Global Journal of Management and Business Research, B: Economics and Commerce, Volume 20 Issue 1

objective is the estimation of the coefficients of variables other than the constant and if they differ a bit, the question of the choice between the two models (fixed effects and random effects) loses its acuity. The random effects model is = + 1 + 2 + 3 + 4 + 5 + +  The Generalized Method of Moment (GMM) model in System GMM in the dynamic panel has several virtues: they solve problems of bias of concurrency, inverse causation, and omitted variables. The GMM estimator is better than the Ordinary Least Squares (OLS) estimator. There are two (2) forms of GMM estimators in dynamic panels: The first difference GMM Estimator and the System GMM Estimator. The Arellano & Bond Model (1991) offers a first-GMM-difference estimator. It consists in taking for each period the first difference of the equation to be estimated to eliminate the country of the specific effects, and to the instrument after that the explanatory variables of the equation in first difference by their values at the level retarded of a period or more. The Blundell & Bond Model (1998) determines a system-GMM estimator that combines the first- difference equations with the level equations in which their primary differences instrument the variables. The GMM estimator in the system appears to be better than the GMM estimator since the latter gives biased results in the case of finite samples when the instruments are weak. The determination of the GMM estimator depends on the validity of the hypothesis that the error terms are not self-correlated and the validity of the instrumental variables used. To ensure the lack of self-correlation of the error terms and the validity of the instruments used, Blundell and Bond (1998) propose two essential tests: The Sargan test which allows to analyze the over- identification of the model and the validity Instruments used for the estimation and common test of lack of self- correlation for error terms, ε it. Basic GMM model is: = + −1 + 1 + 2 + 3 + 4 + 5 + + Where FINANCE is financial development, INTECO is economic institutions, INSTPO is political institutions, INSTFIN is financial institutions, RGDPC is real GDP per capita, the subscripts i and t index countries and time respectively. Also, the specification contains an unobservable country-specific effect and error-term .The data used in this study are mostly from the World Bank. IV. R esults In this part, we will first give the results of our composite financial indicator and then the results of our econometric model with all its tests. a) Composite indicator of financial development To obtain this index, we proceed by applying the Principal Component Analysis method to achieve a weighting that reflects the reality of contributions from different dimensions of financial development. This Principal Component Analysis work focuses on data from institutional and financial variables such as the Voice and accountability, Political Stability and no Violence, Government Effectiveness, regulatory quality, rule of law, Control of Corruption, bank credit to bank deposit, deposit money bank asset to GDP, Domestic credit to private sector, Private credit by deposit money banks and other financial institutions to GDP, Liquid liabilities to GDP and Financial system deposits to GDP. The software used XLSTAT when applying the PCA gives us a table of contribution to the different variables to the construction of the different axes. It is the contributions of the various variables that we use as a weighting in the calculation of our synthetic indicator for the quality of institutions. © 2020 Global Journals 27 Global Journal of Management and Business Research Volume XX Issue I Version I Year 2020 ( ) B Institutional Quality and Financial Development in West Africa Economic and Monetary Union