I am trying to calculate the VaR of a underlying strategy and didn't find any definitive proposal on old posts.
As I understood from the VaR webinar, and assuming that number of trades and leverage are stable, darwinex takes the last 45 daily returns and it's metrics to run a simulation motor. Then, VaR is extracted from the cumulative 20 day (1 month) returns of all simulations.
My question is, are thoose assumptions true? And, In that case, should the correct way to do such calculations look like...:
1) Take the last 45 day's daily return
2) Extract the mean and standard deviation
3) With that metrics, run.. let's say 1000 simulations of length 20
4) Calculate the cumulative return at the end of each simulation
5) Take the 20 periods cumulative simulation return's 5% quantile's value = That's the expected VaR
Am I on the right way?