Pemodelan Yield Curve Obligasi Pemerintah Indonesia Dengan Robust Locally Weighted Regression Smoothing Scatterplots

Anggraeni, Deni Dewi (2009) Pemodelan Yield Curve Obligasi Pemerintah Indonesia Dengan Robust Locally Weighted Regression Smoothing Scatterplots. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 1305100059-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (7MB) | Request a copy

Abstract

Obligasi adalah surat pengakuan utang dengan kesanggupan untuk mengembalikan pokok utang pada waktu yang Ielah ditentukan. Salah satu atribut penting dalam penilaian harga obligasi adalah struktur yield Penilaian yield untuk penelitian ini digunakan pendekatan Yield to Maturity atau YTM. Kajian tentang pemodelan yield curve belum banyak dilakukan di Indonesia. Metode yang digunakan untuk memodelkan yield curve dalam penelitian ini adalah Robust Locally Weighted Regression Smoothing Scatterplots (RLWRSS) dan Nelson Siegel Svensson (NSS). RLWRSS merupakan salah satu metode yang tegar terhadap outlier. Sedangkan NSS merupakan model terbaik pada peneltian yield curve sebelumnya. Data obligasi pemerintah Indonesia didapatkan dari BAPEPAM melalui IBPA. Data terse but merupakan obligasi pemerintah periode Januari hingga April 2009 yang berbentuk 82 pasangan YIM dan 11M. Data yang digunakan dalam penelitian ini sebanyak 12 pasangan YIM-TIM yang dipilih berdasarkan kecukupan data untuk proses validasi. Hasil penelitian ini menunjukkan bahwa ni/ai Root Mean Square Error (RMSE) in sample metode RLWRSS lebih kecil dibandingkan dengan NSS. Sedangkan berdasarkan nilai RMSE out sample, NSS lebih baik dibandingkan RLWRSS. Sehingga NSS merupakan metode yang terbaik dan sesuai untuk obligasi pemerintah Indonesia berdasarkan nilai RMSE out sample dan kehalusan kurva yang dihasilkan
===================================================================================================================================
Bonds are debt instruments with a promise to repay the principal at a specified time. One important attribute in bond pricing is the yield structure. The yield assessment for this study used the Yield to Maturity (YTM) approach. Studies on yield curve modeling have not been widely conducted in Indonesia. The methods used to model the yield curve in this study are Robust Locally Weighted Regression Smoothing Scatterplots (RLWRSS) and Nelson Siegel Svensson (NSS). RLWRSS is a method that is robust to outliers. Meanwhile, NSS is the best model in previous yield curve studies. Indonesian government bond data was obtained from BAPEPAM through IBPA. The data are government bonds from January to April 2009 in the form of 82 YIM and 11M pairs. The data used in this study were 12 YIM-TIM pairs selected based on data sufficiency for the validation process. The results of this study indicate that the Root Mean Square Error (RMSE) value in the RLWRSS sample method is smaller than that of the NSS. Meanwhile, based on the out-sample RMSE value, NSS is better than RLWRSS. Therefore, NSS is the best and most appropriate method for Indonesian government bonds based on the out-sample RMSE value and the smoothness of the resulting curve

Item Type:	Thesis (Other)
Additional Information:	RSSt 519.536 Ang p-1 2009 (weding)
Uncontrolled Keywords:	Obligasi, yield curve, NSS, RLWRSS, outlier, RMSE; Bond, yield curve, NSS, RLWRSS, outlier, RMSE
Subjects:	Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression
Divisions:	Faculty of Mathematics and Science > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	EKO BUDI RAHARJO
Date Deposited:	12 Sep 2025 02:04
Last Modified:	12 Sep 2025 02:04
URI:	http://repository.its.ac.id/id/eprint/128202

Actions (login required)

View Item