Box-Jenkins ve Nonparametrik Regresyon Yöntemlerinin Etkinliklerinin Karsilastirilmasi: IMKB-100 Endeksine Yonelik Bir Uygulama

It is a well known fact that time series and data are commonly and frequently used in applied researches and various methods are developed for analysis of these data. The most common method used for analysis of monovariate time series is Box- Jenkins method which is based on modelling of a time series with its own lagged values and error terms. Another recent method used for analysing the monovariate time series is nonparametric regression method which is based on a certain function instead of coefficients and where the estimations are done through this function. The two methods share the same aim which is to model the time series and to forecast making use of this model. This study aims at realizing a practical comparison of the efficiency of Box-Jenkins method and nonparametric regression method, which are used for analysis of monovariate time series, on the basis of monthly closing prices of ISE National Index 100. As a result of the analyses performed in line with this aim the comparisons performed as for various performance criteria revealed that the nonparametric regression method gives more effective results than Box-Jenkins method.