Everyone Keeps Asking: What Is VAR And How To Calculate It?
What VAR means
Value at Risk, usually shortened to VAR or VaR, is a financial risk measure that estimates the biggest loss a portfolio, position, or business line is likely to face over a chosen time period at a chosen confidence level. In plain English, it answers: "How bad could a normal-market loss get, and how often might that happen?"
For example, a 95% one-day VaR of $1 million means there is a 95% confidence that losses will not exceed $1 million over one day, while 5% of outcomes could be worse. That makes VAR useful for traders, risk managers, and finance teams that need a single number to describe downside risk.
How VAR works
VAR depends on three core inputs: the time horizon, the confidence level, and the size of the portfolio or exposure. Common settings are one day, one month, or one year for the time period, and 95% or 99% for confidence.
The key idea is statistical: instead of asking what the average loss is, VAR asks what loss threshold is only breached in the worst tail of outcomes. In pension and investment settings, a 95% VAR is often described as the "worst 1 in 20" scenario, while 90% is "worst 1 in 10."
Simple calculation
The easiest way to calculate VAR is the historical method, which ranks past returns from best to worst and then selects the loss at the relevant percentile. For example, if you use 250 trading days and want a 95% one-day VAR, you look at the 5th percentile of those daily returns, which is roughly the 13th worst day in the sample.
A simplified formula for the parametric method is:
VAR = Z x σ x Portfolio Value
Here, Z is the confidence multiplier, σ is the portfolio's standard deviation, and Portfolio Value is the current market value of the position. A 95% confidence level often uses a Z-score around 1.65, while a 99% level is higher because the tail is more extreme.
Worked example
Imagine a portfolio worth $20,000 with daily volatility of 2% and a 95% confidence level. Using the parametric approach, one-day VAR is approximately 1.65 x 2% x $20,000 = $660, meaning the portfolio is expected not to lose more than about $660 on 95% of days under the model's assumptions.
That number is not a guarantee. It is a model-based threshold built on historical volatility or assumed return behavior, and it can understate risk when markets move violently or returns are not normally distributed.
Calculation methods
Most practitioners use one of three methods: historical VAR, parametric VAR, or Monte Carlo VAR. Each method estimates the tail of the loss distribution differently, which is why two analysts can produce different VAR numbers for the same portfolio.
- Historical VAR: Uses actual past returns and picks the percentile loss directly from history.
- Parametric VAR: Assumes returns follow a distribution, usually normal, and calculates risk from mean and standard deviation.
- Monte Carlo VAR: Simulates thousands of possible future paths and measures the loss distribution from those scenarios.
Illustrative data
The table below shows how the same hypothetical $1,000,000 portfolio can produce different VAR estimates depending on confidence level and horizon. These figures are illustrative only, but they reflect the way higher confidence and longer horizons increase the estimated loss threshold.
| Portfolio value | Horizon | Confidence | Illustrative VAR |
|---|---|---|---|
| $1,000,000 | 1 day | 95% | $18,000 |
| $1,000,000 | 1 day | 99% | $30,000 |
| $1,000,000 | 10 days | 95% | $57,000 |
| $1,000,000 | 10 days | 99% | $95,000 |
Step by step
If you want to calculate VAR yourself, the basic workflow is straightforward. The exact method changes the math, but the structure stays the same.
- Choose the portfolio or exposure you want to measure.
- Pick a time horizon, such as one day, one month, or one year.
- Pick a confidence level, such as 95% or 99%.
- Estimate return volatility or collect historical returns.
- Apply the historical, parametric, or Monte Carlo method.
- Read the loss threshold at the chosen percentile.
Why firms use it
VAR is popular because it condenses complex portfolio risk into one number that executives, traders, and regulators can understand quickly. It is often used in risk reporting, limit setting, capital planning, and scenario review.
It is also useful because it can compare different portfolios on the same scale. A lower VAR usually means a portfolio is less exposed to large short-term losses, although the figure depends heavily on the chosen assumptions.
Limits of VAR
VAR has important blind spots. It tells you the loss threshold at a confidence level, but it does not tell you how bad losses can be once that threshold is broken.
VAR can also miss tail risk, assume overly smooth return patterns, and produce different answers depending on the model. That is why many risk teams pair it with expected shortfall, stress testing, and scenario analysis.
"A good VAR number is useful, but only if you remember it is a model, not a forecast of certainty."
Historical context
VAR became a mainstream risk metric in the 1990s as banks and investors looked for a standardized way to measure market risk. By the early 2000s, it was widely used in trading desks, asset management, and regulatory reporting because it translated probability into a loss amount people could act on.
Even today, VAR remains one of the most recognized risk statistics in finance because it is simple to communicate, easy to compare, and flexible enough to apply across asset classes. Its simplicity is also its weakness, since real-world markets often behave in ways that the underlying assumptions do not fully capture.
FAQ
Practical takeaway
The simplest way to think about risk threshold is this: VAR tells you how much you could lose under normal market conditions before you enter the worst tail of outcomes. If you want a number that is quick to explain and useful for comparing exposures, VAR is a strong starting point; if you want a full picture of tail danger, it should be paired with stress tests and expected shortfall.
Expert answers to Everyone Keeps Asking What Is Var And How To Calculate It queries
What is VAR in finance?
VAR, or Value at Risk, is a statistical estimate of the maximum expected loss over a specific time period at a chosen confidence level. It is one of the standard ways to summarize market risk.
How do you calculate VAR?
You calculate VAR by choosing a time horizon and confidence level, then using historical returns, a volatility-based formula, or simulation to find the loss at the selected percentile. The most common methods are historical VAR, parametric VAR, and Monte Carlo VAR.
What does 95% VAR mean?
A 95% VAR means the portfolio should not lose more than the stated amount over the chosen period in 95 out of 100 modeled outcomes. The remaining 5% of outcomes represent worse-than-VAR losses.
Is VAR the same as expected loss?
No, VAR is a threshold loss number, while expected loss is the average loss you would expect over many events. VAR does not tell you how severe losses are beyond that threshold.
Which VAR method is best?
There is no single best method. Historical VAR is easy to understand, parametric VAR is fast and efficient, and Monte Carlo VAR is flexible for complex portfolios, but all three depend on assumptions and data quality.