A Quantile Regression Approach to Generating Prediction Intervals

Exponential smoothing methods do not involve a formal procedure for identifying the underlying data generating process. The issue is then whether prediction intervals should be estimated by a theoretical approach, with the assumption that the method is optimal in some sense, or by an emp irical procedure. In this paper we present an alternative hybrid approach which applies quantile regression to the empirical fit errors to produce forecast error quantile models. These models are functions of the lead time, as suggested by the theoretical variance expressions. In addition to avoiding the optimality assumption, the method is nonparametric, so there is no need for the common normality assumption. Application of the new approach to simple, Holt's, and damped Holt's exponential smoothing, using simulated and real data sets, gave encouraging results.

[1]  Kenneth F. Wallis,et al.  Density Forecasting: A Survey , 2000 .

[2]  Nigel Meade,et al.  Prediction intervals for growth curve forecasts , 1995 .

[3]  Víctor M. Guerrero,et al.  Restricted forecasts using exponential smoothing techniques , 1994 .

[4]  Pierre Lefrançois,et al.  On the estimation of time-series quantiles using smoothed order statistics , 1993 .

[5]  Chris Chatfield,et al.  Calculating Interval Forecasts , 1993 .

[6]  William Gould,et al.  Quantile regression with bootstrapped standard errors , 1993 .

[7]  W. Rogers.  Quantile regression standard errors , 1993 .

[8]  C. Chatfield,et al.  Prediction intervals for multiplicative Holt-Winters , 1991 .

[9]  C. Chatfield,et al.  Prediction intervals for the Holt-Winters forecasting procedure , 1990 .

[10]  James H. Bookbinder,et al.  Estimation of Inventory Re-Order Levels Using the Bootstrap Statistical Procedure , 1989 .

[11]  Robert L. Winkler,et al.  Sampling Distributions of Post-sample Forecasting Errors , 1989 .

[12]  Everette S. Gardner,et al.  A Simple Method of Computing Prediction Intervals for Time Series Forecasts , 1988 .

[13]  F. R. Johnston,et al.  The Variance of Lead-Time Demand , 1986 .

[14]  E. S. Gardner,et al.  Forecasting Trends in Time Series , 1985 .

[15]  Wilpen L. Gorr,et al.  An Adaptive Filtering Procedure for Estimating Regression Quantiles , 1985 .

[16]  Everette S. Gardner,et al.  Exponential smoothing: The state of the art , 1985 .

[17]  Bovas Abraham,et al.  Some comments on the initialization of exponential smoothing , 1984 .

[18]  S. A. Roberts.  A General Class of Holt-Winters Type Forecasting Models , 1982 .

[19]  Robert L. Winkler,et al.  The accuracy of extrapolation (time series) methods: Results of a forecasting competition , 1982 .

[20]  R. Koenker,et al.  Robust Tests for Heteroscedasticity Based on Regression Quantiles , 1982 .

[21]  A. J. Kinderman,et al.  Computer Generation of Normal Random Variables , 1976 .

[22]  W. H. Williams,et al.  A Simple Method for the Construction of Empirical Confidence Limits for Economic Forecasts , 1971 .