There is a endogenous variable named ‘zg’ in the code, but in the paper, this variable is never mentioned. What is this variable?

Which code are you talking about?

The code in the supplementary material of Competition, Markups and Predictable Returns by A. Corhay, H. Kung and L. Schimid, published on Review of Financial Studies.

Can you provide it?

Do you also have a copy of the paper?

It looks like a growth factor that appears when detrending the model.

Thank you, professor. My guess is same to yours. Here is another question. What are the techniques used to calculate utility, in the first three rows of model? The first row is the Epstein-Zin preference utility.

The implementation of Epstein-Zin preferences requires the use of auxiliary variables.

I tried to use auxiliary variable by expressing the expectation with the next period deterministic value, as your method shows, but I did not work out due to too large residual. What are the techniques in their code? The exp(eu) on the second row uses many unknown parameters. It is not only the use of auxiliary variables.

It’s only about auxiliary variables. However, their implementation introduces a normalizing constant to improve numerical properties. Essentially, they choose `VCONST`

to get `eu_aux=0`

, i.e. exp(eu_aux)=1`.