions, especially in markets unlike the sample used.

STEP 2

Examine severity beyond the threshold

VaR identifies a quantile, not the average of worse outcomes. Expected shortfall and stress scenarios ask additional questions. Both remain fallible because extreme events, liquidity and correlated losses are hard to estimate.

Follow a worked example.

A hypothetical one-day 99 percent VaR of $10,000 means the model assigns roughly 1 percent probability to a larger daily loss. It does not mean losses above $10,000 are impossible or capped there.

Walk it through, one step at a time

  1. A hypothetical one-day 99 percent VaR is $5,000. Under the model, about 1 percent of comparable days have larger losses.
  2. A $50,000 loss is not ruled out by that statement. VaR alone does not say how large that worse tail can be.
  3. Explore explicit scenarios and funding needs rather than reporting “maximum loss $5,000.”

Where people get stuck

A model can underestimate tail losses when correlations, volatility or market liquidity change.

Now try it yourself.

Pause before opening each answer. Explain your reasoning in your own words.

1. Check the core idea

Does 99 percent VaR tell you the average size of the worst 1 percent of losses?

Reveal the answer & reasoning

No. It describes a threshold. Expected shortfall is designed to summarize tail severity under its specified definition and model.

2. Apply what you learned

Could two portfolios have the same VaR but different extreme-loss severity?

Reveal the answer & reasoning

Yes. Their quantile thresholds can match while losses beyond the threshold differ greatly. Tail severity requires additional information.