Dear Gretl Community,
I trying to estimate a time-varying parameter model (TVP) using the Kalman
filter but I'm getting no success :( The model is a TVP Phillips curve for
Brazilian inflation augmented with the trade openness (my sample has 103
observations):
inf_f = b1(t)*inf(-1) + b2(t)*exp12m + b3(t)*gap(-1) + b4(t)*trade(-1) +
w(t)
b1(t) = b1(t-1) + v1(t)
b2(t) = b2(t-1) + v2(t)
b3(t) = b3(t-1) + v3(t)
b4(t) = b4(t-1) + v4(t)
Where:
inf_f - non monitored prices
exp12m - 12 month expected inflation
gap - output gap
trade - trade openness
Based upon equations (26.1) and (26.2) from the Gretl User's Guide (p. 216)
I'm trying to setup the filter. Please take a look at the following
commands:
list regressors = inf exp12m gap trade
matrix y = {inf_f} # Matrix order:(Txn)=(103x1)
matrix H = *?* # Matrix order:(rxn)=(4x1)
matrix x = {regressors} # Matrix order:(Txk)=(103x4)
matrix F = I(4) # Matrix order:(rxr)=(4x4)
matrix Q = *?* # Matrix order:(rxr)=(4x4)
How can I define H and Q?
Best,
--
*Henrique C. de Andrade*
Doutorando em Economia Aplicada
Universidade Federal do Rio Grande do Sul
www.ufrgs.br/ppge