Page 1 of 1
dsge state espace
Posted: Fri Sep 11, 2015 8:26 am
by jocildo_47
Dear Tom. I am a beginner in DSGE. I would like to know how to treat sums when writing formulas to solve DSGE models. My concern at the moment is with the following equations: (i) 4 and 4b in Jadresic, E. (1999) "Sticky Prices: An Empirical assessment of Alternative models" and, (ii) equations 4, 5 and 6 in Carvalho, C. and Dam, N. A. (2010) "The Cross-Sectional Determination of Price stickiness Implied by Aggregate Data". Also, for future works, I want to know how to treat Infinite sums and Infinite sums with changing timing of expectations. I apologize for not having found in the RATS documentation. I fear having undertaken an incorrect search.
Best Regards.
Jocildo Fernandes.
Re: dsge state espace
Posted: Fri Sep 11, 2015 10:00 am
by TomDoan
In the papers cited, those are sums of finite length, so you just expand them out. (The coefficients are independently specified for the different leads, so there is no way to simplify them). Infinite sums have to be eliminated by adding or subtracting a time-shifted version to eliminate the infinite sum.
You might want to get the
State-Space Models and DSGE course which goes into much greater detail than the main documentation.
dealing with sums in dsge
Posted: Sat Sep 12, 2015 11:17 am
by jocildo_47
Okay, Tom, I appreciate the suggestion. I will buy, soon, the DSGE course. I want to be updated with RATS. Before that, however, could you provide or indicate examples of how to use the RATS code to address the issues that I presented to you in my previous message? Thanks a lot for your indispensable help.
Re: dsge state espace
Posted: Mon Sep 14, 2015 12:18 pm
by TomDoan
In the second paper, (6) for k=4 would be written something like
Code: Select all
frml(identity) xt4eq = xt4-(1-beta)/(1-beta^4)*($
(zeta*(m{0}-yn{0})+(1-zeta)*p{0})+$
beta*(zeta*(m{-1}-yn{-1})+(1-zeta)*p{-1})+$
beta^2*(zeta*(m{-2}-yn{-2})+(1-zeta)*p{-2})+$
beta^3*(zeta*(m{-3}-yn{-3})+(1-zeta)*p{-3}))
Re: dsge state espace
Posted: Wed Sep 16, 2015 2:22 pm
by jocildo_47
Dear Tom, many thanks for your help. Still working and, when necessary, returning with doubts. Best regards.