Early detection of the financial crisis in Indonesia, Thailand, and South Korea based on real exchange rates

Abstract

The 1997 Asian financial crisis and the 2008 global financial crisis had a significant impact on the economies of various countries, including Indonesia, Thailand, and South Korea. Therefore, this study aims to detect early financial crises in these three countries based on real exchange rate indicators. The early detection of financial crises, relevant institutions in applying appropriate policies. Thus, financial crises can be avoided. The real exchange rate indicator contains fluctuations, and changing conditions. As a solution, the volatility models, i.e. autoregressive conditional heteroscedasticity (ARCH) and generalized ARCH (GARCH) explains volatility or fluctuations, and Markov switching (MS) explains changing conditions. The combination of these models yields a smoothed probability value used for detecting financial crises. The real exchange rate of Indonesia, Thailand, and South Korea data against the United States were used from January 1990 to March 2020. The results showed that MSARCH( 2,1), MS-ARCH(2,1), and MS-GARCH(2,1,1) are for Indonesia, Thailand, and South Korea, respectively. The real exchange rate indicator can detect crises in Indonesia and South Korea in the 1997 and 2008. Whereas in Thailand, it was only able to detect crises occured in the 1997. The prediction results of the smoothed probability showed no signal of the financial crisis in Indonesia, Thailand, and South Korea one year later.