Add riskiness of bonds in a closed economy

General DSGE Modeling
1

I want to apply optimal fiscal/monetary policy analysis to my model in a similar manner to SGU. The key difference is that I’m doing the analysis to a developing economy (also closed), my question is, in order to be a little bit more accurate in the modeling maybe adding some sort of notion to the riskiness of developing economies’ govt. bonds, but I’m not sure how to proceed in such modelling. Since usually when I’ve seen this sort of thing it’s modeled through an spread between local and foreign interest rates, as this model is closed I’m not sure how to include this fact.

Then I’d be grateful if you may provide me some literature for modelling this fact without overcomplicate the model (since that’s not going to be its main feature), or even advise me in if it may not be very worthy to include such detail.

Thanks!

2

Is this spread supposed to be time-varying?

1 Like
3

It depends, if it doesn’t overcomplicate the model yes, maybe something like:

(1+i_t)(1+sp_t)?

Where sp_t follows an stochastic process? Or would it be different? Some literature for guidance in the options would be very useful. Also I was checking and may you recommend me some literature on how to properly include the government in an economy with cash, since I have a little confusion with market clearings in that case.

Thanks!!

4

It all depends on your exact model. Which interest rate does the central bank govern? The risk-free one? Smets/Wouters (2007) already have a risk-premium shock in their Euler equation.

1 Like
5

Indeed, CB controls would control i_t, which corresponds to risk-free return of bonds. I was checking Smets and Wouters (2007), and for what I can understand it’s more or less as the form: (1+i_t)(1+sp_t), right? In that sense the government budget constraint should internalize this new cost of debt?

Thanks a lot for your help Professor Pfeifer!!

6

Yes, it’s more or less this. What do you mean with “internalize”? Yes, the government should have to pay these additional costs. This would differ from Smets/Wouters, where only private agents would be affected by the risk premium.

1 Like
7

But at the end, in a closed economy, unless I include it as a shock for bayesian estimation (as in Smets/Wouters), or specifically study responses to an spread shock, including such a risk premium would not be a crucial change, right? I mean, if my model is not focused on studying specifically that interest rate spread, adding it seems not to contribute much to the focus of my model, or at least that’s what I’m realizing. By the way, my model is focused in monetary policy and labor markets. Thanks!!

8

Yes, it sounds that way.

1 Like