Jacob contains elements that are infinite andn

Stochastic simulations
1

selfmademodel2_log_crra.mod (663 Bytes)

I’m a new learner of DSGE and this is a modified version of the standard DSGE model(I change the original log-utility function into a log+CRRA one). Thus I cannot find calibration for “sigma” in the textbook so I set one myself and it got the error saying:

The Jacobian matrix evaluated at the steady state contains elements that are not real or are infinite
Error check (line 76)
print_info(info, 0, options);
Error selfmademodel2_log_crra (line 161)
oo_.dr.eigval = check(M_,options_,oo_);
Error dynare (line 235)
evalin(‘base’,fname) ;

Is it because of my calibration or some other problem??Plz help me, thank u guys very much!
(btw:sigma is the relative risk aversion and sigmaa is the standard error of e)

2

Hi,

I think there is more than one problem in your mod file.

  1. The timing of the variables is wrong. Predetermined variables must be in t-1. For instance, the capital stock in the production function should be K(-1) and not K. You have the same problem in the law of motion for the capital stock which should be:

    K=(1-delta)*K(-1)+I;

    and not

    K(+1)=(1-delta)*K+I;

  2. The Euler equation (the fourth equation) looks wrong. First there are again timing issues: (i) the return of capital should be expected (R(1) instead of R), (ii) the consumption ratio should be shifted one period in the future (C(1)/C instead of C/C(-1)). Also the depreciation rate is missing and the elasticity of the marginal utility \sigma is not where it should be (it should be a power on the consumption ratio).

  3. Maybe the value of \sigma is to low.

You should first check your model…

Best,
Stéphane.

3

thank u stephane, I’ve checked my model euqation and there were indeed a mistake and I got my IRF. However it is not about the timing issues, I think according to my previous equations they are correct, so maybe there are problems about my previous equations? Hope u can tell me! Thank u again!