Estimating technical and scale efficiency in banks and its relationship with economic value addes: A South African study

Abstract

For estimating the relative efficiency of banks, income statement data as well as balance sheet data are used. Monthly balance sheet data are readily available from the DI900 reports published by the department of Bank Supervision on the web site of the South African Reserve Bank. Monthly income statement data, on the other hand, are not readily available as banks regard it as confidential. Because of the unavailability of monthly income statement data, only monthly balance sheet data were used to estimate the relative efficiency of the four largest banks in South Africa. These four banks handle about 85.17% (March 2005) of the banking business in South Africa.Data Envelopment Analysis (DEA) is applied to estimate the monthly technical and scale efficiency for the four banks over a period of 36 months. DEA is used because it lends itself more easily to the analysis of multiple-output firms. Bank B appears to be the most technically efficient bank. For fifteen months during the sample period the bank could be regarded as fully technically efficient. However this bank operated mainly in the region of increasing return to scale, implying that it was operating at a scale that is too small.Bank C has an average technical efficiency estimate 0.951 (input-orientated), followed by Bank A with an average technical efficiency estimate of 0.917. On seven occasions Bank C can be regarded as being fully technically efficient. Bank C operated mainly in the region of increasing returns to scale. On only three occasions Bank A can be regarded as being fully technically efficient. For twenty one months Bank A has been operating in the region of increasing returns to scale, implying that it was operating at a scale that is too small. Bank D could at no time during the sample period, be regarded as being fully technically efficient. The technical efficiency estimates range from 0.751 to 0.900 with an average value of 0.806 (input-orientated) and from 0.758 to 0.895 with an average value of 0.809 (output-orientated). This bank operated mainly in the region of decreasing return to scale implying that it was operating at a scale that was too large.