Journal of Advanced Research in Business and Management Studies 20, Issue 1 (2020) 1-13 1 Journal of Advanced Research in Business and Management Studies Journal homepage: www.akademiabaru.com/arbms.html ISSN: 2462-1935 Identifying Relevant Predictor Variables for a Credit Scoring Model using Compromised-Analytic Hierarchy Process (Compromised-AHP) Yosi Lizar Eddy 1 , Engku Muhammad Nazri 2, , Nor Idayu Mahat 2 1 Risk Quantification Section, Risk Management Department, Bank Muamalat Malaysia Berhad, 21 Jalan Melaka, 50100 Kuala Lumpur, Malaysia 2 School of Quantitative Sciences, Universiti Utara Malaysia, 06010 UUM Sintok, Kedah, Malaysia ABSTRACT Developing an efficient credit scoring model to reduce the risk of personal-loan defaulters involves the selection of manageable reliable predictor variables in order to avoid the potential clients from providing too much information and to reduce the burden of a bank from keeping huge historical data, which can be burdensome and costly. The objective of this paper is therefore to illustrate how compromised-AHP can be used as one the methods to select such relevant reliable predictor variables before the final credit scoring model is constructed. A case study involving four experts from a bank was conducted. A set of sub-predictor variables under four main predictor variables namely financial indicators, demographic Indicators, employment indicators, and behavioural indicators was rated based on the perception of the four experts. The results reveal that, based on the experts’ perception, the number of payments per year and payment interval, the loan or credit history, total income, total debt, the checking accounts, and age are the six most influential predictor variables while race, gender, and social status are the three least influential predictor variables. Keywords: Credit scoring model; predictor variables; loan defaulters; compromised-AHP Copyright © 2020 PENERBIT AKADEMIA BARU - All rights reserved 1. Introduction The banking industry has developed into one of the comprehensive and competitive markets in contributing to economic development over the past few decades. One of its many businesses is providing personal loans to potential clients. Giving out personal loans is an insecure business but at the same time, it is one of the major sources of income to most banks. Banks would prefer not to allow credit to those customers who lack the capacity to pay back the credit given. Be that as it may, after some time, a certain percentage of the credits will eventually transform into bad loan regardless of the possibility that the banks tighten its credit policy [1-2]. Analysing the non-performing loans data will effectively measure the quality of credit endorsement process. The loan granting process must be observed vigilantly, and banks should formulate an effective credit risk management. After  Corresponding author. E-mail address: enazri@uum.edu.my https://doi.org/10.37934/arbms.20.1.113 Open Access