Selection of estimation window in the presence of breaks

Abstract In situations where a regression model is subject to one or more breaks it is shown that it can be optimal to use pre-break data to estimate the parameters of the model used to compute out-of-sample forecasts. The issue of how best to exploit the trade-off that might exist between bias and forecast error variance is explored and illustrated for the multivariate regression model under the assumption of strictly exogenous regressors. In practice when this assumption cannot be maintained and both the time and size of the breaks are unknown, the optimal choice of the observation window will be subject to further uncertainties that make exploiting the bias–variance trade-off difficult. To that end we propose a new set of cross-validation methods for selection of a single estimation window and weighting or pooling methods for combination of forecasts based on estimation windows of different lengths. Monte Carlo simulations are used to show when these procedures work well compared with methods that ignore the presence of breaks.

[1]  J. Stock,et al.  A Comparison of Linear and Nonlinear Univariate Models for Forecasting Macroeconomic Time Series , 1998 .

[2]  G. C. Tiao,et al.  Use of Cumulative Sums of Squares for Retrospective Detection of Changes of Variance , 1994 .

[3]  Donald W. K. Andrews,et al.  Optimal changepoint tests for normal linear regression , 1996 .

[4]  P. Patnaik.  THE NON-CENTRAL χ2- AND F-DISTRIBUTIONS AND THEIR APPLICATIONS , 1949 .

[5]  A. Timmermann,et al.  Small Sample Properties of Forecasts from Autoregressive Models Under Structural Breaks , 2003, SSRN Electronic Journal.

[6]  C. Granger,et al.  Handbook of Economic Forecasting , 2006 .

[7]  Allan Timmermann,et al.  Instability of Return Prediction Models , 2005 .

[8]  R. Clemen.  Combining forecasts: A review and annotated bibliography , 1989 .

[9]  Bruce E. Hansen,et al.  Tests for Parameter Instability in Regressions with I(1) Processes , 1992 .

[10]  J. Bai,et al.  Estimating Multiple Breaks One at a Time , 1997, Econometric Theory.

[11]  J. Bai,et al.  Estimation of a Change Point in Multiple Regression Models , 1997, Review of Economics and Statistics.

[12]  H. White,et al.  Monitoring Structural Change , 1996 .

[13]  C. Granger.  Invited review combining forecasts—twenty years later , 1989 .

[14]  J. Durbin,et al.  Techniques for Testing the Constancy of Regression Relationships Over Time , 1975 .

[15]  Davide Pettenuzzo,et al.  Forecasting Time Series Subject to Multiple Structural Breaks , 2004, SSRN Electronic Journal.

[16]  L. Kilian,et al.  In-Sample or Out-of-Sample Tests of Predictability: Which One Should We Use? , 2002, SSRN Electronic Journal.

[17]  Todd E. Clark,et al.  The power of tests of predictive ability in the presence of structural breaks , 2005 .

[18]  F. Diebold,et al.  Forecast Evaluation and Combination , 1996 .

[19]  W. Krämer,et al.  A new test for structural stability in the linear regression model , 1989 .

[20]  P. Perron,et al.  Computation and Analysis of Multiple Structural-Change Models , 1998 .

[21]  D. Andrews.  Tests for Parameter Instability and Structural Change with Unknown Change Point , 1993 .

[22]  Filippo Altissimo,et al.  Strong Rules for Detecting the Number of Breaks in a Time Series , 2003 .

[23]  Michael P. Clements,et al.  Forecast Combination and Encompassing , 2009 .

[24]  P. Perron,et al.  Estimating and testing linear models with multiple structural changes , 1995 .

[25]  Michael P. Clements,et al.  A companion to economic forecasting , 2004 .

[26]  G. Chow.  Tests of equality between sets of coefficients in two linear regressions (econometrics voi 28 , 1960 .

[27]  Todd E. Clark,et al.  Tests of Equal Forecast Accuracy and Encompassing for Nested Models , 1999 .