Memprediksi gejolak perbankan di Indonesia dengan pendekatan Markov Switching VAR

Abstract

This research aims to detect early indicators of conventional banking and Islamic banking crises, identify the longest period of crisis between the two banks and compare the stability between the two. The method used is the Markov Switching Vector Autoregressive (MS-VAR) approach, where the advantage of this approach is the threshold value of the crisis index (threshold) is an endogenous variable in other words the crisis period and the duration of the crisis is part of the estimation results. Leading micro indicators for conventional banking are cash ratio and leading macro indicators for conventional banking are interest rates. While the leading micro indicators for sharia banking are bank deposits and cash ratios as well as leading macro indicators for sharia banking, namely interest rates, inflation, domestic credit, money supply and current account/gdp. The Z-score of conventional banking is higher (10.98) than the Z-score of Islamic banking (9.93) meaning that in general conventional banking is more stable than Islamic banking. But the longest distress period experienced by conventional banking is around October 2014–January 2016 while the period of Islamic banking distress around January 2008 December 2008