An Investigation On Improvement Of Explanatory Power Of ARCH And State Space Models Using The Haar Wavelet Transform Approach And The Monte Carlo Simulation Method (Case Study: Forecasting the TEPIX Index)

Regarding The Importance Of Forecasting And Its Precision And Accuracy Significance In Various Economic Conditions, This Research Is An Attempt For Denoising Of Stock, By Implementing Of Wavelet Transform (A Branch Of Physic And Specially Signal Analysis) Considering TEPIX Index. Results Indicate That Existence Of Noises Lead To Reduction In Predictive Power Of The Models And Denoising Cause To Improvement In Compatibility Of Dataset With The Models And Finally Improve Predictive Power Of The Models. In This Respect, The ARCH And State Space Models And In-Sample Of 739 Daily Observations, Using The TEPIX Index Data Between Years 1389 – 1392, Are Implemented. The Results Clearly Indicate The Impactful Role Of Haar Wavelet Transform For Denoising TEPIX Dataset. This Method, Strongly, Lead To Increase In The Predictive Power Of The Models And Accuracy Of Estimated Coefficients.