Ferrero, Gertler and Svensson (2008)

General DSGE Modeling Replication Attempts
1

Hello together,

I am trying to replicate “Current Account Dynamics and Monetary Policy” from Ferrero, Gertler, Svensson (2008). My problem is that I am not able to build the model correctly in Dynare, because I am not able to derivate all log-linear equations, so that I cant model the foreign country equations. I know this is a Dynare forum, but I dont know where to ask elsewhere.

With help from teaching slides of Ferreros coursework in Oxford, I was able to solve the most basic non-log-linearized equations, but for some of the log-linear equations I have no approach/idea!

Here my biggest problems:

Equation (41) (demand for home tradables):
It seems not to be a normal log-linear version of the non-log-linearized version of the demand for home tradables (equation (34)). So I tried it another way (uploaded-file: Derivation_41), but I cannot reach the exactly same equation like Ferrero.

Equations (47) (48) - NKPCs:
Here the Problems begins by the derivation of the optimal rest price (21). I tried it with following a lot of literature, but I don`t know how to combine the optimal reset price with the evolution of the price index ((22) to reach the NKPCs.

Equation (52) - Net exports:
Here I have also really no idea…

I found a Dynare code for the paper in this forum (from “st.josef” Mar 21, 2016). But at first I don`t know if it is correct and second I cannot explain how he derived the equations for the foreign country, when I am not able to derive all home country equations (moreover it is a version of a currency union, but think this should only be a little change).

I am really thankful for any information (technical appendix?), hint or maybe literature recommendations for DSGE models with trade and current account dynamics, because I am despairing for weeks.

Thanks and greetings!

2

Sorry, but this is not the type of model should be working with in a master’s thesis without strong support by your professor. Even a small model as Gali/Moncelli can be challenging. Typically, even when you manage to rederive every equation, getting the model to correctly run can easily take 6 to 8 weeks.

Big picture question: why do you need to derive all equiations to model the foreign country?

Regarding the NKPC: take a look at Gali’s textbook. Unfortunately, the algebra is rather involved and can take a couple of pages. You might want to take a look at Corsetti’s handbook article as well: sites.google.com/site/giancarlocorsetti/

(52): the paper states what to do, but this is rather involved.