Page 92 - CITN 2017 Journal

sample of fifty (50) companies out of non-financial quoted companies that covers 10
sectors were purposively selected on stratified random sampling basis.

The rationale for the exclusion of financial related quoted companies like banks, insurance
companies and Mortgage Fund Companies was due to the peculiarity in their operations
which are substantially different from non-financial quoted companies. The data used in
the analysis were collected from the audited financial statement of the selected non-
financial quoted companies in Nigeria and Nigeria Stock Exchange Fact books. The data
collected cover 2004 to 2014 financial years of the selected companies.

ETR = (Total tax expense / Pre-tax income) *100

Variables Code Measurement
Effective Tax Rate ETR ETR=(Total Tax Expenses/Pre Tax Income) *100

Firm Value TOBINQ Total market value/Total Asset Value of firm
Measured by Cash flow to net assets ratio = Pre-tax
Liquidity LIQ profits + Depreciation/(Total assets – cash and
equivalents)
Leverage LEV Long-term debt/total asset
Profitability ROA Operating profits/Total assets
Capital intensity CIN tangible assets to total assets
Net Working Capital NWC Net current assets/Total assets
Size SIZE Natural logarithm of Total Asset =Ln(Total Asset)
Total Assets in year (t)/Total assets in year (t-1)
Growth Opportunities MTB

In panel estimation, neither the Generalized Least Squares (GLS) estimator nor Fixed
Effect (FE) estimator produces consistent estimates in the presence of dynamics and
endogenous regressors. Since TOBINQ* , in Equation 1 was lagged endogenous
i t
regressors as well as unobserved firm fixed effects which are correlated with the regressor,
hence the orthogonality condition is not likely to be met for a GLS or FE estimator to
produce consistent estimates. This explains the use of GMM approach.
According to Arellano and Bond's (1991), the basic GMM panel estimators are based on
moments of the form,

85