Purpose

This study aims to test whether loan loss provisions in Islamic banks is procyclical by explicitly examining the link between non-discretionary provisions and loan growth. In the next stage, this paper tests whether the link between non-discretionary provisions and loan growth is conditional on bank capitalization and lending. This is to identify whether bank-specific factors affect the procyclicality of non-discretionary provisions and whether such procyclicality can be explained by income smoothing in banks with different capitalization and loan profiles.

Design/methodology/approach

This study is conducted in four stages. The first stage identifies the determinants of loan loss provisions. The second stage investigates whether income smoothing is affected by capitalization and lending activities. In the third stage, the link between non-discretionary provisions and loan growth is examined. In the fourth stage, this paper tests whether the link between non-discretionary provisions and loan growth is affected by bank capitalization and lending. A two-way panel-fixed effect model is used.

Findings

Non-discretionary provisions are procyclical, particularly for banks with lower capitalization and lending activities, because such banks do not conduct income smoothing. Specifically, banks with lower capitalization experience a decline in loan growth when non-discretionary provisions to cover credit risk increase.

Research limitations/implications

The dataset used in this study follows Soedarmono et al. (2017) and does not enable to differentiate types of financing products in Islamic banks that may exacerbate or mitigate the procyclicality of non-discretionary provisions.

Originality/value

This paper extends prior literature on the procyclicality of loan loss provisions by specifically investigating the influence of non-discretionary provisions on loan growth in Islamic banks and whether such relationship depends on the role of income smoothing undertaken by banks with different levels of capitalization and lending. This paper builds on the work of Soedarmono et al. (2017) in which they do not explicitly examine the relationship between loan loss provisions and loan growth.