VaR and CVaR are two downside-risk measures used to understand severe losses.
The difference is straightforward:
VaR (Value at Risk) = a severe-loss threshold.
CVaR (Conditional Value at Risk) = the average loss beyond that threshold.
If VaR tells you where the bad tail begins, CVaR helps explain how bad losses become once you are inside it.
That makes the two measures complementary rather than interchangeable.
Educational disclaimer: TradingSimuLab is an educational research platform. VaR and CVaR are model-based risk estimates, not guaranteed loss limits, forecasts, or financial advice.
What Is Value at Risk?
Value at Risk (VaR) estimates a downside threshold at a chosen confidence level and time horizon.
Suppose a simulation reports:
95% VaR: -12%
A simplified interpretation is:
The modeled loss threshold is about 12% at the 95% confidence level.
But this does not mean 12% is the maximum possible loss.
Outcomes can be worse than VaR.
That is the main limitation users need to remember.
What Is CVaR?
Conditional Value at Risk (CVaR) looks beyond the VaR threshold.
It asks:
When outcomes fall into the severe-loss tail, how bad are those losses on average?
CVaR is also commonly called Expected Shortfall.
For example:
95% VaR: -12%
95% CVaR: -18%
The first number identifies the tail threshold.
The second suggests that outcomes beyond that threshold average a substantially larger loss.
This reveals information that VaR alone cannot show.
VaR vs CVaR
| Metric | Main question |
|---|---|
| VaR | Where does the severe-loss tail begin? |
| CVaR | How severe are losses inside that tail on average? |
A useful shortcut is:
VaR = threshold.
CVaR = tail severity.
Neither represents the worst possible outcome.
Why VaR Alone Can Be Misleading
Imagine two assets with the same VaR:
Asset A VaR: -10%
Asset B VaR: -10%
Looking only at VaR, their downside may appear similar.
Now add CVaR:
Asset A CVaR: -13%
Asset B CVaR: -25%
The difference becomes much clearer.
Both have a similar loss threshold.
But once Asset B moves into its bad tail, simulated losses become much more severe.
This is why CVaR can reveal tail risk that VaR hides.
Why CVaR Matters for Tail Risk
Tail risk refers to unusually severe outcomes at the extreme end of a return distribution.
VaR identifies a point near that tail.
CVaR examines the losses beyond it.
When CVaR is much worse than VaR, the modeled tail may be particularly deep.
That can materially change the interpretation of an otherwise attractive setup.
For example:
Expected Return: Positive
Probability of Gain: Constructive
VaR: Moderate
CVaR: Severe
The upside metrics may look encouraging.
But the downside tail remains uncomfortable.
Risk analysis should keep that conflict visible.
VaR and CVaR vs Max Drawdown
VaR, CVaR, and Max Drawdown answer different questions.
VaR: downside threshold.
CVaR: average severity beyond the threshold.
Max Drawdown: peak-to-trough decline along a path.
This distinction matters because tail losses and path risk are not identical.
An asset can have an acceptable VaR reading while still producing uncomfortable simulated drawdowns.
That is why TradingSimuLab presents multiple risk metrics together.
How TradingSimuLab Uses VaR and CVaR
Inside Risk Simulation, VaR and CVaR are read alongside:
- Expected Return;
- Probability of Gain;
- terminal price ranges;
- Average Max Drawdown;
- Worst Max Drawdown;
- broader risk-reward context.
The goal is not to predict an exact future loss.
It is to make the shape and severity of modeled downside easier to understand.
TradingSimuLab explains these outputs at the interpretation level without publishing proprietary simulation code, thresholds, or backend implementation.
Example: Same VaR, Different Risk
Consider:
Scenario A
VaR: -9%
CVaR: -12%
The tail worsens beyond VaR, but not dramatically.
Scenario B
VaR: -9%
CVaR: -22%
The threshold is identical.
The tail is not.
Scenario B contains much greater modeled loss severity once the VaR threshold is breached.
This is the central reason to read VaR and CVaR together.
Common Mistakes
Treating VaR as maximum loss
Losses can exceed VaR.
Treating CVaR as worst-case loss
CVaR is an average across severe tail outcomes, not the single worst scenario.
Comparing risk without matching assumptions
Confidence level, horizon, volatility, and model assumptions matter.
Ignoring drawdown
Tail metrics do not replace path-risk analysis.
Using one risk number as the conclusion
VaR and CVaR work best as part of a broader risk distribution.
Frequently Asked Questions
What is the difference between VaR and CVaR?
VaR identifies a downside threshold. CVaR estimates the average loss beyond that threshold.
Is CVaR the same as Expected Shortfall?
The terms are commonly used for the same tail-risk concept in this context.
Is VaR the maximum possible loss?
No. Losses can exceed the VaR threshold.
Is CVaR always worse than VaR?
For the same modeled distribution and confidence level, CVaR reflects losses deeper in the tail and therefore represents greater average loss severity.
Which is better, VaR or CVaR?
They answer different questions. VaR is useful for locating the tail threshold; CVaR adds information about what happens beyond it.
Does TradingSimuLab use both?
Yes. Risk Simulation presents VaR and CVaR alongside other downside and return metrics.
Final Takeaway
The difference between VaR and CVaR is simple:
VaR tells you where severe losses begin.
CVaR tells you how severe losses become beyond that point.
VaR is useful.
But VaR alone does not reveal the depth of the tail.
For a fuller downside-risk picture, read:
VaR → CVaR → Drawdown → Probability of Gain → Expected Return
That gives you a clearer view of both the opportunity and what can happen when the simulation moves against it.