Harun ÖZTÜRKLER, Harun ÖZTÜRKLER, Selim YILDIRIM

İSTANBUL MENKUL KIYMETLER BORSASI 100 ENDEKSİNİN DEK DEĞİŞKENLİ DOĞRUSAL OLMAYAN BİR MODELİ

Ekonomi ve finans alanında asimetrik davranış yaygın olarak gözlemlenir. Asimetrik davranışı doğrusal modellerle modellemek olanaklı olmadığından, bu tür zaman serilerinin sergilediği asimetrik davranışı açıklamak için uygun doğrusal olmayan modeller geliştirilir. Bu çalışmanın bulguları İMKB 100 endeksinin 2000 yılı sonrasında sergilediği davranışının doğrusal modeller ile tahmin edilemeyeceğini göstermektedir. Bu nedenle bu çalışmanın amacı İMKB 100 endeksinin zaman serisinin doğrusal olmayan bir modelini kurgulamak ve tahmin etmektir. Kurgulanan ve hesaplanan modelin sonuçları İMKB 100 endeksinin doğrusal olmayan bir davranış sergilediğini doğrulamaktadır

Anahtar Kelimeler:

İMKB 100, doğrusal olmayan zaman serileri, dek değişkenli zaman serileri

A UNIVARIATE NONLINEAR MODEL OF THE RETURNS ON ISTANBUL STOCK EXCHANGE 100 INDEX

Asymmetric behaviors are common in economics and finance. Since it is not possible to capture asymmetric behaviors by linear models, nonlinear models are developed in order to explain asymmetric behaviors exhibited by such time series. Findings in this study show that ISE 100 index’s behavior cannot be estimated by linear univariate models for the period after 2000. Therefore, it is our aim to construct and estimate nonlinear time series models of ISE 100 index. The results obtained also confirm that ISE 100 index exhibits nonlinear behavior.

Keywords:

ISE 100, nonlinear time series, univariate time series.,

PDF

___

Baragona, R., & Battaglia, F. (2006). Genetic Algorithms for building Double Threshold Generalized Autoregressive Conditional Heteroscedastic Model of Time Series. In Alfredo Rizzi, Maurizio Vichi (Ed.), COMPSTAT 2006 – Proceedings in Computational Statistics, Part VI. Germany: Physica-Verlag.

Baragona, R., & Cucina, D. (2008). Double Threshold Autoregressive Conditionally Heteroscedastic Model Building Genetic Algorithms. Journal of Statistical Computation and Simulation, 78(6), 541-558. Chan, K. (February 14, 2012 – last update). Package “TSA” Reference Manual [Online]. Available: http://cran.r-project.org/web/packages/ TSA/TSA. pdf [March 25, 2012].

Engle, R. F. (1982). Autoregressive Conditional Heteroscedasticity with Estimates of the Variance of United Kingdom Inflation. Econometrica, 50(4), 987-1007.

Grassberger, P. & Procaccio I. (1983). Measuring the Strangeness of Strange Attractors. Physica D, 9(1-2), 189-208.

Ince, H. (2005). Non Parametric Regression Methods and anApplication to Istanbul Stock Exchange 100 (ISE 100) Index. Yapı Kredi Economic Review, 16(1), 17-28.

Knatz, H. & Schreiber, T. (2004). Nonlinear Time Series Analysis. New York: Cambridge University Press.

Kocenda, E. & Cerny, A. (2007). Elements of Time Series Econometrics: An Applied Approach., Prague: Charles University, Karolium Press.

Li, C. W. & Li, W. K. (1996). On a Double-Threshold Autoregressive Hetereoscedastic Time Series Model. Journal of Applied Econometrics, 11(3), 253-274.

Schwert, G. W. (2011). Stock Volatility During the Recent Financial Crises. NBER Working Paper Series, No. 16976.

Stigler, Matthieu. (February 16, 2012- last update). Threshold Cointegration: Overview and Implementation in R [Online]. Available: http://cran.rproject.org/web/packages/tsDyn/ vignettes/ThCointOverview.pdf [March 25, 2012]

Tong, H. (1993). Non-linear Time Series: A Dynamical System Approach, UK: Oxford University Press.

Tsay, R. S. (2005). Analysis of Financial Time Series, New Jersey: WileyInterscience.