Putri, DinaEka (2014) Perbandingan Regresi Robust Penduga Least Trimmed Squares (Lts) Dan Penduga Mm Untuk Pendugaan Model Penilaian Aset Modal (Studi Kasus Saham Pt. Telekomunikasi Indonesia (Persero), Tbk.). Sarjana thesis, Universitas Brawijaya.
Abstract
Metode Kuadrat Terkecil (MKT) merupakan metode pendugaan parameter regresi yang paling umum digunakan. MKT harus memenuhi asumsi regresi agar menghasilkan penduga bersifat Best Linear Unbiased Estimator (BLUE). Menurut teori Capital Asset Pricing Modal (CAPM), parameter risiko sitematik suatu saham dapat diduga menggunakan analisis regresi linier sederhana. Namun jenis data saham rentan terhadap pencilan, sehingga asumsi kenormalan galat tidak terpenuhi. Regresi robust merupakan metode pendugaan parameter yang dapat digunakan apabila galat tidak menyebar normal atau terdapat pencilan yang mempengaruhi model regresi. Peneliti ingin menduga parameter risiko sitematik saham PT. Telekomunikasi Indonesia (Persero), Tbk. di masa mendatang dengan teori CAPM, serta memilih regresi robust dengan penduga mana yang lebih baik berdasarkan kriteria Root Mean Square Error dan salah baku. Parameter risiko diduga dengan penduga Least Trimmed Square (LTS) dan penduga MM pada data saham harian TLKM periode 3, 6, 9, 12 dan 12 dengan jumlah pencilan berbeda. Hasil analisis dengan kedua metode pendugaan menunjukkan saham TLKM diprediksikan bergerak searah harga pasar dengan nilai risiko dan return harapan yang tinggi di masa mendatang. Berdasarkan kriteria Root Mean Square Error dan salah baku, penduga LTS lebih direkomendasikan daripada penduga MM pada teori CAPM.
English Abstract
Ordinary Least Square (OLS) is one of the most commonly used methods on regression analysis. OLS should meet the assumptions of regression analysis to produce Best Linear Unbiased Estimator (BLUE). Based on the theory of Capital Asset Pricing Model (CAPM), parameter of a stock’s systematic risk can be estimated with simple linear regression. But sensitiveness of stock data to outlier makes the assumption of normality is not met. Robust regression is a parameter estimation method which can be used if errors are not normally distributed or there are one or more outliers affecting regression model. This study would like to estimate systematic risk parameter of PT.Telekomunikasi Indonesia (Persero) Tbk. in the future based on CAPM theory, as well as to determine which estimator is the best for robust regression based on criteria Root Mean Square Error (RMSE) and Standard Error (SE). Risk parameter was estimated with Least Trimmed Square (LTS) and MM estimator. Data used was TLKM daily stock price in period of 3, 6, 9, 12 and 15 months with different number of outliers. Result of both estimators showed TLKM stock was predicted to move in direction of market price with higher risk and expected return in the future. Based on criteria RMSE and SE LTS estimator is more recommended than MM estimator for CAPM theory.
| Item Type: | Thesis (Sarjana) |
|---|---|
| Identification Number: | SKR/MIPA/2014/189/051403846 |
| Subjects: | 500 Natural sciences and mathematics > 510 Mathematics |
| Divisions: | Fakultas Matematika dan Ilmu Pengetahuan Alam > Matematika |
| Depositing User: | Budi Wahyono Wahyono |
| Date Deposited: | 14 Jul 2014 13:08 |
| Last Modified: | 21 Oct 2021 03:48 |
| URI: | http://repository.ub.ac.id/id/eprint/153807 |
Preview |
Text
SKRIPSI.pdf Download (3MB) | Preview |
Actions (login required)
 |
View Item |