Risk measures
What is Value at Risk and how is it calculated?
Value at Risk (VaR) is the most widely used risk measure among banks, asset managers and insurers. It answers a simple question: how much could I lose, at most, over a given period, on a normal market day?
It reads like this: "one-day 95% VaR of €1,000" means that on 95 days out of 100 the loss should not exceed €1,000. On the other 5 days out of 100 it can be worse.
The three inputs
- Portfolio value, for example €100,000.
- Volatility, i.e. how much the portfolio fluctuates. It comes from the past returns of each holding and from how holdings move together (correlation).
- Confidence level and horizon, usually 95% or 99%, over one day, ten days or a month.
A worked example
Take a €100,000 portfolio with 10% annual volatility. Daily volatility is the annual figure divided by the square root of 252 trading days: about 0.63%.
With the parametric method, 95% VaR is 1.645 times daily volatility: about 1.04%, or roughly €1,040 in one day. At 99% the multiplier is 2.326: about €1,470.
These are illustrative numbers to show the mechanics, not an estimate for a real portfolio.
Three ways to calculate it
- Parametric (variance-covariance): uses volatilities and correlations; fast, and the method used above.
- Historical: applies the returns actually recorded over recent years to today's portfolio and reads the 5th or 1st percentile loss.
- Monte Carlo: simulates thousands of possible scenarios from a statistical model.
The limits of VaR
VaR describes a normal day, not the worst day. In crises assets tend to fall together and correlations rise, so losses can be far larger than VaR. That is why it should always be paired with a stress test that replays real crises such as 2008 or the Covid crash.
A VaR model must also be checked: count the days when the actual loss exceeded VaR (exceptions) and verify there are about as many as expected. This is the backtesting used by banking supervisors.
How Radar del rischio calculates it
Upload a bank statement in Excel or CSV to Analyse your portfolio: each holding is recognised by ISIN, volatilities and correlations come from real market prices, and you get volatility, VaR and each holding's contribution to total risk. For subscribers, Radar Pro adds model backtesting.
Want to see these numbers on your own portfolio? Upload your bank statement: volatility, VaR, each holding's risk contribution and stress tests on real crises, in seconds.
Analyse your portfolioFor information only. Not investment advice: past data does not guarantee future results.
More guides
- Portfolio volatility: what it is and how to read it
- Risk contribution: what really drives portfolio risk
- Portfolio stress test: if a crisis came back
- What happens to markets when a war breaks out
- Correlation and diversification: why many holdings are not enough
- A Bloomberg alternative for portfolio risk analysis