Dear all,

I am new on this forum, so I hope I'm posting these questions on the right place.

I am new with Stata and have been using it for just a couple of weeks now for a research skills course I'm following. I'm also in the early stages of writing my MSc thesis for which I'll need Stata as well.

Right now, I'm working on an assignment about the collateral channel, in other words how certain independent variables (Net PPE Ratio, Gross PPE Ratio, Inventories Ratio and Receivables Ratio) impact short and long-term leverage, while also using several control/moderating variables. The dataset we use is a panel dataset consisting of multiple companies and across multiple years.
The variables we use in this assignment are defined as follows:Array
:


Now, I have played around with Stata for a bit and have the following questions:

1. For one question, we had to determine how the variables are correlated with each other. We used a correlation table for this, but in such a table it is difficult to detect multicollinearity. As such, we used the VIF command (Variance inflation factors). This command provides you with the VIF factors for each of the independent variables.
Array

As you see in this table, not all independent variables are there (e.g. Gross PPE is lacking). Does anyone know how to fix this? We are not aware of any other method besides simply using the VIF command.

2. As you see in the first table, many variables are defined as ratios. After running a pooled OLS regression, the coefficient for Net PPE Ratio is 0.0935 (t-stat is 4.50) and the coefficient for Inventories Ratio is -0.197 (t-stat is -3.05). Since the independent variables are defined as ratios, we are not 100% what the correct interpretation is. When considering the Net PPE Ratio, would the correct interpretation be: a one unit change in the Net PPE Ratio is associated with a 0.095 unit change in leverage? Or should we use percentage changes instead of unit changes since the variables are ratios? My same question applies to the other independent variables (Inventories Ratio, Receivables Ratio). We dropped the Gross PPE variable from this regression analysis on purpose. Also the w and 1's behind the variables stand for winsorized and lagged by 1 period.
Array


I hope I was able to phrase my questions somewhat clearly. Please let me know if additional data/information is needed to be able to answer these questions.
Any help would be thoroughly appreciated!

Thanks a lot.