ARMA Models and the Box-Jenkins Methodology

Abstract

The purpose of this paper is to apply the Box±Jenkins methodology to ARIMA models and determine the reasons why in empirical tests it is found that the post-sample forecasting the accuracy of such models is generally worse than much simpler time series methods. The paper concludes that the major problem is the way of making the series stationary in its mean (i.e. the method of dierencing) that has been proposed by Box and Jenkins. If alternative approaches are utilized to remove and extrapolate the trend in the data, ARMA models outperform the models selected through Box±Jenkins methodology. In addition, it is shown that using ARMA models to seasonally adjusted data slightly improves post-sample accuracies while simplifying the use of ARMA models. It is also con®rmed that transformations slightly improve post-sample forecasting accuracy, particularly for long forecasting horizons. Finally, it is demonstrated that AR(1), AR(2) and ARMA(1,1) models can produce more accurate post-sample forecasts than those found through the application of Box± Jenkins methodology. # 1997 by John Wiley & Sons, Ltd.

Forecasting on the long-term sustainability of the employees provident fund in Malaysia via the Box-Jenkins’ ARIMA model

“…Here, one benefit of the conversion is that the use of log form ascertains each series to possess stationary properties in its variance (Box & Jenkins, 1970 as cited in Makridakis & Hibon, 1997). Besides that, it enables for the statistical meanings of coefficients to be interpreted as the long run elasticity effects.…”

Statistical Modeling for Real Domestic Hot Water Consumption Forecasting

“…Since these observations ( ) are following a temporal dimension in discrete time, they can be defined as time series model like ARIMA [3] (Auto Regressive Integrated Moving Average) in order to analyze the previous behavior and predict its future behavior. This model is a generalized form of the stationary ARMA model [6] (Auto Regressive Moving Average) and allows to take into account the non-stationary and the periodicity of the data. An ARIMA process ( , ) of order ( , , ) is defined by three components: the autoregressive model of order , the integration of order , the moving average of order and given by:…”

Study of Time Series Data Mining for the Real Time Hydrological Forecasting: A Review

“…In their research, ARIMA (1, 0, 0) (0, 1,1) 12 model was developed. This model is used to forecasting the monthly rainfall for the upcoming 10 years to help decision makers establish priorities in terms of water demand management [30]. Kothyari et al (1997) studied rainfall and temperature (i.e., long-term monsoon rainfall, number of rainy days during the monsoon season, and annual maximum temperature) from three stations at Agra, Dehradun and Delhi for evaluating the changes in regimes in the upper and middle parts of the Ganga basin in northern India [31].…”

“…Makridakis S. and Hison M. presented various aspects of the Box-Jenkins methodology to ARMA models in this paper. The major concern of this study is to identify the post sample forecasting methods and improve the accuracy in order to ascertain static form of data [30]. Boudaghpour, S. et al (2014) explain the stochastic AR (1) and AR (2) models which used for the case study of TSS in Lighvan Chia basin.…”