Questions and answers

Q: What is RRR?

A: RRR.ox and RRR.h are files for a program for estimating Reduced Rank Regressions.

Q: How do I run RRR?

A: To run RRR interactively you need both OxMetrics and Ox Professional.

You can also run RRR as a part of Ox Professional.

Q: What sort of systems does RRR estimate?

A: Consider the system Y_t = \Pi R_t + \Gamma U_t + error, where Y_t, R_t and U_t are vectors of

endogenous variables (Y_t), restricted exogenous variables (R_t) and unrestricted exogenous

variables (U_t). RRR can be used to test for reduced rank on the coefficient matrix \Pi

in this system.

Q: Does there exixt an example file for how to use RRR by commando line?

A: Yes, see "CoNonLinearity.ox". This program estimates all the reduced rank regressions in

"Testing for co-nonlinearity by H. Hungnes, forthcoming in Studies in Nonlinear Dynamics &

Econometrics (SNDE).

Related publications

"Testing for co-nonlinearity", forthcomming in Studies in Nonlinear Dynamics & Econometrics. [Discussion Paper version]