نمذجة الانحدار الذاتي للعتبة الذاتية مع تطبيق عملي

الملخص EN

Although linear time series models have wide applications for economic phenomena in general, they are not able to capture the behavior of many economic phenomena and applications, especially financial ones.

As this type of series is characterized by modeling the kinetic state of the phenomena of asymmetry, structural changes, threshold, and others.

Therefore, this shortcoming in linear modeling led to the emergence of non-linear models, which are models of various formats and not a model in one general format, as is the case in linear modeling.

In order to overcome this shortcoming, most recent studies have adopted non-linear modeling, and (Tong, 1978) was one of the first who made a qualitative leap in the application of this type of models that depend on the analysis of the dynamics of financial and monetary time series and others, including autoregressive models of the self-threshold.

(SETAR).

This research aims to apply the (SETAR) model to a sample that represents the series of the percentage change in the shares of the Iraq Stock Exchange for the index (ISX60) for a group of companies, and then make an applied comparison of the methods of the mentioned model.

Through the results of the applied side, it was found that the appropriate model for the research sample is the model SETAR (2; 2, 0) with the threshold (gamma = 0.25).

The model has outperformed SETAR (2; 2, 0) for having the lowest values of the criteria (AIC, BIC, pooled-AIC, MAPE), and this model also outperformed the linear model AR (1).