[Gretl-users] OPG inversion problem

Dan Běsoň eubie at centrum.cz
Sun Nov 20 06:31:15 EST 2011

Dear everybody, 

I am trying to estimate an ARMA process using MLE and then to extend it by jointly estimating ARMA-GARCH. The following code

series in = log_ret*100

scalar mu = mean(in)
scalar h = var(in)

scalar c = 0.001
scalar phi1 = -0.9
scalar phi2 = -0.89
scalar phi3 = -0.47
scalar phi4 = -0.90
scalar phi5 = -0.45
scalar phi6 = -0.31
scalar phi7 = -0.14
scalar theta1 = 0.87
scalar theta2 = 0.81
scalar theta3 = 0.29
scalar theta4 = 0.82
scalar theta5 = 0.33
scalar theta6 = 0.17

mle ll=-0.5*(log(h) + (e^2)/h)
	series e = in - mu
	params c phi1 phi2 phi3 phi4 phi5 phi6 phi7 theta1 theta2 theta3 theta4 theta5 theta6 
end mle --hessian

produces an "failed to invert OPG matrix GG" error regardless of whether I use --hessian. log_ret are daily logreturns of the FTSE index. What could be the problem? The initial values of phis and thetas are taken close to the real values estimated by the ARMA window. Any hint is much appreciated. 

More information about the Gretl-users mailing list