On Forecast Performance Using a Class of Weighted Moving Average Processes for Time Series

Abstract

In this paper, we consider a class of weighted moving average models called the k-th moving average, the k-th weighted moving average and the k-th exponential weighted moving average models for modeling and forecasting economic time series data. Using real time series data set, we compare the ability of these various models to smooth the available data and also use an out of sample forecast performance to determine the best model among the various competitive models. Our findings is that the k-th exponential weighted moving average model performed best when all the three models where used to smooth our  time series data, while the k-th simple moving average model outperformed the others in terms of future forecasting.

Keywords: k-th moving average; k-th weighted moving average; k-th exponential weighted moving average; forecasting