MODEL VOLATILITAS ARCH(1) DENGAN RETURN ERROR BERDISTRIBUSI SKEWED STUDENT-T

E. D. Saputri, D. B. Nugroho, A. Setiawan

Abstract


Model volatilitas Autoregressive Conditional Heteroscedasticity (ARCH)lag 1, dimana return error berdistribusi skewed Student-t, diaplikasikan untuk runtun waktu return kurs beli harian Euro (EUR) dan Japanese Yen (JPY) terhadap Indonesian Rupiah (IDR) dari Januari 2009 sampai Desember 2014. Metode indepence chain Metropolis-Hastings (IC-MH) yang efisien dibangun dalam algoritmaMarkov Chain Monte Carlo (MCMC) untuk memperbarui nilai-nilai parameter dalam model yang tidak bisa dibangkitkan secara langsung dari distribusi posterior. Meskipun 95% interval highest posterior density dari parameter skewness memuat nol untuk semua data pengamatan, tetapi sebagian besar distribusi posteriornya berada di daerah negatif, yang mengindikasikan dukungan terhadap distribusiskewed Student-t. Selain itu diperoleh nilai derajat kebebasan disekitar 15 dan 18, yang mengindikasikan dukungan terhadap heavy-tailedness.

Autoregressive Conditional Heteroscedasticity (ARCH) volatility model of lag 1, where return error has a skewed Student-t distribution,  for the buying rate Euro (EUR) and Japanese Yen (JPY) to Indonesian Rupiah (IDR) from January 2009 to December 2014,. An efficient independence chain Metropolis-Hastings (IC-MH) method is developed in an algorithm Markov Chain Monte Carlo (MCMC) to update the parameters of the model that could not be sampled directly from their posterior distributions. Although 95% highest posterior density interval from skewness parameter contains zero for all the data, most of the posterior distribution located in the negative area, indicating support for the skewed Student-t distribution into the return error. Furthermore the value of degrees of freedom is found around 15 and 18, indicating support for the heavy-tailedness.



Keywords


distribusi skewed Student-t, independence-chain Metropolis–Hastings, kurs beli, MCMC, model ARCH

Full Text:

PDF

References


Aas, K., & Haff, I. H. (2006). The generalized hyperbolic skew Student’s t-distribution, Journal of Financial Econometrics, 4, 275309.

Albert, J. (2009). Bayesian computation with R (2nd ed.). Springer.

Chen, M. H, & Shao, Q. M. (1999). Monte Carlo estimation of Bayesian credible and HPD intervals. Journal of Computational and Graphical Statistics, 8, 6992.

Engle, R. F. (1982). Autoregressive conditional heteroskedasticity with estimates of the variance of The United Kingdom Inflation. Journal of Econometrica, 50 (4), 987-1007.

Geweke, J. (1992). Evaluating the accuracy of sampling-based approaches to the calculation of posterior moments dalam Bayesian Statistics 4 (eds. J. M. Bernardo, J. O. Beger, A. P. Dawid dan A. F. M. Smith), 169194.

Johnson, N. L., Kotz, S. & Balakrishnan, N. (1995). Continuous Univariate Distributions (2nd ed.). John Wiley & Sons.

Lo, M. S. (2003). Generalized autoregressive conditional hetroscedastic time series model. A project submitted in partial fulfillment of requirements fordegree of master of science, Simon Fraser University.

Nakajima, J., & Omori, Y. (2012). Stochastic volatility model with leverage and asymmetrically heavy-tailed error using GH skew Student’s t-distribution. Comput. Stat. Data Anal., 56, 36903704.

Safrudin, I. M, Nugroho, D. B, & Setiawan, A. (2015). Estimasi MCMC untuk return volatility dalam model ARCH dengan return error berdistribusi Student-t. Prosiding Sendika Seminar Nasional Matematika dan Pendidikan Matematika, FKIP UMP, 3439.

Saputri, E. D, Nugroho, D. B, & Setiawan, A. (2015). Model volatilitas ARCH(1) dengan returns error berdistribusi non-central Student-t. Prosiding Seminar Nasional Matematika dan Pendidikan Matematika, FMIPA UNY, 233240.

Tierney, L. (1994). Markov chain for exploring posterior distributions. Annals of Statistics, 22 (4), 17011762.

Tsay, R. S. (2002). Analysis of financial time series (2nd ed.). John Wiley & sons.




DOI: https://doi.org/10.15294/ijmns.v39i1.7707

Refbacks

  • There are currently no refbacks.




Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.