Comparison of Predictive Accuracy of the Indonesia Composite Index (ICI)

Abstract

This study aims to examine the effect of exchange rate and interest rates on the movement of the Composite Stock Price Index using the causality method and pattern approach. The pattern approach method predicts the movement of the stock price index through the movement pattern itself like the ARIMA method while the causality method predicts the movement of the stock price index through the variables that influence it such as the ARCH and GARCH methods. This study was conducted using monthly IHSG data, exchange rates and interest rates as of January 2014 to November 2018. From the results of the study it can be concluded that the exchange rate does not significantly influence the IDX value while the interest rate has a significant effect on the IDX, but the exchange rates and interest rates together the same has a significant negative effect on the value of the IDX. And after the two methods were compared the results were obtained that the ARIMA method gave more accurate predictive results in predicting the IHSG movement. Abstrak Penelitian ini bertujuan untuk menguji pengaruh variabel kurs dan suku bunga terhadap pergerakan Indeks Harga Saham Gabungan dengan menggunakan metode kausalitas dan pendekatan pola. Metode pendekatan pola memprediksi pergerakan indeks harga saham melalui pola pergerakan itu sendiri seperti metode ARIMA sedangkan metode kausalitas memprediksipergerakan indeks harga saham melalui variabel-variabel yang mempengaruhinya seperti metode ARCH dan GARCH. Penelitian ini dilakukan menggunakan data bulanan IHSG, kurs dan suku bunga per januari 2014 hingga November 2018. Dari hasil penelitian dapat disimpulkan bahwa kurs tidak berpengaruh secara signifikan terhadap nilai IHSG sedangkan suku bunga berpengaruh signifikan terhadap nilai IHSG, tetapi kurs dan suku bunga secara bersama-sama berpengaruh negatif secara signifikan terhadap nilai IHSG. Dan setelah kedua metode dibandingkan diperoleh hasil bahwa metode ARIMA memberikan hasil prediksi yang lebih akurat dalam memprediksi pergerakan IHSG.