Predicting Loan Defaults in Development Financing in South Africa

Sibiya, Petros

Predicting Loan Defaults in Development Financing in South Africa

Thesis / Dissertation

2023

Abstract

Non-Performing Loans (NPLs) remain a pertinent issue faced by financial institutions, including Development Finance Institutions (DFIs). The escalation of this problem has a detrimental impact on the DFI's profitability and sustainability in the long-term, including hampering its crucial role in addressing the failures of the market mechanism to allocate financial resources to the development agendas of developing countries. This study examines the client and loan factors linked to the predictability of loan defaults in a DFI loan portfolio, using the secondary loan portfolio data of a major DFI from 2016 to 2021. The logistic estimation technique was employed to examine the effect of variables including firm industry, firm size, firm development stage, deal complexity, credit scoring, and type of financial instrument on default predictability of DFI loans. The empirical findings of the study show that the size of the firm and the industry in which it operates, are client factors linked to the predictability of loan defaults in a DFI loan portfolio. The type of financial instrument, complexity of the deal, and the credit scoring represented characteristic loan factors that were investigated in this study. All three variables were found to be linked to loan default predictability. Some of the recommendations put forward for the consideration of management includes close monitoring of all clients in the loan book, offering business support to SMEs, and thorough due diligence process amongst others.

Keywords

Development Finance

Reference:

Collections

Masters

Full item page