Measure and quantify expected loss from unlikely scenarios by assessing conditional value-at-risk (CVaR)
Conditional value-at-risk (CVaR) is the extended risk measure of value-at-risk that quantifies the average loss over a specified time period of unlikely scenarios beyond the confidence level. For example, a one-day 99% CVaR of $12 million means that the expected loss of the worst 1% scenarios over a one-day period is $12 million. CVaR is also known as expected shortfall.
Practitioners in both risk management and portfolio management are increasingly using CVaR. For example:
- CVaR is replacing VaR for calculating market risk capital in the Fundamental Review of the Trading Book (FRTB) by Basel Committee on Banking Supervision (BCBS).
- CVaR is being adopted for portfolio optimization.
Depending on the asset classes and types of risk exposure, risk managers employ various mathematical techniques to calculate CVaR, including:
- Monte Carlo simulation
- Copula-based portfolio simulation
- Pricing and valuation of financial derivatives
- Econometrics models (e.g., interest rate models and GARCH models)
For more information, see Statistics and Machine Learning Toolbox™, Financial Toolbox™, Financial Instruments Toolbox™, and Risk Management Toolbox™.
Examples and How To
Software Reference
-
Market Risk - Documentation
-
Conditional Value-at-Risk Portfolio Optimization - Documentation
-
portvrisk
: Portfolio value at risk - Function -
varbacktest
: VaR backtesting - Function
See also: risk management, market risk, value-at-risk, backtesting, Basel III, systemic risk, credit scoring model, concentration risk