An investigation on the x-efficiency of commercial banks in Kenya

Abstract

This paper seeks to determine the X-efficiency of commercial banks in Kenya and to establish whether the X-efficiency of these banks is affected by economies of scale. The data set consists of annual operation costs of banks including interest expense. Deposits and borrowed funds are the inputs, and the loans to customers and investment and other incomes are the outputs. The data was collected from 33 banks for the period 2000 to 2005. To measure the X-efficiency level of commercial banks in Kenya, we used the Stochastic Econometric Cost Frontier approach which involves the estimation of the cost function and the derivation of the X-efficiency estimate based on the deviation from the efficient cost frontier. The empirical results obtained showed that X-efficiency exists in the commercial banks in Kenya and that X-efficiency of the banks is affected by economies of scale. We found out thatthe level of X-efficiency in Kenya's commercial banks industry is 18%. After controlling for scale differences, the average small bank is found to be relatively less efficient than the average large bank. The persistency of Xefficiency in relation to bank size was measured to find out if inefficient banks tend to remain inefficient over time. We found out that the average large bank was more persistent than the average small bank at the level of23%. We also found out that bank size affects X-efficiency for large banks. These findings were consistent with the results found in other related studies in US (Kwan and Eisenbeis, 1996), Hong Kong (Kwan, 2001) and Namibia (Ikhide, 2000).