MENDETEKSI KRISIS PERBANKAN DI INDONESIA DENGAN EKSPERIMEN MODEL SIGNAL

Abstract

The objective of this paper is to develop a signal model for detecting banking crisis in Indonesia. The model uses macroeconomic indicators to make composite index for forecasting banking crisis. There are macroeconomic indicators selected for composite index such as GDP growth, ratio of short-term foreign debt to international reserves,ratio of BIS claim to international reserves, ratio of Central Bank credit to public to GDP and ratio of growth of M2 multiplier to international reserves.The model is quite robust to detect banking crisis in sample period, but it relatively less powerful to predict out of sample period. This might be due to the significant change of economic and banking structure between in sample and out of sample period. However, this model can detect the potential risk of banking crisis in 2008 refle by the increase of probability of crisis. Further research is needed to refine the model by including microprudential indicators to reflect internal banking sectors condition and shorten the crisis window period such as 15, 12, or even 9 months.