Pemodelan Box-Jenkins (Arima) Untuk Peramalan Indeks Harga Saham Gabungan

Primaditya, Vincentius Iwan (2015) Pemodelan Box-Jenkins (Arima) Untuk Peramalan Indeks Harga Saham Gabungan. Masters thesis, Institut Teknologi Sepuluh Nopember.

[img]
Preview
Text
9110205441-Master Thesis.pdf - Published Version

Download (2MB) | Preview

Abstract

Indeks Harga Saham Gabungan (IHSG) merupakan indikator utama yang digunakan di Bursa Efek Indonesia (BEI) untuk mengukur kinerja pasar saham secara keseluruhan. Peramalan yang akurat mengenai pergerakan indeks dapat menghasilkan keuntungan bagi investor dan untuk mengembangkan strategi perdagangan pasar saham yang efektif. Autoregressive Integrated Moving Average (ARIMA) adalah model deret waktu yang berguna untuk peramalan indeks harga saham dengan menggunakan autokorelasi dan variasi residual deret waktu. Masing-masing tahap dalam pemodelan Box-Jenkins dilakukan dan menghasilkan model ARIMA(3,1,2). Jumlah lag autoregresi yang dihasilkan kemudian digunakan sebagai input ke dalam sistem Jaringan Syaraf Tiruan (JST) untuk menghasilkan model prediksi Neural Network Autoregression (NAR). Untuk mengetahui apakah pasar saham di BEI sesuai dengan hipotesis pasar efisien bentuk lemah, maka peramalan dengan model Random Walk with Drift juga dilakukan. Hasil evaluasi akurasi peramalan ketiga model pada dataset testing menggunakan Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), dan Mean Absolute Percentage Error (MAPE) menunjukkan bahwa NAR(3) menghasilkan kesalahan peramalan terkecil dengan akurasi yang berada dalam kisaran 98%. Hasil penelitian keseluruhan memberikan indikasi bahwa pasar saham di BEI tidak sesuai dengan hipotesis pasar efisien bentuk lemah. ===================================================================================================== The Jakarta Composite Index (JCI) is the main indicator used in the Indonesia Stock Exchange (IDX) to measure the overall performance of the stock market. Accurate forecasting of the movement of the index would benefit investors and to develop strategies for effective stock market trading. Autoregressive Integrated Moving Average (ARIMA) time series model is useful for forecasting stock price index by using autocorrelation and residual variation in the time series. Each stage in the Box-Jenkins modeling were performed and produced ARIMA (3,1,2). The order of lag of the autoregression generated is then used as input to an Artificial Neural Network (ANN) system to generate a predictive model Neural Network Autoregression (NAR). To determine whether the IDX stock market is in accordance with the weak form of the efficient-market hypothesis, the Random Walk with Drift forecasting model was also performed. Forecasting accuracy evaluation of the three models on testing dataset using the Mean Absolute Error (MAE), Root Mean Squared Error (RMSE) and Mean Absolute Percentage Error (MAPE) indicates that the NAR (3) generates the smallest forecasting error with accuracy within the range of 98%. Overall study results provide an indication that the IDX is not in accordance with the weak form of the efficient-market hypothesis.

Item Type: Thesis (Masters)
Additional Information: RTMT 519.535 Pri p
Uncontrolled Keywords: arima, autokorelasi, box-jenkins, indeks harga saham gabungan, peramalan, prediksi, random walk
Subjects: H Social Sciences > HB Economic Theory > Economic forecasting--Mathematical models.
Divisions: Magister Management Technology
Depositing User: Mr. Tondo Indra Nyata
Date Deposited: 02 Apr 2018 07:24
Last Modified: 24 Aug 2018 04:22
URI: http://repository.its.ac.id/id/eprint/51665

Actions (login required)

View Item View Item