Country Risk, Corporate Strategy and Profitability: A System GMM Estimation for Multinational Banks in Sub Saharan Africa

Abstract

Over the past 2 decades, foreign investment flows in Africa have steadily increased due to; high growth in GDP, economic liberalization, improved business climate (political and socioeconomic), the fast growth and continuous shifts in customer needs, improved general infrastructure development, technological advancement and the formation of free trade zones. However, political instability, government interference through deregulation, corruption, sharp practices, stiff competition, inadequate demand, cultural barriers and failing public infrastructures are factors to worry about. The dynamic environment in the region provides both opportunities and threats for businesses. This research examines the effect of country risk as a whole, and corporate strategy (diversification, debt and equity financing) on performance of multinational banks within sub Saharan Africa. To accomplish this, a panel dataset (2007-2017) on best performing multinational banks in the region is used. Accounting for the dynamic nature and persistent country risk, corporate strategy and bank profitability, a two-step system GMM estimator is used in the analysis. Results indicate country risk negatively and significantly affecting multinational bank performance. Corporate strategies: diversification and equity financing indicate a positive and significant effect on multinational banks’ profitability in terms of return on assets. The effect however, is higher for diversification than equity financing. Debt financing indicates a negative and significant effect on profitability of these banks. Also, the study finds no performance difference between the local indigenous, and overseas multinational banks. The resultant Arellano-Bond AR(1&2) test for zero autocorrelation, the Hansen test of overidentifying restrictions and the difference-in-Hansen test for the validity of the additional moment restriction necessary for system GMM; all, suggest that the underlying assumptions under the two-step system GMM estimator, are not violated since their p-values are very high. The implication of these tests is that the results obtained are robust, efficient and reliable.