Chaos Process Testing (Using Local Polynomial Approximation Model) in Predicting Stock Returns in Tehran Stock Exchange

Abstract

Nowadays, the benefits of predicting are undeniably accepted in decision and policy making from different dimensions. Error in predicting makes a model unapplied and transient. Recently, structural models which were relatively successful in explaining the current situation have not been paid much attention in the field of forecasting. Thus, other tests such as local polynomial approximation model have been proposed. This model has introduced to test chaos processes in time-series of daily returns on Tehran Stock Exchange over a period from 2007 to 2013. The obtained findings prove the existence of such process in the evolution of this index.