CVaR / Expected Shortfall

média perdas > VaR.
Created by
Renato Passos, Eng. de Software
Reviewed by
Renato Passos, Eng. de Software

Last updated: Apr 18, 2026

CVaR R$
3.000,00

About this calculator

The CVaR (Conditional Value at Risk) calculator measures the average losses exceeding the Value at Risk (VaR) at a given confidence level. Unlike VaR, which estimates maximum loss within a confidence interval, CVaR focuses on the tail risk by calculating the expected loss beyond VaR's threshold. It is widely used in risk management to assess the impact of extreme scenarios, particularly in financial portfolios and investment strategies.

This tool is ideal for analyzing investment portfolios, funds, or individual assets. It operates using historical return data or stochastic simulations, computing the average losses that occur when VaR is exceeded. For instance, if VaR at 95% confidence is $100,000, CVaR might show an average loss of $150,000 in those 5% of cases. This helps managers understand tail risk more comprehensively.

When using this calculator, note that CVaR depends on the quality and breadth of input data. Rare extreme events may not be captured effectively in short samples. Additionally, results are sensitive to the return distribution: asymmetric or high-kurtosis distributions can produce volatile estimates. It is advisable to complement CVaR with other risk metrics like VaR or Sharpe ratio for a balanced analysis.

Frequently asked questions

How does CVaR differ from VaR?

CVaR goes beyond VaR by calculating the average losses exceeding the VaR threshold, whereas VaR only estimates the worst-case scenario within a confidence interval.

For which applications is CVaR most useful?

It is commonly used in portfolio optimization, stress testing, and financial regulations requiring tail risk analysis.

Do I need specific historical data?

Yes, reliable historical return data or simulations are required to compute CVaR accurately.

What are the limitations of CVaR?

CVaR may underestimate risks in rare events not captured by historical data and is sensitive to asymmetric return distributions.

Other VaR calculators