Yapısal kırılmalar olduğunda uyarlanmış ve basit üstel düzeltme yöntemlerinin karşılaştırılması

The essential aim of the time series modelling is applied for the forecasting as well as the examination of correlation. One of the most widely used methods in the literature is exponential smoothing (ES) methods. It is a method preferred by many researchers because of its easy application, calculation efficiency, high accuracy and automatic prediction. Such as policy changes, financial crises, natural disasters in the data production processes of the series, permanent structural changes can change affect model parameters as well as analysis results. Having no consideration of such affects, leads parameter estimator to be biased, tests tend to be useless in terms of power and incorrect modelling arise. The main purpose of this study is to compare the predictive performances of the newly developed Modified Exponential Smoothing (MSES) (2016) methods with the simple exponential smoothing (SES) when there are structural breaks in the series. Determining the initial value and misspecification in the selection of the optimum smoothing parameter, as a disadvantage, adversely affect the estimation results. The MSES method gives more weight to the current observations on the series, so that the predictions that are calculated give better performance than the classical method. The MSES method against structural break has not been examined yet. In this study, received from the Central Bank of the Republic of Turkey and traded on the Istanbul Gold Exchange "weighted average price of gold (TL/kg)" data are used. This data set with different break points compares the forecast performance of MSES and SES methods.