Predicting Efficiency in Angolan Banks: A Two‐Stage TOPSIS and Neural Networks Approach

• Author: Peter Wanke,Carlos Barros,Nkanga Pedro João Macanda
• Published date: 01 September 2016
• DOI: http://doi.org/10.1111/saje.12103
• Date: 01 September 2016

PREDICTING EFFICIENCY IN ANGOLAN BANKS:

A TWO-STAGE TOPSIS AND NEURAL

NETWORKS APPROACH

PETER WANKE

*,

CARLOS BARROS

‡

AND NKANGA PEDRO JOÃO MACANDA

†

Abstract

This paper presents an efficiency assessment of the Angolan banks using Technique for Order

Preference by Similarity to the Ideal Solution (TOPSIS). TOPSIS is a multi-criteria decision-

making technique similar to data envelopment analysis, which ranks a finite set of units based on the

minimisation of distance from an ideal point and the maximisation of distance from an anti-ideal

point. In this research, TOPSIS is used first in a two-stage approach to assess the relative efficiency

of Angolan banks using the most frequent indicators adopted by the literature. Then, in the second

stage, neural networks are combined with TOPSIS results as part of an attempt to produce a model

for banking performance with effective predictive ability. The results reveal that variables related to

cost structure have a prominent negative impact on efficiency. Findings also indicate that the

Angolan banking market would benefit from higher level of competition between institutions.

JEL Classification: D24, D2, G21, G2

Keywords: Banks, Angola, TOPSIS, two-stage, neural networks, efficiency ranks

1. INTRODUCTION

One of the major research streams in banking is to measure the relative importance of

banks using popular multi-criteria decision making (MCDM), such as the data

envelopment analysis (DEA) and the Technique for Order Preference by Similarity to the

Ideal Solution (TOPSIS) (Hemmati et al., 2013). This paper analyses the efficiency of

Angolan banks with TOPSIS. Thus far, applications of TOPSIS to measure bank

efficiency have been scarce (Seçme et al., 2009; Shaverdi et al., 2011; Hemmati et al.,

2013; Bilbao-Terol et al., 2014), although efficiency per se has been the focus of much

recent research (Briec and Lemaire, 1999; Camanho and Dyson, 2005; Boussemart et al.,

2009; Briec and Liang, 2011), especially in relation to banking (Berger and Humphrey,

1992, 1997; De Borger et al., 1998; DeYoung, 1998; Camanho and Dyson, 1999; Ariff

and Can, 2009; Drake et al., 2009; Sahoo and Tone, 2009; Fukuyama and Weber 2009a,

2009b, 2010; Ray and Das, 2010; Tone and Tsutsui, 2010; Epure et al., 2011; Wu et al.,

2014). However, efficiency in African banking has been addressed in few studies (see, e.g.,

O’Donnell and Vander Westhuizen, 2002; Ikhide, 2008; Barros, et al., 2014), and a deep

* Corresponding author: Associate professor, COPPEAD Graduate Business School, Center for

Studies in Logistics, Infrastructure and Management, Rua Paschoal Lemme 355, Rio de Janeiro,

Brazil 20520120. E-mail: peter@coppead.ufrj.br.

†Faculdade de Economia, Universidade Agostinho Neto, Luanda, Angola.

‡University of Lisbon, Lisbon, Portugal.

Please acknowledge this Research made with support of Calouste Gulbenkian Foundation.

South African Journal of Economics Vol. ••:•• •• 2015

© 2015 Economic Society of South Africa. doi: 10.1111/saje.12103

1

V

C2015 Economic Society of South Africa. doi: 10.1111/saje.12103

461

South African Journal of Economics Vol. 84:3 September 2016

South African Journal

of Economics

analysis of Angolan banks has not yet been undertaken in terms of the assessment of its

efficiency drivers, which justifies this research.

Therefore, this paper innovates in this context first by focusing on Angolan banks and

second by adopting TOPSIScombined with neural networks in a two-stage approach. The

motivations for the present research are the following. First is to evaluate the relative

efficiency among Angolan banks. Efficiency is the relative position of the units analysed in

the frontier of best practices, which is defined by the target group of banks. In this research,

a TOPSIS analysis of Angolan banks is undertaken for the first time. Second, the paper

expands the existing literature by virtue of its use of neural networks to predict and interpret

the role of major contextual variables in achieving higher levels of efficiency in Angolan

banks. Typically, contextual variables, such as financial ratios related to cost structure or

other business characteristics, are used in fuzzy analytical hierarchy process (AHP)

techniques for determining the criteria weights (Seçme et al., 2009; Shaverdi et al., 2011).

In this research, however,these contextual variables are used in neural networks in order to

build an efficiency model with predictive ability. Third, our analysis covers the period from

2006 to 2012. Finally, our analysis is based on a representative sample of Angolan banks.

The purpose of this study is to propose a predictive model for banking efficiency in

Angola based on the financial and operational criteria commonly found in the literature.

Angola attracts banks every year, and there is almost evidence that the market is saturated

because profits have started decreasing and mergers have already started (Russian bank

VTB África and Angola Banco Privado Atlântico merged in 2014). Therefore, in this

context, efficiency calls the attention of the banks for the relative competition in which

they have been engaged.

In order to achieve this objective, neural networks are presented in a two-stage

approach; TOPSIS analysis is then carried out because it aims to render prediction of

performance more flexible and informative than traditional statistical methods (Brockett

et al., 1997). The remainder of the paper is organised as follows: Section 2 presents the

contextual setting, then Section 3 covers the literature review. Section 4 presents the data

and the model. The empirical results are presented and discussed in terms of policy

implications in Section 5. Conclusions follow in Section 6.

2. CONTEXTUAL SETTING

Banking activity in Angola started in 1865 when the National Ultramarine colonial bank

was established to fund colonial activities. Until 1926 this bank served as Angola’s central

bank, but due to the uncontrolled growth of money, the Bank of Angola (BA) was

established. In 1957, other banks established themselves in Angola: Banco Comercial de

Angola, Banco de Crédito Comercial e Industrial, Banco Totta Standard de Angola,

Banco Pinto e Sotto Mayor and Banco Inter-Unido. After gaining independence from the

colonial power, Portugal, in 1975, the Angolan government nationalised the BA and

Banco Comercial de Angola. The government also established Banco Nacional de Angola

(BNA), which today serves as the central bank, as well as Banco Popular de Angola (BPA).

In 1978, the banking system was reduced to two public banks: BNA and BPA. However,

in 1987, with the abandonment of the socialist organisation, the Angolan government

allowed new banks to enter the country and forced BNA to end its commercial activity,

allocating it to a new bank, Caixa de Credito Agro-Pecuário, which was shut down in

2000. In 2005, Angola’s banking system included three public banks (BNA, which is the

South African Journal of Economics Vol. ••:•• •• 20152

© 2015 Economic Society of South Africa.

462 South African Journal of Economics Vol. 84:3 September 2016

V

C2015 Economic Society of South Africa.