The determinants of bank profitability in Zambia

Thumbnail Image
Momba, Naminda
Journal Title
Journal ISSN
Volume Title
The University of Zambia
Commercial banks perform an important part in financial intermediation. Zambia, like any other nation, needs a stable banking system in order to finance private and public investments. The fruitfulness of the banking system in channelling funds from depositors to lenders is often gauged by examining bank profitability. Although the Zambian banking system continues to grow, the level of profitability remains low. This paper examines the determinants of the profitability of banks in Zambia during the period January 2010 to December 2016. The empirical analysis uses bank specific variables and employs a fixed effect model after carrying out a Hausman test. The study employs panel data of 17 commercial banks obtained from the Bank of Zambia. The measures of profitability used in the study are return on assets (ROA), return on equity (ROE) and net interest margin (NIM). In estimating the models, these measures of profitability were regressed on bank size as measured by the log of asset and on credit and liquidity risk measures and bank efficiency. Empirical findings show that bank size, the ratio of loan loss provision to total assets and total loans to assets significantly affect bank profitability regardless of the profitability measure employed. Results indicate that banks pay more to depositors than they receive from loans and that most of their profit is relatively derived from operational income rather than interest based income. Bank efficiency is observed to have a positive significant impact on NIM only. In light of these finding, it is recommended that banks develop a policy that limits the amount of loans they extend without collateral. This will allow commercial banks to reduce and mitigate the high risk of default observed in this study. Keywords: Bank profitability, Panel data, Bank specific variables, Zambia.
Bank profitability. , Banks and banking--Econometric models.