Pemodelan Multiple Discriminant Analysis untuk Memprediksi Financial Distress Bank Umum Syariah di Indonesia

Abstract

The aims of this study is to set financial distress prediction model and to identify the best accuraction and classification from the financial distress prediction model. The objects were 9 Islamic Banks in Indonesia since 2012 to 2017 using Multiple Discriminant Analysis modelling. The variables used financial ratios, consist of ROA, BOPO, Current Assets to Current Liabilities Ratio, NPF, Equity to Total Liabilities, and FDR. The outcome shows that a variable tend to cause an Islamic bank fall into financial distress condition dominantly was NPF ratio. The accuration prediction power with 42 from 42 obervations predicted fall into health bank category were classified correctly (100%), and 2 from 12 Islamic Banks fall into financial distress category were classified incorrectly (16.7%) and were corrected into helath bank category. The classification power created by multiple discriminant analysis was 81.48%.