Integrated early warning prediction model for Islamic banks: the Malaysian case

It is increasingly becoming important to predict the performance of Islamic banks in order to anticipate a problem before it materializes and negatively affects banks’ performance and financial standing. Benefiting from the earlier research on the subject, this study aims to develop a preliminary in...

Full description

Saved in:
Bibliographic Details
Published inJournal of banking regulation Vol. 19; no. 2; pp. 118 - 130
Main Authors Othman, Jaizah, Asutay, Mehmet
Format Journal Article
LanguageEnglish
Published London Palgrave Macmillan UK 01.04.2018
Palgrave Macmillan
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:It is increasingly becoming important to predict the performance of Islamic banks in order to anticipate a problem before it materializes and negatively affects banks’ performance and financial standing. Benefiting from the earlier research on the subject, this study aims to develop a preliminary integrated early warning model for Islamic banks in Malaysia to assess their financial standing by using quarterly data for the 2005–2010 period. Factor analysis and three parametric models (discriminant analysis, logit analysis, and probit analysis) are used in this study. Out of 29 variables used in the early stage of study, only 13 were selected as predictor variables in this study. Results show that, overall, classification accuracy is relatively high in the first few quarters before the benchmark quarter (2010 Q3) for all the estimated models. Correct classification rates are high during the first few quarters and decrease subsequently. Based on these results, therefore, it is obvious that the first few quarters before the benchmark quarter are the most important for making a correct prediction. These results show the predictive ability of the integrated model to differentiate healthy and non-healthy Islamic banks, thus reducing the expected cost of bank failure.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1745-6452
1750-2071
DOI:10.1057/s41261-017-0040-5