Modelling Bank Lending in Cameroon Using Two Steps Cointegration Technique

Abstract

This paper has as objective to model the key determinants of bank lending in Cameroon using time series data collected mainly from the World Bank database for the period spanning from 1979 to 2019. After exploring related theoretical and empirical literature, Engle and Granger two steps cointegration testing procedure was employed together with the necessary pretesting (unit root tests) and post estimation testing. Results of stationarity tests indicate that variable were a mixture of stationary and first difference. The lagged residual term possess unit root at level indicating that the variables are actually cointegarted and that ECM is relevant. The estimated cointegration regression model and the corresponding ECM were globally significant at one percent and the models were equally adequate with explanatory power of over 73 and 52 percent respectively. The results of the two models revealed that the effect of economic growth on bank lending was positive and highly significant, the effect of interest rate is negative and equally significant, then the effect of Gross Investment rate was also positive but not very significant in the short run. Based on the major findings, bank lending can be encouraged by stimulating the level of economic activities, by reducing interest rate on loans and by encouraging more investment projects.