Python for Finance Cookbook

There's more...

Calculating VaR using different approaches has its drawbacks, some of which are:

• Assuming a parametric distribution (variance-covariance approach).
• Not capturing enough tail risk.
• Not considering the so-called Black Swan events (unless they are already in the historical sample).
• Historical VaR can be slow to adapt to new market conditions.
• The Historical Simulation Approach assumes that past returns are sufficient to evaluate future risk (connects to the previous points).

Another general drawback of VaR is that it does not contain information about the size of the potential loss when it exceeds the threshold given by VaR. This is when Expected Shortfall (also known as conditional VaR or expected tail loss) comes into play. ...