Principal Component Analysis (PCA) is a stable and good way of calculating VaR on a portfolio using a model building approach There has been several attempts to estimate factors for a PCA analysis. The well known model, for a long period of time, below can be used Factor loadings PC1 1 0,216 2 0,331 3 0,372 4 0,392 5 0,404 Factor scores Factor PC 1 PC 2 PC 3 PCA model PC2 -0,501 -0,429 -0,267 -0,11 0,019 Score 17,55 4,77 2,08 PC3 0,627 0,129 -0,157 -0,256 -0,355 What is the 99% 10 day VaR for the portfolio using all three factors estimated in the PCA model Would you use all the factors in reality