Value at Risk can also be computed parametrically using a method known as variance/co-variance VaR. This method allows you to simulate a range of possibilities based on historical return distribution properties rather than actual return values. It's actually more accurate even though it assumes a normal distribution especially when not much data is available
Hi, is there a reason VAR should not be defined for shorts as well? Thx
Yelian
⋅
Thank you for sharing your work! Can you please provide some guidance on the script? Do bars extending past VAR imply returns outside of VAR? If so, should there not also be a line to the positive side? It seems that only negative returns can spike to the VAR line
mks17
⋅
@Yelian, Yeas only negative returns can spike the VAR cause is defined for long only positions. Its simply calculating a normal distribution, made with standard deviation and mean from the price sample to gauge the %VAR