Panel: interaction dummy
Posted: Thu Oct 29, 2015 2:33 am
I would like to estimate a model with an interaction dummy between one variable and the country dummy.
Here, I generate the interaction term with the intercept
for %c ES IT PT FR NL BE IE AT FI DE US UK
smpl @all
genr stock_{%c} = 0
smpl 2014:08 2015:1
series stock_{%c} = 1
next
But then I cannot run the following
pool yieldd3a_{%m}y es it pt fr nl be ie at fi de us uk
yieldd3a.ls y_? c x_? stock_ES stock_US
I see that there is problem of "Near singular matrix", because stock_ES and stock_US are the same variable.
Therefore, how can I introduce in a panel setting an interaction dummy between one variable and the country dummy?
Needless to say that the following specification works, but I do not achieve my objective of getting country-specific estimates on the stock variable: yieldd3a.ls y_? c x_? stock_?
Here, I generate the interaction term with the intercept
for %c ES IT PT FR NL BE IE AT FI DE US UK
smpl @all
genr stock_{%c} = 0
smpl 2014:08 2015:1
series stock_{%c} = 1
next
But then I cannot run the following
pool yieldd3a_{%m}y es it pt fr nl be ie at fi de us uk
yieldd3a.ls y_? c x_? stock_ES stock_US
I see that there is problem of "Near singular matrix", because stock_ES and stock_US are the same variable.
Therefore, how can I introduce in a panel setting an interaction dummy between one variable and the country dummy?
Needless to say that the following specification works, but I do not achieve my objective of getting country-specific estimates on the stock variable: yieldd3a.ls y_? c x_? stock_?