Hello everyone,
I want to study how firms' investment response to its cash flows across two different groups of firms, firms with financial constraints and firms without financial constraints.
So I divide my sample into two sub-samples based on a financial constraint measure. The two sub-samples has the same number of years, but for each year, one consists of firms with financial constraints and another consists of firms without financial constraints.
I then estimated two panel-data regressions for the two sub-samples with firm fixed effect.
Investment_i,t= a_i + b*cashflow_i,t + e_i,t where i indexes firms and t indexes year and a_i is firm fixed effect.
I want to compare the coefficient of investment, b, across the two subsamples. This is currently what I am doing is that:
Since the single fixed effect regression is just like OLS with firm dummy
regress investment cashflow i.firm_id if constraint=1
est store eqn1
regress investment cashflow i.firm_id if constraint=0
est store eqn0
suest constraint0 constraint1,cluster(firm_id)
test [eqn1_mean]cashflow=[eqn0_mean]cashflow
I am wondering whether it is right? I see there are lots of discussions on SUR with panel data fixed effect regressions but I am still not very clear how to do that. Hopefully someone can help me!! Thanks very much!!
I want to study how firms' investment response to its cash flows across two different groups of firms, firms with financial constraints and firms without financial constraints.
So I divide my sample into two sub-samples based on a financial constraint measure. The two sub-samples has the same number of years, but for each year, one consists of firms with financial constraints and another consists of firms without financial constraints.
I then estimated two panel-data regressions for the two sub-samples with firm fixed effect.
Investment_i,t= a_i + b*cashflow_i,t + e_i,t where i indexes firms and t indexes year and a_i is firm fixed effect.
I want to compare the coefficient of investment, b, across the two subsamples. This is currently what I am doing is that:
Since the single fixed effect regression is just like OLS with firm dummy
regress investment cashflow i.firm_id if constraint=1
est store eqn1
regress investment cashflow i.firm_id if constraint=0
est store eqn0
suest constraint0 constraint1,cluster(firm_id)
test [eqn1_mean]cashflow=[eqn0_mean]cashflow
I am wondering whether it is right? I see there are lots of discussions on SUR with panel data fixed effect regressions but I am still not very clear how to do that. Hopefully someone can help me!! Thanks very much!!
Comment