Abstract
In this papers we examine the relative increase in mean square forecast error from fitting a weakly stationary process to the series of interest when in fact the true model is a so-called perturbed long-memory process recently introduced by Granger and Marmol (1997). This model has the property of being unidentifiable from a white noise process on the basis of the correlogram and the usual rule-of-thumbs in the Box-Jenkins methodology. We prove that this kind of misspecification can lead to serious errors in terms of forecasting. We also show that corrections based on the AR(1) model can in some cases partially solve the problem.
Similar content being viewed by others
References
Andersson, M. (1998), On the effects of imposing or ignoring long memory when forecasting, Working Paper 225, Dpt. of Economics and Statistics, Stockholm School of Economics.
Beran, J. (1994), Statistics for Long Memory Processes, New York: Chapman and Hall.
Breidt, F. N. Crato and P. de Lima (1998), ‘The detection and estimation of long memory in stochastic volatility’, Journal of Econometrics 83. 325–348.
Crato, N. and B. Ray (1996), ‘Model selection and forecasting for long-range dependent processes’, Journal of Forecasting 15, 107–125.
Crato, N. and H. Taylor (1996), ‘Stationary persistent time series misspecified as nonstationary ARIMA models’, Statistical Papers 37, 215–223.
Granger, C. and F. Marmol (1997), The correlogram of a long memory process plus a simple noise, Discussion paper 97–29, Dpt. of Economics, UCSD.
Granger, C. and R. Joyeux (1980), ‘An introduction to long memory time series and fractional differencing’, Journal of Time Series Analysis 1, 15–29.
Hampel, F., P. Rousseew, P. Rochetti and W. Stahel (1986), Robust Statistics, New York: John Wiley and Sons.
Harvey, A. C. (1993), ‘Long memory in stochastic volatility’, mimeo, LSE.
Hassler, U. (1994), ‘(Mis)specification of long—memory in seasonal time series’, Journal of Time Series Analysis 15, 19–30.
Hosking, J. (1981), ‘Fractional differencing’, Biometrika 68, 165–176.
Ray, B. (1993), ‘Modeling long-memory processes for optimal long-range prediction’, Journal of Time Series Analysis 14, 511–525.
Smith, J., N. Taylor and S. Yadav (1997), ‘Comparing the bias and misspecification in ARFIMA models’, Journal of Time Series Analysis 18, 507–527.
Smith, J. and S. Yadav (1994), ‘Forecasting costs incurred from unit differencing fractional integrated processes’, International Journal of Forecasting 10, 507–514.
Yajima, Y. (1993), Asymptotic properties of estimates in incorrect ARMA models for long-memory time series, in D. Brillinger, P. Caines, J. Geweke, E. Parzen, M. Rosenblatt and M. Taqqu, eds, ‘New Directions in Time Series Analysis’, number 46 in ‘IMA Series’, New York: Springer-Verlag, pp. 375–382.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Arranz, M.A., Marmol, F. Out-of-sample forecast errors in misspecific perturbed long memory processes. Statistical Papers 42, 423–436 (2001). https://doi.org/10.1007/s003620100071
Received:
Revised:
Issue Date:
DOI: https://doi.org/10.1007/s003620100071