I am wondering if there are some notes or paper to offer a guideline to use Blaschke matrices. It is about the non-invertiblity of the MA components in a time series model. The idea is to flip the roots of a “fundamental” solution to get those “non-fundamentals”, but I don’t know how to find and use the so-called “Blaschke matrix” appropriately. The Lippi and Reichlin (1994 J of Econometrics) only shows a bivariate a example, but no clue how to find the Blaschke matrix.
If you have to have a note or a set of coding to do it, I really appreciate it.