给自己顶一个，真心求大神帮忙！

You need to linearise your equation to a specification the same as your equation (1). Except that the meaning of the variable will be changed to the deviation of output from steady state (normalised to 0) at time t.

Put all the equations in your dynare file. Remember to specify that your model is linearised.

Yvette

Dear Yvette,
i really appreciate your generous help.
i'm still a little confused though.
in the equation(4), i thought i had  already log-linearised,using the Ulig log-linearised method.was i wrong?
Suppose equation (4) is correct,what bothered me all the time is that how can i input equation (4)into dynare,since the initial value of steady state  Y is still unkown?
or is there any other method to log-linearise,so that i can give all the variable the initial value(zero)?
Help me please ,i really appreciate that.
By the way ,there is a similiar problem,the same,problem,actually.
I've already log-linearised ,and i don't know how to deal with it in dynare.i describe the problem in the attachment,Could you please have a look  at it for me ?I really really appreciate your help!
The 300 forum currency will be sent to you account soon .
Looking forward for your reply.
Yale

Thank you so much for your time, i've uploaded the attachment once again,would you please have a look at it for  me?
In fact ,this is not a problem of a single equation for me, but a series of equations  which i cannont always calibrate a percentage like  yc/y. So ,i  sincerely hope you can have a look at my attachment, my problem has been discribed in detail ,and i really appreciate your help!!
Looking forward for your reply.
Yale

Dear Yvette,
i tried several times,it seems there's a little bug in this forum ,the "pdf" i uploaded can't be opened after i downloaded it by myself!! please,try the "doc"first,and maybe what doesn't work for me may work for you!
Best wishes!
Yale

Hi,

Sorry for the delay. I don't come here often.

For endogenous variables, you will be able to work out their steady state functions in terms of exogenous variables and coefficients. Just like you may find your steady state capital being a function of depreciation rate and other coefficients. These coefficients are all predetermined. A common way is to calibrated according to literature/'stylised fact'. Another way is to estimate using methods such as GMM, Bayesian and MSM, etc. But of course you need some calibrations/priors before your estimation.

And yes, working out SS is difficult. You will be able to work out SS using Dynare. If you input your non-linear version of model into Dynare, it has internal solver to work it out. Or you can specify an external file to solve SS. But yes, it is quite sensitive to initial value. These initial values can be explored from literature. You might have to search quite a few in order to find proper initials.

Just curious, you don't need to answer, or you can answer privately. Are you doing Macroprudential policies?

It is possible that the paper and source code are subject to modifications. You can either wait for their publications or you can email the author. You might get a reply from them.

Hope this is helpful.

Yvette

顶一个！