Principal Components Analysis

#1
Hello everybody,
I have 15 time series consisting of returns on classic asset class investment (Stocks, Bonds, etc.) and some more advanced strategies which are less correlated with the classic one. The result of my PCA shows that the first PC mostly loads on these classic investments whereas the loadings of other PC are rather split among different strategies so I do not have a clear cut picture. In the next step I regress the first three components on some macro variables like market volatility and inflation to test how sensitive they are towards different macro regimes.
So the question here is whether it is legitimate to apply the varimax rotation to the three components and use rotated scores in the further regression analysis. On one hand it would facilitate a better interpretation of the components (what I need) but on the other hand it also affects the significance levels in the further analysis.
As I’m new to this topic I would very appreciate any comments.
Thanks a lot!
 
Last edited by a moderator: