Error: Blanchard & Kahn conditions are not satisfied: no stable equilibrium

[Maida]

Hi there

I’m getting the in the title mentioned error warning when running my mod file below together with:

There are 7 eigenvalue(s) larger than 1 in modulus
for 4 forward-looking variable(s)

The rank condition ISN’T verified!

I’ve already read the forum questions with regards to that error and it seems like it could be the consequence of wrong timing conventions wihtin the equations. However I’ve checked it for my code and it does not seem to be the case.
Could maybe anybody assist herewith?
I’m quite new to this kind of modelling and cannot debug the error.

Thanks in advance
v1.mod (7.8 KB)

[jpfeifer]

My guess is the problem is related to

Timing of capital in two sector economy

Hello all, I am building a two-sectoral model, and I have a question about the timing of sectoral-capital variables: In my model, T and N are sectors, hence kT is capital in sector T and kN is capital in sector N. Then, I assume that the economy’s aggregate capital is simply: k = kT + kN. However, I assume only one law of motion for aggregate capital such that: k = (1-delta)*k(-1) + i - written in Dynare language. It is clear to me that k is a predetermined variable and should be treated as su…

It seems strange to have

k=(1-lambda)*kNLC;

where both variables are predetermined.

[Maida]

Dear Professor Pfeiffer

Thanks for replying that fast. I just read through the post you suggested. Just to make sure I understand it correctly in my context: : the capital chosen by non-liquidity-constraint households (kNLC) would be predetermined, however the total capital k characterized by the equation you find strange is determined in t. Hence implying that the equation becomes: k=(1-lambda)kNLC(-1).
Does that make sense now?

I would appreciate very much some feedback.

[jpfeifer]

I think it’s the other way round. Your law of motion is for k, not kNLC. So kNLC is not predetermined (ditto for bNLC)

[Maida]

Actually I posted an older version of my mod file. The law of motion i for kNLC. I adjusted my code now accordingly, meaning kNLC and bNLC are predetermined, k and b not (see below mod file). However it gives me the same error message. Could you please have a look and share some thoughts.

That would be very valuable for me.

Many thanks.

v1.mod (7.8 KB)

[jpfeifer]

Now the timing looks better, but the fiscal rules with b(+1) may be an issue. Does the model work with lump sum taxes?

[Maida]

Those are actually lump sum taxes in a J. Gali, J. D. Lopez-Salido, and J. Valles (2007) fashion.

[jpfeifer]

What did you change compared to that model?

[Maida]

I added 4 shocks: a demand shock, production shock, monetary policy shock and a fiscal policy shock. I redistributed all intermediate firm’s profits to the non-liquidity constraint households (which are the dividends). Not sure how the paper deals with it. I do not have yet different lump sum taxes for the two households(the papers has that). And I have another taylor rule than in the paper. Additionally what I want to investigate specifically are the reactions of the consumption of LC and NLC households, rather than total consumption. Hence this is the reason why my equations do not enter in aggregate form only within the code. I wanted to implement a baseline model which works and build on that.

[Maida]

I tried to find the problem by checking again the timing, varying parameters and going through the equations, however I cannot find the mistake. Could you please elaborate on why the fiscal rules with b(+1) may be a problem?

Your thoughts and tipps would be of great help to me.

Many thanks

[jpfeifer]

I would recommend starting from a working version of the Galí et al paper. Otherwise, it’s extremely hard to debug.