Synopsis of support: Significant T-tests and Principal Component Analysis

Single distribution vs Mixture of Distributions

single distribution vs mixture of distributions

Single distribution vs Mixture of Distributions

Shaded in green are significant T-tests by row and column for the following data sets: 0) 120 trade days, 2) 60 trade days offset 0 and 3) 60 trade days offset 1 trade day. Significant at alpha = 10% with some significant at 5%.

Principal Component Analysis

retrospective chgpts pcm

Principal Component Analysis

The maximal eigenvalues and associated principal eigenvector are shown by trade date order for a moving 60 trade day window. Of note, wherever we see the simultaneous negation of the principle eigenvector coefficients from prior measurement, a retrospective change point has been detected.