Page 321 - Contributed Paper Session (CPS) - Volume 4

CPS2245 Azrie Tamjis
an asymmetric half-normal distribution. ln is the random error term that
permits the random variation of the frontier across banks and captures the
effects of measurement error, other statistical noise and random shocks
outside the bank’s control. This error term is assumed to consist of
independently and identically distributed normal random variables, with zero
mean and variance (Coelli et al., 2005).
The cost-efficiency of the i-th bank is the estimated cost needed to
produce bank i’s output vector if the bank were as efficient as the best-practice

bank (on the frontier curve) in the sample facing the same inputs and outputs,
control, and environmental variables(, , , ), divided by the actual cost of i-
th bank, and adjusted for random error. It can be written as:
exp((, , , , ) + 1 (5)

where is the cost-efficiency of i-th bank. The numerator in equation 5
indicates the minimum cost that can be incurred by the best practice banks
and the denominator in equation 5 denotes the actual cost incurred by i-th
bank at time t. Hence, cost-efficiency is measured against the ratio of
minimum cost banks (best-practice banks on the frontier) and the actual cost
of i-th bank. Cost-efficiency could also be seen as a proportion of cost
that is either being used efficiently or being wasted. For example, if of i-
th bank is 0.60, it indicates that i-th bank is 60.0% efficient and 40.0% of its
cost is being wasted when compared to the best-practice bank. Costefficiency
ranges between 0 and 1. Banks with a cost-efficiency of 1 are considered to
be best-practice banks within the observed data.

3. Result
The cost efficiency scores for the Malaysian banks for the years 2000–2011 are
shown in Table 1. The average cost efficiency score for Malaysian banks
between 2000 and 2011 is 82.7%. The average score of cost efficiency was at
85.1% in 2000 and ended with 75.0% in 2011. The average cost efficiency
scores suggest that Malaysian banks wasted around 20.0% of their input to
produce the same level of outputs of the best performing banks. This finding
(approximately 20% inefficiency) is consistent with past findings in the
literature, where SFA was performed for the cost efficiency function (e.g De
Young, 1997; Berger and De Young, 2001; Bonin et al., 2005).

310 | I S I W S C 2 0 1 9