No abstract

Öz: Yatırımcılar, piyasa riskinden kaçınmak için, portföy çeşitlendirmesinin yanı sıra ekonometrik modeller ile volatiliteyi en iyi şekilde modelleyerek belirsizliği azaltmaya çalışmaktadırlar. Volatiliteyi modellemek için en sık başvurulan yöntemler Otoregresif Koşullu Değişen Varyans ailesi modelleridir. Ancak son yıllarda yapılan bazı çalışmalar, Otoregresif Koşullu Değişen Varyans ailesi modellerinin Yapay Sinir Ağları algoritması ile entegrasyonundan oluşan yarı parametrik hibrit modellerin, yalın modellere göre daha iyi performans sergilediğini göstermektedir. Bundan yola çıkılarak bu çalışmada, bahsi geçen yalın ve hibrit modeller ile Borsa İstanbul 100 fiyat endeks getirisinin volatilitesi tahmin edilmiş ve hibrit modellerinin tahmin başarısı, bileşenlerinin tahmin başarısı ile karşılaştırılmıştır. Tahminlerde, veriler ile ilgili iki farklı dağılım -Normal Dağılım ve Genelleştirilmiş Hata Dağılımı -varsayımı yapılmış ve karşılaştırmalarda Hata Kareleri Ortalaması ve Mutlak Sapma Ortalaması kriterleri kullanılmıştır. Her iki kritere göre, Üssel Genelleştirilmiş Otoregresif Koşullu Değişen Varyans -Yapay Sinir Ağları bileşkesi olan hibrit model en iyi performansı sergilemiştir. Bu bulgular doğrultusunda, finansal araçların dinamik risk analizinde hibrit modellerin sağlayabileceği üstünlüklerin değerlendirilmesi önerilmektedir.

Multiple linear regression model is a widely used statistical technique in social and life sciences. Ordinary Least Squares (OLS) estimator is the Best Linear Unbiased Estimator (BLUE) for the unknown population parameters of this model. Unfortunately, sometimes two or more of the regressors may be moderately or highly correlated causing multicollinearity problem. Various biased estimators are proposed to refine the ill-conditioning of X'X matrix and shrink the variance under the multicollinearity. The most popular of them is Ridge estimator. But it may worsen the fit when solving the ill-conditioning problem. Two-Parameter Ridge (2PR) and Liu Type (LT) estimators are proposed to overcome the fitting degeneration of Ridge estimator by using a tuning parameter. In this study, holding the parameter refining the ill-conditioning of X'X matrix fixed, the success of the tuning parameters of these estimators is investigated. Minimizers of Predicted Sum of Squares (PRESS) and Generalized Cross Validation (GCV) statistics are used as estimates of tuning parameters. Optimum parameter estimates are compared via their Scalar Mean Squared Errors (SMSE). It is observed that the SMSEs of estimates obtained by LT and 2PR estimators decreases when estimates of parameter refining the ill-conditioning of X'X matrix increases, and in all cases estimates obtained by the 2PR estimator are much more efficient than estimates obtained by LT and OLS estimators.

One particular area in financial economics that has received a great deal of attention is the link between exchange rate and the stock prices. The interaction between exchange rate and stock prices has been of special interest because they are regarded among the leading economic variables. The effect of exchange rate on the stock market can work in two avenues. Many studies have documented that changes in the exchange rate have the capacity to increase the volatility of the stock prices, while some other researchers indicated the effect of exchange rate on average returns. In this study, we investigate both of these issues for the case of Istanbul Stock Exchange, using monthly US Dollar-Turkish Lira (USD-TRY) exchange rate and the Istanbul Stock Exchange (BIST) 100 indicex for the period 2009M01-2015M11, employing GARCH approach. Our main findings show that an increase in exchange rate decreases expected returns and increases the riskiness of BIST 100 in Turkey.

A large number of studies have examined the relationship between the trade balance and the exchange rate, focusing on whether depreciations improve the trade balance in the long run and whether this effect is different in the short run. In general, it is argued that the short run effect of currency depreciation is to worsen the trade balance, but this is reversed in the long run; thus, producing the J-curve shape. The aim of this study is to investigate the short run and long run effects of the real effective exchange rate on the trade balance of Turkey and to infer whether Marshall-Lerner condition and J-curve effect exist. To do that, monthly data covering the period 2003:M1-2017:M3 is used. The relationship between variables is estimated using autoregressive distributed lag (ARDL) model, and Bounds test is applied to examine the existence of a long run relationship. Overall results are in favor of a long run relationship and suggest that Marshall-Lerner condition is satisfied. No evidence is found supporting J-curve effect.