Identifikasi Korelasi Nilai Tukar Rupiah dan Indeks Harga Saham Gabungan Melalui Ensemble Empirical Mode Decomposition

Abstract

Capital market indicators such as the composite stock price index may fluctuate with changes in economic indicators such as exchange rates also fluctuated. This study identified a correlation between the exchange rate and composite stock price index at different time scales. The difference in the time scale is obtained by decomposition of the ensemble of the initial data using Empirical Mode Decomposition Ensemble (EEMD) as a pre-condition. The EEMD results show that IMF-IMF contribute greatly to the volatility of each of the monthly exchange rate and composite stock price index, the IMF’s pair relationship more closely. In the period from January 1990 until November 2013, cyclical events around 3.2 years and 12 years to dominate the movement of data, with correlation coefficients IMF’s pair higher is equal to -0.499 and -0.726. Meanwhile, in the period August 1997 to November 2013, the events that have approached the mean period of 1.5 years and 2.7 years also dominates the movement of data, so it has a correlation coefficient higher IMF’s pair are -0.268 and -0.539. EEMD shown to reveal the local relationship between the exchange rate and composite stock price index was overall positively related, but after exploring the different cycles, both of which also have a negative relationship when the appreciation or depreciation.