Publication: A generalised regression neural network model of financing imbalance: Shari'ah compliance as the roadmap for sustainability of capital markets
No Thumbnail Available
Date
2020
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
IOS Press BV
Abstract
The current study looks at the impact of compliance to Shari'ah principles on the capital structure for Malaysian firms. Examination of impact of compliance is based on the classification by the Securities Commission of Malaysia. Given that the literature on adjustment tends to ignore non-linear models, the current study utilises Generalised Regression Neural Network (GRNNs). Results are compared to conventional panel data regression models via performing a hold-out sample. Initial results confirm stability of the data allowing predictive ability. The results indicate that compliant firms tend to finance a greater portion of their financing imbalance via equities relative to non-compliant firms. This provides a strong indication towards compliant firms reducing overall risk taking where the financing pattern incorporates a greater aspect of risk sharing which is in-line with Shari'ah principles. In addition, two more factors are ranked as important in deciding compliant firms issue choice to resolve financial imbalance: profitability and size. The rest of the determinants have low impact on explaining net debt issues. Diagnostics for results provide evidence of lower RMSE and MSE for GRNNs for the training, testing and overall datasets. The potential benefit of this research allows managers and investors of Islamic capital markets to understand potential risk exposure and financing costs of compliant firms. Findings also provide a roadmap for development of a sustainable capital market model which has wider implications on a global scale. � 2020-IOS Press and the authors.
Description
Keywords
Capital structure, generalised regression neural networks, Islamic capital markets, Islamic finance, sustainable capital markets
Citation
Journal of Intelligent & Fuzzy Systems, vol. 39, no. 4, pp. 5387-5395, 2020