Analisis perbandingan metoda Value-at-Risk dan Traditional Risk Measures dalam menjelaskan hubungan risiko dan return Portofolio Saham

MUAL, Tania Maureen, Sukmawati, Prof.,Dr.,MM

2007 | Tesis | Magister Manajemen

Penelitian ini bertujuan untuk mengetahui metode pengukuran risiko mana diantara traditional risk measures (standar deviasi dan beta) dan metode Value at Risk (VaR Variance Covariance dan VaR Historical Simulation) yang dapat lebih menjelaskan return saham portofolio. Prosedur penelitian dilakukan dengan mengambil sampel saham-saham yang termasuk dalam anggota LQ45 dengan periode penelitian selama 5 tahun (2 Januari 2002 s/d 28 Desember 2006). Sampel sebanyak 13 emiten kemudian dihitung nilai risikonya menggunakan traditional risk measures dan metode Value at Risk. Pengukuran dilakukan dalam sampel yang dibagi ke dalam 3 tipe portofolio yang kemudian hasil pengukuran risiko masing-masing portofolio diregresikan terhadap returnnya. Hasil dari analisis regresi berganda adalah variabel independent yang signifikan berpengaruh terhadap return portofolio I, II, maupun III adalah VARHS 95% dan VARHS99% yang juga memiliki pengaruh positif terhadap pergerakan return. Hasil uji Multikolinearitas dan Autokorelasi yang dilakukan terhadap model regresi memberikan hasil bahwa model regresi terbebas dari multikolineritas dan autokorelasi sehingga model regresi tersebut dapat dipakai.

This research planned to know the comparison between risk measurement method among traditional risk measures ( standard of deviation and beta) and Value at Risk method ( VaR Variance Covariance and VaR Historical Simulation) which can more explaining the return of stock portfolio. Research procedure conducted by taking samples of stocks which are included in member LQ45 with the research period during 5 year ( 2 January 2002 s / d 28 December 2006). Samples as much 13 stocks was calculated by its risk value using the traditional risk measures and method of Value at Risk. The measurement taking stocks samples into three type portfolio. Then the risk measurement result of each portfolio regressed for it each return. The result from multiple regression analysis found that variable independent which have significance effect to return portfolio I, II, and also III are VARHS 95% and VARHS99%. And they also own the positive influence to movement of the portfolio return. From the test of Multicollinearity and Autocorrelation there was some findings can give conclusion that this portfolios regression models are free from the multicollinearity and autocorrelation so these model are reliable to use.

Kata Kunci : Manajemen Resiko,Value at Risk, Risk measurement method, traditional risk measures, Value at Risk. Regrresed