0
$\begingroup$

My sample is balanced panel data and includes 6,071 firm-year observations for the period 2002-2014.

I test the book leverage (total debt/total asset) by using lagged factors as follows: 9 firm-specific factors, 2 industry-specific factors, 3 macro-specific factors (inflation, interest rate, and GDP growth), and 3 dummies. At the same time I use 12 year dummies for 13 year analysis.

Can I use only time dummies instead of macro-specific dummies or opposite? I cannot find any empirical evidence about this usage.

$\endgroup$
0
$\begingroup$

Dummy variables are artificial variables created to represent attributes, usually with two or more distinct categories (in your case 13). At first, if your model is correctly specified the firm, industry, and macro factors will capture all the relevant variation. The macro effect will be controlled by the macro factors and your dummies will be not significant (redundant). But what if your macro factors are incomplete (wrong model specification), some dummies will show significance. Then, in your context is a matter of trusting in your model specification to consider or not the dummies. I will choose dummies for years where there is information that a macro effect is not being capture by the other variables.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.