Quant Learning Lab

VaR and Expected Shortfall

Learn how probabilistic risk thresholds and tail-loss averages work together in institutional risk reporting.

Model Overview

Value at Risk and Expected Shortfall are core downside-risk tools used by portfolio managers, risk officers, and regulators. VaR summarizes a loss threshold at a chosen confidence level, while Expected Shortfall goes deeper by describing the average loss once that threshold has been breached.

Intuition

VaR tells you where the tail starts; Expected Shortfall helps tell you how painful that tail can be. Together, they move the conversation from ordinary volatility into portfolio resilience under bad outcomes.

Key Formula

VaR_alpha = loss quantile at confidence alpha

ES_alpha = E[Loss | Loss > VaR_alpha]

Tail-risk analysis often complements stress scenarios and drawdown review

Practical Use Case

A risk team can track daily VaR, Expected Shortfall, and scenario stress losses across a portfolio to understand ordinary risk, tail severity, and the types of events that could materially damage capital.

Learning Outcome

This lesson is designed to connect quantitative theory with the exact kind of institutional workflow QuantModels.ai exposes in its pricing and analytics modules.

Static Question Bank

Work through the curated model question bank first, then generate additional mock AI question sets below.

1. What does Value at Risk estimate?

A threshold loss level not expected to be exceeded at a chosen confidence over a horizon

The exact worst possible loss in all scenarios

The average gain in the right tail

The option delta of a portfolio

2. What does Expected Shortfall measure?

The average loss conditional on losses beyond the VaR cutoff

Only the best-case scenario

The realized volatility of a bond

The drift term in Black-Scholes

3. Why is Expected Shortfall often preferred to VaR in tail-risk discussions?

Because it captures tail severity beyond the cutoff

Because it ignores extreme losses

Because it never needs simulation

Because it is always smaller than VaR

4. What does a 99% one-day VaR of INR 10 million mean?

There is a 1% chance losses exceed INR 10 million over one day under the model

The maximum possible loss is INR 10 million

The portfolio earns INR 10 million with 99% certainty

Losses cannot occur on 99% of days

5. Which method can be used to estimate VaR?

Historical simulation, parametric methods, or Monte Carlo

Only a bond duration formula

Only implied volatility surfaces

Only option theta calculations

6. What is a weakness of VaR?

It does not describe how large losses can be once the threshold is breached

It cannot be computed from returns data

It always exceeds Expected Shortfall

It is only valid for bonds

7. Why do regulators and risk teams care about tail metrics?

Because severe but rare losses can threaten solvency and liquidity

Because tail events are always impossible

Because returns are normally distributed in all markets

Because expected return is irrelevant

8. What is historical simulation VaR based on?

Reapplying actual past return moves to the current portfolio

Ignoring the portfolio composition entirely

A deterministic flat-loss assumption

Only the current risk-free rate

9. What is a parametric VaR assumption often made for simple portfolios?

Returns follow an approximate normal distribution

Returns are always zero on average

All assets are perfectly correlated

The portfolio is riskless

10. How are VaR and stress testing different?

VaR is probabilistic while stress testing examines specified adverse scenarios

They are exactly the same concept

Stress testing ignores losses completely

VaR only applies to derivatives and stress only to equities

Generate Unlimited Questions

Use the mock AI agent panel to create additional practice sets by topic and difficulty. The component is already shaped for a future API-backed generation workflow.

AI Placeholder

Model topicDifficultyNumber of questions

Generated Questions

Mock generated set for VaR and Expected Shortfall at beginner difficulty.

1. What does VaR try to summarize for a portfolio? (VaR and Expected Shortfall · beginner · Set 1)

A loss threshold at a chosen confidence level

The exact worst-case loss in all universes

Only positive returns

The option delta of the portfolio

2. Why do risk teams also look at Expected Shortfall? (VaR and Expected Shortfall · beginner · Set 2)

Because it describes the average severity of losses beyond VaR

Because it removes the need for confidence levels

Because it ignores extreme events

Because it only applies to options

3. What is one simple interpretation of a 95% VaR? (VaR and Expected Shortfall · beginner · Set 3)

Losses are expected to exceed that threshold about 5% of the time under the model

The portfolio cannot lose money

Returns are always normally distributed

The portfolio always earns the risk-free rate

Try the model Back to Learning Lab