Dear community, I have a data set of almost 12 thousand observations of CEOs and their firm's performance when they were CEO. Most CEOs have data from several years. I've added for each CEO whether they migrated (0=no 1=yes) as I am interested in whether CEOs that migrated have a better effect on shareholder value over time than CEOs that did not.
Since it is panel data I've performed [xtset execid year], however when I try to perform [xtreg shareholder_value migration age_exec employees, fe] to see the effect migration has on shareholder value, I get the prompt that migration is omitted due to collinearity. Eventhough when I look at the correlation matrix, migration's highest correlation with another variable is 0.0224. Some sources say that the collinearity comes from the fact that migration is a time invariant variable as it stays the same over time. However, other sources say that this should not itself be a problem. Therefore, my question is how I can fix this problem of collinearity or is there another way to investigate whether the migration of CEOs has a positive or negative influence on the shareholder value over time.
Thank you in advance!
Since it is panel data I've performed [xtset execid year], however when I try to perform [xtreg shareholder_value migration age_exec employees, fe] to see the effect migration has on shareholder value, I get the prompt that migration is omitted due to collinearity. Eventhough when I look at the correlation matrix, migration's highest correlation with another variable is 0.0224. Some sources say that the collinearity comes from the fact that migration is a time invariant variable as it stays the same over time. However, other sources say that this should not itself be a problem. Therefore, my question is how I can fix this problem of collinearity or is there another way to investigate whether the migration of CEOs has a positive or negative influence on the shareholder value over time.
Thank you in advance!
Comment