Forecasting an explosive time series


Forecasting is an important exercise in Time series analysis. For a statio-nary time series, there are theoretically strong forecasting methods which canprovide most accurate forecasts for the future (Karlin and Taylor (1975)).For most non stationary time series Box Jenkins methodology is a usefulforecasting technique. Essentially, the Box Jenkins methodology assumesthat any non stationarity time series can be conveniently modeled as anAutoregressive Intregrated Moving Averages (ARIMA) model with sucientnumber of unit roots in the linear stochastic dierence equation generatingthe time series. The non stationarity in such time series is then removed bysuccessively dierencing of the series until one obtains a stationary series,for which optimal forecasts can be computed. The forecasts for the originalseries are then computed by `inverting' the dierence operators that wereused ( Makridakis et al. (1998)) on the forecasts computed for the statio-nary series. The main objective of this study is to demonstrate that the BoxJenkins methodology is not useful, especially in large time series, when thenon stationarity in the time series is due to `explosive' roots. An alternativemethod is proposed in such a situation and its performance is assessed bothon a simulated as well as on a real life data.

DOI Code: 10.1285/i20705948v12n3p674

Keywords: keywords: Stochastic dierence equation; Unit and explosive roots; ARIMA model;Rate of convergence in probability; auxiliary processes.


Bhat, B. R. (1999). Modern Probability Theory. 3/e, New Age International (P) Ltd.,

Publishers, New Delhi.

Box, G. E. P. and Jenkins, G. M. (1976). Time Series Analysis Forecasting and Control.

Holden Day, San Francisco.

Dickey, D. A. and Fuller, W. A. (1979). Distribution of the estimators for autoregressive time series with a unit root. Journal of the American Statistical Association, 74:427-431.

Karlin, S. and Taylor, H. M. (1975). A First Course in Stochastic Processes. 2/e,Academic Press, Inc. New York.

Kendall, M. and Ord, J. (1990). Time Series. Edward Arnold.

Lehmann, E. L. (1986). Testing Statistical Hypothesis. (2/e), John Wiley and Sons, New York.

Makridakis, S., Wheelwright, S. C., and Hyndman, R. J. (1998). Forecasting Methodsand Applications. (3/e), John Wiley and Sons. Inc., New York.

Prabakaran, S. (2015). Statistical analysis of heinous crimes in tamilnadu and related statistical inferences. Ph.D thesis submitted to the University of Madras, Chennai.India.

Suresh Chandra, K., Bhat, B, R., and Srinivasan, T, M. (1994). A modied box-jenkins methodology for explosive time series. In Proceedings of the Third Ramanujan Symposium on Stochastic Processes and their applications, volume 2, pages 89-96. Rama-

nujan Institute, University of Madras.

Suresh Chandra, K. and Janhavi, J. (2008). Unit root tests for time series in the presence of an explosive rootunit root tests for time series in the presence of an explosive root.

Working Paper 27/2008.

Suresh Chandra, K., Manjunath, S. M., and Vaman, H. J. (1999). Least squares estimation for linear explosive models and its ramications. Statistical Methods, 1:54-64.

Venkataraman, K. N. (1968). Limit theorems on a linear stochastic difference equation

with a constant term and their statistical applications. Sankhya, 10:51-74.

Venkataraman, K. N. (1973). Some convergence theorems on a second order linear explosive stochastic difference equation with a constant term. Journal of the Indian Statistical Association, pages 47-69.

Venkataraman, K. N. (1974). Convergence theorems on the least squares estimators of a linear explosive model. Annals of the Institute of Statistical Mathematics, pages 61-72.

Full Text: pdf

Creative Commons License
This work is licensed under a Creative Commons Attribuzione - Non commerciale - Non opere derivate 3.0 Italia License.