I wanted to create a financial conditions index using principal component analysis. The variables are arranged in a manner that higher/positive values indicate tight monetary conditions. However, the loading factors of PC1 are coming out to be negative. So, finally, when I am creating an index, the positive values of original series are being multiplied by negative factor loadings resulting in negative values leading to difficulty in interpretation of results. The time period with positive values are now reflecting negative index values. Can I multiply all factor loadings with -1 for the sake of my interpretation or would that be inaccurate to do?
Asked
Active
Viewed 14 times
0
-
Why are the negative numbers hard to interpret? – Dave Jul 19 '21 at 04:04
-
Dave, Because higher values of variables indicate tight monetary conditions. But correspondingly, I am getting lower value of index constructed using PCA (which would mean easy conditions) and is a contradiction. Am I missing something? – Abhinandan Jul 19 '21 at 04:26