Bank Efficiency and Its Determinants in a Small Island Developing Economy: Evidence from Fiji Using Data Envelopment Analysis and a Two-Limit Tobit Approach

This study examines the efficiency of Fiji’s banking sector from 2000 to 2025 using a two-stage analytical framework. First, data envelopment analysis (DEA) under the variable returns to scale (VRS) assumption is employed to estimate the efficiency scores of seven financial institutions: five commercial banks—four foreign-owned and one locally owned—and two locally owned non-bank financial institutions. Second, a two-limit Tobit model is used to investigate whether the estimated efficiency scores are associated with credit risk, return on assets (ROA), return on equity (ROE), bank size, foreign ownership, loan-loss provisions relative to net income, and real GDP growth. The balanced panel comprises 182 institution-year observations over 26 years. The DEA results indicate a high mean efficiency score of 0.923, although meaningful variation is observed across institutions and over time. Foreign-owned banks record a marginally higher mean efficiency score than locally owned institutions (0.924 compared with 0.921); however, foreign ownership is not statistically significant in the multivariate Tobit model. ROA has a positive and statistically significant association with efficiency, whereas ROE has a negative and statistically significant association, suggesting that asset profitability and equity profitability capture distinct balance-sheet and capital-structure channels. Bank size is positively associated with efficiency, while credit risk, loan-loss provisioning, and real GDP growth are statistically insignificant. Overall, the findings suggest that Fiji’s financial institutions operate relatively close to the estimated best-practice frontier. Nevertheless, the small number of institutions and the resulting dense DEA frontier warrant cautious interpretation. The study concludes that policies aimed at strengthening institutional efficiency should prioritise cost discipline, productive asset utilisation, appropriate capital management, and technology diffusion rather than ownership status alone.