The impact of financial technology on loan risk and payment systems: evidence from Africa

Obuobie, James (2023) The impact of financial technology on loan risk and payment systems: evidence from Africa. Doctoral thesis, University of East Anglia.

Preview

PDF Download (6MB) | Preview

Abstract

Fintech can disrupt the delivery of traditional financial services but can equally improve banking for the poor and has the potential to provide new solutions to old problems. Further, signalling theory provides a firm ground to understand how signals can be used by lenders to screen for quality borrowers, reduce uncertainty and facilitate lending. Particularly to informationally opaque borrowers in developing countries. My thesis proposes a model to screen loan applicants based on clients’ adoption of financial technology. Using machine learning algorithms, my results show that, adopters of Fintech are associated with lower default likelihood. I find asymmetric relationship between new and repeat borrowers who adopt Fintech and loan spread. However, clients with bank account only are more likely to default compared to those who adopt Fintech and have bank account. This suggest that Fintech can unlock opportunities for adopters to improve their credit score by linking their Fintech account to their respective bank accounts.

Further, over the past decade, payments using Fintech has become critical to the financial systems in most countries in Sub-Saharan Africa. In the third chapter of my thesis, I empirically investigate the impact of the use of financial technology on the payments systems in Africa, financial inclusion and signiorage. Using the VAR-VEC model, my results show that Fintech provides financial inclusion. A variance decomposition analysis show that majority of the forecast error variance is due to own shock. Further, my results show a long-run causal relation between mobile money and payment system transaction, and the use of currency. This confirm that Fintech, can transfer informal cash to the formal banking system and lead to a significant reduction in the social burden of signiorage.

The final chapter of my thesis investigate the impact of borrowers’ self-declared religiosity and religious connectedness on loan risk. Using a credit scoring algorithm, I find a significant reduction in default probabilities when borrowers signal as trustable to lenders via their voluntary self-declared religiosity. Further, I find that, all things being equal, borrowers who voluntarily self-declare their religiosity to signal their credit risk are likely to be charged higher interest rate. However, those who self-declare their religious connectedness are associated with the likelihood of receiving lower interest rate, and female borrowers who signal their credit by self-declaring their religiosity are associated with lower default likelihood.

Item Type:	Thesis (Doctoral)
Faculty \ School:	Faculty of Social Sciences > Norwich Business School
Depositing User:	James Tweddle
Date Deposited:	11 Jul 2024 10:22
Last Modified:	11 Jul 2024 10:22
URI:	https://ueaeprints.uea.ac.uk/id/eprint/95869
DOI:

Downloads

Actions (login required)

View Item