Munich Personal RePEc Archive

Soft Data Modeling via Type 2 Fuzzy Distributions for Corporate Credit Risk Assessment in Commercial Banking

Brkic, Sabina and Hodzic, Migdat and Dzanic, Enis (2018): Soft Data Modeling via Type 2 Fuzzy Distributions for Corporate Credit Risk Assessment in Commercial Banking. Forthcoming in:

[img]
Preview
PDF
MPRA_paper_87652.pdf

Download (374kB) | Preview

Abstract

The work reported in this paper aims to present possibility distribution model of soft data used for corporate client credit risk assessment in commercial banking by applying Type 2 fuzzy membership functions (distributions) for the purpose of developing a new expert decision-making fuzzy model for evaluating credit risk of corporate clients in a bank. The paper is an extension of previous research conducted on the same subject which was based on Type 1 fuzzy distributions. Our aim in this paper is to address inherent limitations of Type 1 fuzzy dis-tributions so that broader range of banking data uncertainties can be handled and combined with the corresponding hard data, which all affect banking credit deci-sion making process. Banking experts were interviewed about the types of soft variables used for credit risk assessment of corporate clients, as well as for providing the inputs for generating Type 2 fuzzy logic membership functions of these soft variables. Similar to our analysis with Type 1 fuzzy distributions, all identified soft variables can be grouped into a number of segments, which may depend on the specific bank case. In this paper we looked into the following segments: (i) stability, (ii) capability and (iii) readiness/willingness of the bank client to repay a loan. The results of this work represent a new approach for soft data modeling and usage with an aim of being incorporated into a new and superior soft-hard data fusion model for client credit risk assessment.

UB_LMU-Logo
MPRA is a RePEc service hosted by
the Munich University Library in Germany.