我正在研究一个真正的商业周期模型。我是Dynare计划的新手。当我运行模型时,它说:模数中有7个大于1的特征值 对于10个前瞻性变量
等级条件ISN经过验证!
使用print_info时出错(第45行) Blanchard Kahn的条件不满意:不确定性。
有人知道如何克服这个问题吗?
var
chat
ihat
nhat
qhat
Rhat
Rkhat
khat
zhat
what
hhat
xihat
vahat
vbhat
nthat
vthat
shat
;
varexo eps_z
eps_xi
;
parameters betta
upsilon
alppha
phi
delta
lambda
epsilon
theta
siggma
rho_z
sigma_z
rho_xi
sigma_xi
covariance_z_xi
;
betta=0.9942;
upsilon=1.7167;
alppha=0.36;
phi=3.6;
delta=0.025;
lambda=0.1548;
epsilon=0.001;
theta=0.9685;
siggma=1;
rho_z=0.9315;
sigma_z=0.006424;
rho_xi=0.3744;
sigma_xi=0.0512;
covariance_z_xi=0;
model(linear);
#zstar=1;
#xistar=1;
#qstar=1;
#Rstar=1/betta-1;
#Rkstar=1/betta-1;
#nstar=((1-alppha)/theta*((1/betta-1+delta)/alppha)^(alppha/(alppha-
1)))/((theta+1-alppha)/theta*((1/betta-
1+delta)/alppha)^(alppha/(alppha-
1))-delta*((1/betta-1+delta)/alppha)^(1/(alppha-1)));
#kstar=((1/betta-1+delta)/alppha)^(1/(alppha-1))*nstar;
#cstar =(kstar/nstar)^alppha*(1-alppha)/theta*(1-nstar);
#ystar=(kstar/nstar)^alppha*nstar
#istar=delta*kstar;
#sstar=kstar;
#ntstar=(1-theta)/(1+betta*theta);
#vastar =qstar*sstar*(lambda/ntstar);
#vbstar=vastar;
#vtstar=0;
//1
1/(1-nstar)*nstar*nhat+siggma*chat=what;
//2
siggma*chat(+1)-siggma*chat=Rhat(+1);
//3
Rkhat(+1)*Rkstar=(zhat(+1)+(1-alppha)*khat+(1-alppha)*nhat(+1))*
(nstar/kstar)^(1-alppha)+(qhat(+1)-qhat)*qstar;
//4
what=zhat+alppha*khat(-1)-alppha*hhat;
//5
vahat*vastar=xihat+vbhat*vbstar;
//6
qhat+shat=nthat-vthat*vtstar/(lambda-vtstar)+vahat;
//7
nthat*ntstar=(1-theta)*betta*Rhat(+1)*Rstar+betta*theta*((vahat(+1)-
vahat)+nthat(+1));
//8
vthat*vtstar=(1-theta)*betta*(Rkhat(+1)*Rkstar-
Rhat(+1)*Rstar)+betta*theta*(qhat(+1)-qhat+shat(+1)-shat+vahat(+1));
//9
vbhat(+1)*vbstar=theta*(Rkhat(+1)*Rkstar-
Rhat(+1)*Rstar+nthat+vthat*vtstar/(lambda-
vtstar)+Rhat(+1)*Rstar)+epsilon*vbhat;
//10
shat=khat;
//11
khat=khat(-1)-delta*khat(-1)-ihat*(istar/kstar)+(-phi*delta*(ihat-
khat(-1)))*istar/kstar;
//12
qhat*qstar=(phi*(ihat-khat(-1)))*(istar/kstar);
//13
hhat=nhat;
//14
chat*cstar/ystar+ihat*istar/ystar=zhat+alppha*khat(-1)+(1-
alppha)*hhat;
//15
zhat=rho_z*zhat(-1)+eps_z;
//16
xihat=rho_xi*xihat(-1)+eps_xi;
end;
steady_state_model;
chat=0;
ihat=0;
nhat=0;
qhat=0;
Rhat=0;
Rkhat=0;
khat=0;
zhat=0;
what=0;
hhat=0;
xihat=0;
vahat=0;
vbhat=0;
nthat=0;
vthat=0;
shat=0;
end;
shocks;
var eps_xi; stderr 0.01;
var eps_z; stderr 0.01;
var eps_z,eps_xi=covariance_z_xi ;
end;
resid(1);
steady;
check;
stoch_simul(order=1,irf=20,periods=250);
答案 0 :(得分:0)
通常这意味着模型中存在基本的计时错误,通常与预定变量(如资本或债券或资本)相关。不知道你的模型,这很难说。请注意,stackoverflow可能不是提出这样一个问题的正确位置。 Dynare在https://forum.dynare.org/
有一个专门的论坛在https://forum.dynare.org/t/blanchard-kahn-conditions-not-satisfied-indeterminacy/10484/1发帖后编辑: 你的资本时机结构似乎是正确的,但净值的几乎肯定是错误的。 如果你有
vahat // bank net worth (after the shock) deviations from trend
vbhat // bank net worth (before the shock) deviations from trend
然后,第二个必须相对于时间t的冲击预先确定。但是,他们怎么能像那样同时相关
vahat*vastar=xihat+vbhat*vbstar;