Publication: Finansal risk değerinin belirlenmesinde kullanılan sayısal yöntemler: ARCH/ GARCH modelleriyle İMKB uygulaması
Abstract
Riskin temel göstergesi olan volatilite, özellikle finansal piyasaların büyük bölümünü oluşturan bankalar için hayati önem taşımaktadır. Finansal risk kavramı çok geniş olup, çalışmada finans kesiminin temelini oluşturan bankaların sektör riski ele alınmış ve bu riski ölçebilmek için İMKB Bankacılık Sektör Endeksinin günlük kapanış değerlerinden oluşan veri setinden yararlanılmıştır. Finansal riski modellemeyebilmek için, hem bankaların taşıdıkları riskleri hem de piyasa risklerini aynı anda sunabilen İMKB Bankacılık Sektör Endeksi tercih edilmiştir. Bu çalışmada finansal riskin belirlenmesinde kullanılan sayısal yöntemler ele alınmış ve bu yöntemlerden biri olan Ardışık Bağımlı Koşullu Değişen Varyans (ARCH) ve Genelleştirilmiş ARCH (GARCH) Modelleri ile uygulama yapılmıştır. Bu yöntemin en büyük avantajı, zaman serilerinin hemen hemen hepsinde görülen, dönemler arası bağımlılığı (otokorelasyon) ve değişken yapıya sahip, sabit olmayan varyansı yani volatiliteyi, serinin geçmiş değerlerinden başka veriye ihtiyaç duymayan modelleyebilme gücüne sahip olmasıdır. Bu çalışmada, finansal verilerde sıkça rastlanan volatilite kümelenmesi, asimetrik fiyat hareketleri, kaldıraç etkisi ve kalın kuyruk özellikleri araştırılmış, ARCH/ GARCH ve türevi modeller tanıtılmış ve bu modellerin İMKB Bankacılık Sektör Endeksinin volatilitesini yani riskini tahmin etme ve öngörme kapasiteleri incelenmiştir. 2007’de dünyada uygulanmaya başlanan ve 2008’de ülkemizde de uygulanacak olan Basel II gerekliliklerinin uygulamaya konmasıyla daha da önemli hale gelecek olan Türkiye’nin Bankacılık Sektör riski global sermayenin de yakından takip ettiği bir konudur.
Volatility, the basic indicator of risk, bears vital importance particularly for banks that represent the majority of the financial markets. Financial risk is a very comprehensive concept, and this study considers the sectoral risk of the banks constituting the backbone of the finance sector and benefits from a set of data consisting of daily closing figures of ISE Banking Sector Index to be able to measure the risk. To be able to model the financial risk, ISE Banking Sector Index has been preferred as it is capable of presenting the banks’ own risks and the market risks. In this study, quantitative methods used in determining financial risk have been evaluated, and Autoregressive Conditional Heteroskedasticity (ARCH) and Generalized ARCH (GARCH) have been used in practice. The greatest advantage of this method is that it is able to model autocorrelation, which we see almost in all time series, and the heteroskedasticity, i.e. volatility, which has an unsteady character, without requiring any data other than the past values of the series. In this study, volatility clustering, asymmetrical price movements, leverage effect and fat tail characteristics which are frequently seen in financial data have been investigated, ARCH/ GARCH and derivative models have been explained, and the capabilities of these models to forecast and estimate the volatility, i.e. the risk of ISE Banking Sector Index, have been examined. The Banking Sector risk in Turkey that will become much more important with the introduction of Basel II requirements which are effective as of 2007 internationally and shall become effective nationally as of 2008 is an issue which is closely monitored by the global capital.
Volatility, the basic indicator of risk, bears vital importance particularly for banks that represent the majority of the financial markets. Financial risk is a very comprehensive concept, and this study considers the sectoral risk of the banks constituting the backbone of the finance sector and benefits from a set of data consisting of daily closing figures of ISE Banking Sector Index to be able to measure the risk. To be able to model the financial risk, ISE Banking Sector Index has been preferred as it is capable of presenting the banks’ own risks and the market risks. In this study, quantitative methods used in determining financial risk have been evaluated, and Autoregressive Conditional Heteroskedasticity (ARCH) and Generalized ARCH (GARCH) have been used in practice. The greatest advantage of this method is that it is able to model autocorrelation, which we see almost in all time series, and the heteroskedasticity, i.e. volatility, which has an unsteady character, without requiring any data other than the past values of the series. In this study, volatility clustering, asymmetrical price movements, leverage effect and fat tail characteristics which are frequently seen in financial data have been investigated, ARCH/ GARCH and derivative models have been explained, and the capabilities of these models to forecast and estimate the volatility, i.e. the risk of ISE Banking Sector Index, have been examined. The Banking Sector risk in Turkey that will become much more important with the introduction of Basel II requirements which are effective as of 2007 internationally and shall become effective nationally as of 2008 is an issue which is closely monitored by the global capital.