The effect of bank m&As on efficiency: the portuguese experience victor Mendes
Download 0.67 Mb.
|
| 4. Data and results
Research in banking has followed two alternative approaches, intermediation and production. From an empirical standpoint one does not seem to have a clear edge over the other. On the other hand, there does not exist consensus on the variables that best define bank output. In this paper we use the intermediation approach, summarized in the following definition of the variables used: Outputs:
y1= loans to clients, net of provisions; y2 = loans to credit institutions + bonds (net of provisions); y3 = off-balance products (proxied by commissions received + net profits from financial operations).
x1 = deposits (from clients + credit institutions + bonds); x2 = number of employees; x3 = fixed assets. Input prices: w1 = price of x1, defined as (interest paid and similar costs + commissions paid)/x1; w2 = price of x2, defined as labor costs/number of employees; w3 = price of x3, defined as (depreciation + other administrative costs + other operating costs)/net fixed assets. C = variable cost, defined as the sum of financial costs, labor costs and operating costs. We use data from banks’ annual balance sheets and income statements, for the years 1990 to 1997. The information was collected from the Boletim Informativo da Associação Portuguesa de Bancos and banks’ annual reports, and we use non-consolidated data. The sample includes almost all banks operating in Portugal in that period4. The sample does not include banks in their first (incomplete) year of activity. Table 1: Summary information on the input/output variables.
* Deflated values, at 1990 prices. Summary information on the input/output variables is in table 1. There are no ‘strange’ surprises; the sample seems to combine well different-sized institutions (as shown by the Coefficient of Variation – CV). On the other hand, loans to financial institutions represent the largest output share and price variability is lower than input variability, suggesting strong competition in the input markets. Inefficiency indices estimates are presented in annexes 2, 3 and 4, respectively for DEA, the SFA-T and the SFA-E methods. Summary information of our results is in table 2. Download 0.67 Mb. Share with your friends:
|