KCOrtho: Orthogonalize a set of factors with respect to all other...

Description Usage Arguments Details Value

View source: R/KCOrtho.R

Description

This is an implementation of Klien and Chow (2014?) "Orthogonalized Equity Risk Premia and Systematic Risk Decomposition." Original code by Greg Murphy, 2016.

Usage

1
KCOrtho(factors)

Arguments

factors

an xts object containing a time series of independent variables.

Details

This function will return a matrix of factors that have been completed orthogonalized with respect to every other factor in the set while still preserving the orginal sizes of the variances. This is helpful in producing better factor models that have senisble coefficients. Orthogonalizing the brute force way will yield very tiny variances in the resultant factor set and leave you with massive coefficients in your model. This routine nicely solves this problem.

Value

an xts object of factors that have been orthogonalized with respect to all of the other factors in the set.


gtog/dMisc documentation built on May 17, 2019, 8:57 a.m.