Dynare forums

Hi there,

I have been having a lot of trouble with a simple model with housing, firms who own capital, and a government sector. I cannot get it to work, and the error that is really giving me trouble is the following:

There are 2 eigenvalue(s) larger than 1 in modulus
for 2 forward-looking variable(s)

The rank conditions ISN'T verified!

Warning: Matrix is close to singular or badly scaled.
Results may be inaccurate. RCOND = 7.277823e-17.

In reading other posts my best guess is that there is a problem with my steady-state, but I can't find the error and I don't know what to do. Can anyone help me? Any hints welcome

Hi, I think there are two problems. First, Lambda has the wrong timing. It seems to be an auxiliary variable for the stochastic discount factor. Hence, there should be no (+1). The problem is that it does not appear as a variable in the current period. Second, Ricardian equivalence seems to hold in your model, meaning that lump sum transfers t and bonds b cannot be determined individually. As they are perfectly collinear, this might be the reason for the near singularity warning. What is missing may be an exogenous process for t. Finally, I am not sure about the timing of r. Having it predetermined seems a bit strange.

jpfeifer wrote:Hi, I think there are two problems. First, Lambda has the wrong timing. It seems to be an auxiliary variable for the stochastic discount factor. Hence, there should be no (+1). The problem is that it does not appear as a variable in the current period. Second, Ricardian equivalence seems to hold in your model, meaning that lump sum transfers t and bonds b cannot be determined individually. As they are perfectly collinear, this might be the reason for the near singularity warning. What is missing may be an exogenous process for t. Finally, I am not sure about the timing of r. Having it predetermined seems a bit strange.

Hey, sorry for the late reply, and thanks

You are right, there are some timing mistakes. But even correcting for them I could not get a solution.

After trying every possible combination (I am not proud of it!), finally I decided to get rid of bonds (and the associated interest rate) and taxes, and so I got rid of 3 variables and 2 equations (the government budget constraint and the interest equation from the household optimality conditions). Once that was done I could get rid of the resource constraint (which is derived using the household budget constraint and the firm's profit equation) and everything worked fine.

I was thinking this was the problem: I had an equation (resource constraint) which is a linear combination of two other equations, and hence my matrix was singular. I thought that this could be due to two problems:

1. The interest rates (r and rk) are the same, so bonds are redundant as a source of savings

2. I was missing an equation.

Your suggestion that maybe tax is the source of the problem and that I am missing the process for why seems reasonable and I will investigate that (I remember trying to get rid of taxes and not getting an answer, but it might have been due to the wrong timings).

Again, I really appreciate the help