Please use this identifier to cite or link to this item: http://repository.ipb.ac.id/handle/123456789/114182
Title: Peramalan Nilai Tukar Rupiah Terhadap Dolar Amerika dengan Pendekatan Analisis Regresi dan ARIMA
Authors: Budiarti, Retno
Septyanto, Fendy
Nurpiah, Siti
Issue Date: 2022
Publisher: IPB University
Abstract: Nilai tukar merupakan harga mata uang satu negara yang dinyatakan dalam mata uang negara lain. Nilai tukar berfluktuatif sehingga perlu dilakukan peramalan agar dapat membantu pelaku ekonomi dalam membuat keputusan yang tepat. Fluktuasi nilai tukar dipengaruhi oleh faktor-faktor makroekonomi, seperti inflasi dan jumlah uang beredar. Pengaruh faktor-faktor makroekonomi terhadap nilai tukar dapat dimodelkan dengan regresi linear berganda. Parameter model regresi diestimasi menggunakan metode Ordinary Least Square (OLS). Pada model regresi linear berganda terdapat asumsi-asumsi klasik yang harus dipenuhi. Apabila terjadi korelasi antara residual ke-t dengan residual ke-(t-1), dikatakan residual model mengalami autokorelasi. Dalam karya ilmiah ini, digunakan metode Cochrane-Orcutt untuk mengatasi pelanggaran asumsi tidak adanya autokorelasi pada residual model. Selain faktor makroekonomi, fluktuasi nilai tukar juga dipengaruhi oleh nilai tukar pada periode sebelumnya. Pada karya ilmiah ini dilakukan peramalan nilai tukar rupiah terhadap dolar AS menggunakan metode regresi linear dan ARIMA. Setelah itu, akan dilakukan perbandingan hasil ramalan menggunakan Mean Absolute Percentage Error (MAPE). Hasil peramalan menunjukkan bahwa metode regresi linear memiliki tingkat keakuratan yang lebih baik dibandingkan dengan model ARIMA karena memiliki nilai MAPE yang lebih kecil.
Exchange rate is the price of one country’s currency which is expressed in other country’s currency. The exchange rate is volatile so it is necessary to forecast in order to assist economic actors in making the right decisions. Exchange rate fluctuations are influenced by macroeconomic factors such as inflation and money supply. The effect of macroeconomic factors on the exchange rate can be estimated using multiple linear regression. The regression parameters are estimated using the Ordinary Least Square (OLS) method. There are classical assumptions that must be satisfied in multiple linear regression. If there is correlation between the (t)-th residual and the (t-1)-th residual then the residual model has autocorrelation. In this study, the Cochrane-Orcutt method is used to overcome the violation of assumption that there is no autocorrelation in the residual model. The exchange rate fluctuations are also influenced by exchange rate in previous period. In this study, the exchange rate is forecasted using linear regression and ARIMA. Then, their accuracy are compared by evaluating Mean Absolute Percentage Error (MAPE). The forecasting results showed that linear regression has a better accuracy rate than ARIMA model because it has a smaller MAPE value.
URI: http://repository.ipb.ac.id/handle/123456789/114182
Appears in Collections:UT - Actuaria

Files in This Item:
File Description SizeFormat 
Cover.pdf
  Restricted Access
Cover460.2 kBAdobe PDFView/Open
G94180017_Siti Nurpiah.pdf
  Restricted Access
Fullteks1.02 MBAdobe PDFView/Open
Lampiran.pdf
  Restricted Access
Lampiran898.85 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.