Intermediation Costs in an Agricultural Development Bank: A Cost‐Function Approach to Measuring Scale Economies

Abstract

The cost-output relationships and production technology in the Agricultural Development Bank of Honduras are analyzed using a translog model. Scale economies for the average branch are not significantly different from one. They are a function of output levels reflecting U-shaped cost surfaces. Product-specific economies of scale are substantially different. Returns to scale oflending activities approach unity, whereas there are important unexploited economies to the expansion of deposit mobilization. Cost complementarities between lending and deposit mobilization indicate advantages of joint provision of financial services over the traditional specialization in lending of agricultural development banks in developing countries.Knowledge of the production technology of financial institutions is essential for analyzing market structure and institutional performance. Many regulatory and managerial decisions are based on specific assumptions about economies of scale and other features of the cost-output relationships in these institutions. Consequently, several recent studies have addressed the measurement of scale economies and cost complementarities in the production of financial services (Benston, Hanweck, and Humphrey; Hunter and Timme; Mullineaux; Murray and White; Panzar and Willig 1977). However, few attempts have been made to analyze-these cost-output relationships in development banks operating in less developed countries.Development banks function primarily as lending institutions intermediating government funds or lines of credit established by foreign agencies and providing limited deposit services, if any, to the rural clientele. The lack of empirical evidence regarding the costoutput relationships in these banks has so far prevented the assessment of their overall perCarlos E. Cuevas is an assistant professor in the Department of Agricultural Economics and Rural Sociology, Ohio State University.The author acknowledges helpful comments and suggestions from Journal reviewers.formance and the evaluation of their operational strategies that emphasize lending over deposit mobilization. Furthermore, considerable regulation exists in less developed countries regarding bank branching, bank size, and loan pricing. These regulations are formulated with imperfect knowledge, if any, of the production structure of financial institutions. In particular, government subsidization of development banks should consider the cost structure of these banks to determine the magnitude of the subsidy and the expected impact of the policy on output expansion. Thus, the knowledge of the production structure of financial institutions becomes essential to assess the likely consequences of financial policy.Studies on development banks in less developed countries by Gheen and by Nyanin have provided limited insights into the cost structure and underlying technology of these institutions because of the choice of very restrictive functional forms for the cost function. In general, the use of Cobb-Douglas or constant elasticity of substit...

Cost structure and technical change in rural banking

“…as the 'partial' economies of scale (PES). This is one way 7 of indicating the returns to scale with respect to a single output holding aIl other outputs constant (Bothwell and Cooley, 1982;Cuevas, 1988;Noulas et al, 1990). There are increasing, decreasing or constant returns to scale with respect to output i as K; is less than, greater th an or equal to 1, respectively.…”

“…(1) Cost. The total non-interest operating expenses nett of depreciation and provisions for bad debt (Cuevas, 1988).…”

“…Three outputs (QI' Q2 and Q.l) arc utiliscd in this study. Following Cuevas (1988), QI and Q2 are used to represent, respeetively, the total amount of deposit balances outstanding as of the 31st Deccmbcr of each year and the total value of loans granted during the year, Historically (i.e. 1976-87), government securities (ail of short-term duration) have averaged 21% of GRB asscts.…”

“…Following the practice in previous studies (c.g. Benston et al, 1982;Hunter and Tirnrne, 1986;Cuevas, 1988), the priee for labour, Pl' was defined as the total remuneration and benefits for the staff (minus the manager) dividcd by thcir total number. The unit priee of capital services.…”

“…The unit priee of capital services. P2' is proxicd by the ratio of "depreciation plus rents paid" to "the total value of loans plus deposit balances" (Cuevas, 1988). P J was used to represent cost of management and was proxied by the total expcnse on the Board of Directors plus the manager's remuneration.…”