Dear Stata community,

I am currently looking to estimate a somewhat complex regression model to analyze the market value of banks. The model can be found in Calomiris and Nissim (2007), and is pictured below.
or every bank I have data over multiple quarters. Further, as some variables are year on year growth variables in a quarterly data set I expect auto-correlation issues and would like to have the variables follow an AR(q) process. Because of this I thought about using -xtregar-, but I am not sure how to incorporate the FV variables with no coefficient (or a coefficient set ==1) and the brackets. The brackets are needed as the w variables indicate the amount of capital invested in the activity relative to total capital, and the calculation in the brackets yields the value generated for every dollar invested in an activity.

Does anyone have any suggestions?