Faiz Oranları Oynaklığının Modellenmesinde Ardışık Bağlanımlı Koşullu Değişen Varyans Yaklaşımlarının Karşılaştırılarak Değerlendirilmesi (An empirical comparison of alternative AR conditional heteroscedasticity models of volatility of interest rates)

Abstract

With their ability to completely change the term structure of interest rates, short term interest rates provide the most important tool for monetary policy. While short term rates are frequently used in econometric models, no consensus exists in the time series literature with respect to the dynamics of the series or that of their volatility. In this study, different models are used to explain the dynamics of the 12-month interest rate series. The first claim that is tested is that the level of the interest rate is a determinant regressor in explaining the dynamics of volatility of the first difference of the interest rate. This claim is found invalid for the market under consideration. The second claim tested is that as the degree of uncertainty in asset returns varies over time, the compensation required by risk averse economic agents for holding these assets must also be varying. But it is found that the conditional variance is not a determinant regressor on the premium required for holding these risky assets. Using the ‘autoregressive conditional heteroscedasticity’ model and some of its extensions, this study shows that it is not the level but the news effect that determines the volatility of interest rates.