Please use this identifier to cite or link to this item:
|Title||Comparative study on forecasting accuracy among moving average models with simulation and PALTEL stock market data in Palestine|
In this paper, we discuss three analytical time series models for selecting the more effective with an accurate forecasting models, among others. We analytically modify the stochastic realization utilizing (i) k-th moving average,(ii) k-th weighted moving average, and (iii) k-th exponential weighted moving average processes. The examining methods have been applied for 1000 independent datasets for five different parameters with possible orders pq 5+≤. We consider stationary data () 0 d=, and non-stationary data with first and second differences () 1, 2 d= for ARIMA models. We consider short term () 50 n= and long term,() 500 n= observations. A similar forecasting models was developed and evaluated for the daily closing price of Stock Price of the PALTEL company in Palestine. The main finding is that, in most simulated datasets one or more of the proposed models give better forecast accuracy than the classical model (ARIMA). Specially, in most simulated datasets 3–time Exponential Weighted Moving Average based on Autoregressive Integrated Moving Average (EWMA3-ARIMA) is the best forecasting model among all other models. For PALTEL Stock Price, the best forecasting model is 3–time Moving Average based on Autoregressive Integrated Moving Average (MA3-ARIMA) among all other models.
|Published in||American Journal of Theoretical and Applied Statistics|
|Series||Volume: 2, Number: 6|
|Item link||Item Link|
|Files in this item|
|Safi, Samir K._0.pdf||223.8Kb|