3.2 Test Models
3.2.1 Data Envelopment Analysis
This study employs two types of the DEA approaches. First, it applies the standard DEA approach to measure allocative and technical efficiency. Second, the Malmquist factor productivity measure identifies efficiency gain/loss. The model utilizes data as inputs and output quantities of a group of banks, the decision making units (DMUs), to construct a piece-wise frontier over the data points. This frontier is constructed by the solutions from a sequence of linear programming problems, one for each bank over the time periods.
Efficiency scores or measures are then estimated relative to this frontier, which corresponds
to an efficient technology used by the firm. In addition, it allows efficiency to be estimated
without having to stipulate either the structure of production function or the weights for the
inputs and outputs used. Charnes et al. (1978) introduced the DEA constant return to scale
(CRS) method.