Premium
The performance evaluation of banks considering risk: an application of undesirable relation network DEA
Author(s) -
Chen YuChuan,
Chiu YungHo,
Chiu ChingJen
Publication year - 2020
Publication title -
international transactions in operational research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.032
H-Index - 52
eISSN - 1475-3995
pISSN - 0969-6016
DOI - 10.1111/itor.12446
Subject(s) - financial institution , data envelopment analysis , investment (military) , financial risk , production (economics) , business , order (exchange) , dual (grammatical number) , relation (database) , risk management , capital requirement , commodity , service (business) , financial services , finance , duality (order theory) , economics , actuarial science , computer science , microeconomics , marketing , art , mathematical optimization , mathematics , literature , database , incentive , politics , political science , law , discrete mathematics
The probability of losses for financial institutions fluctuates in accordance with commodity prices and risk managers’ investment tolerance. For the financial service sector, when the capital base increases through lending and investment instruments, the associated financial risks often grow proportionally higher. In other words, the inherent risk of a loss is the financial institution's principal trade. However, with risk comes reward, and therefore banks often engage in complex financial risk management in order to capitalize on their profits. To assess a bank's performance, traditional data envelopment analysis (DEA) incorporates various inputs and outputs and runs through preset formulae to arrive at conclusions. As a result, traditional DEA models often overlook the duality of risks. This paper discusses network DEA by considering the inputs and outputs of a bank's surrounding production processes as additional undesirable factors and integrates the dual nature of risks. We employ and analyze Taiwanese banks’ data to assess the local financial service sector's performance efficiency. The results provide a basis for future recommendations.