Forecasting Vector ARMA Processes with Systematically Missing Observations

The following two predictors are compared for time series with systematically missing observations: (a) A time series model is fitted to the full series Xt , and forecasts are based on this model, (b) A time series model is fitted to the series with systematically missing observations Y τ, and forecasts are based on the resulting model. If the data generation processes are known vector autoregressive moving average (ARMA) processes, the first predictor is at least as efficient as the second one in a mean squared error sense. Conditions are given for the two predictors to be identical. If only the ARMA orders of the generation processes are known and the coefficients are estimated, or if the process orders and coefficients are estimated, the first predictor is again, in general, superior. There are, however, exceptions in which the second predictor, using seemingly less information, may be better. These results are discussed, using both asymptotic theory and small sample simulations. Some economic time ser...

[1]  Peter C. B. Phillips,et al.  The sampling distribution of forecasts from a first-order autoregression , 1979 .

[2]  Barry G. Quinn,et al.  Order Determination for a Multivariate Autoregression , 1980 .

[3]  G. Reinsel,et al.  Prediction of multivariate time series by autoregressive model fitting , 1985 .

[4]  Richard H. Jones,et al.  Maximum Likelihood Fitting of ARMA Models to Time Series With Missing Observations , 1980 .

[5]  William W. S. Wei,et al.  Effect of systematic sampling on arima models , 1981 .

[6]  Helmut Lütkepohl,et al.  Linear aggregation of vector autoregressive moving average processes , 1984 .

[7]  Helmut Lütkepohl,et al.  Forecasting Contemporaneously Aggregated Vector ARMA Processes , 1984 .

[8]  J. Burbidge,et al.  Testing for the Effects of Oil-Price Rises using Vector Autoregressions , 1984 .

[9]  R. Baillie.  PREDICTIONS FROM ARMAX MODELS , 1980 .

[10]  E. J. Hannan,et al.  Multiple time series , 1970 .

[11]  Jostein Paulsen,et al.  ORDER DETERMINATION OF MULTIVARIATE AUTOREGRESSIVE TIME SERIES WITH UNIT ROOTS , 1984 .

[12]  R. Shibata.  Asymptotically Efficient Selection of the Order of the Model for Estimating Parameters of a Linear Process , 1980 .

[13]  G. Schwarz.  Estimating the Dimension of a Model , 1978 .

[14]  Roque B Fernandez,et al.  A Methodological Note on the Estimation of Time Series , 1981 .

[15]  C. Sims.  MACROECONOMICS AND REALITY , 1977 .

[16]  Taku Yamamoto,et al.  Predictions of multivariate autoregressive-moving average models , 1981 .

[17]  W. Fuller,et al.  Properties of Predictors for Autoregressive Time Series , 1981 .

[18]  Cheng Hsiao,et al.  AUTOREGRESSIVE MODELLING AND MONEY-INCOME CAUSALITY DETECTION , 1981 .

[19]  Taku Yamamoto,et al.  Asymptotic mean square prediction error for an autoregressive model with estimated coefficients , 1976 .

[20]  B. G. Quinn,et al.  The determination of the order of an autoregression , 1979 .

[21]  A. A. Weiss.  Systematic sampling and temporal aggregation in time series models , 1984 .

[22]  Robert B. Litterman.  A random walk, Markov model for the distribution of time series , 1983 .

[23]  H. Lütkepohl.  COMPARISON OF CRITERIA FOR ESTIMATING THE ORDER OF A VECTOR AUTOREGRESSIVE PROCESS , 1985 .

[24]  M. H. Quenouille.  Discrete autoregressive schemes with varying time-intervals , 1958 .

[25]  Andrew Harvey,et al.  Estimating Missing Observations in Economic Time Series , 1984 .

[26]  F. Palm,et al.  Missing observations in the dynamic regression model , 1984 .

[27]  G. Reinsel,et al.  Asymptotic Properties of Prediction Errors for the Multivariate Autoregressive Model Using Estimated Parameters , 1980 .

[28]  Arnold Zellner,et al.  On the Analysis of First Order Autoregressive Models with Incomplete Data , 1966 .

[29]  R. Shibata.  Selection of the order of an autoregressive model by Akaike's information criterion , 1976 .

[30]  Richard T. Baillie,et al.  Asymptotic prediction mean squared error for vector autoregressive models , 1979 .

[31]  T. Nijman.  Missing observations in dynamic macroeconomic modeling , 1985 .

[32]  G. Chow,et al.  Best Linear Unbiased Interpolation, Distribution, and Extrapolation of Time Series by Related Series , 1971 .

[33]  K. Brewer.  Some consequences of temporal aggregation and systematic sampling for ARMA and ARMAX models , 1973 .

[34]  Taku Yamamoto.  On the Treatment of Autocorrelated Errors in the Multiperiod Prediction of Dynamic Simultaneous Equation Models , 1980 .

[35]  H. Werner.  To the temporal aggregation in discrete dynamical systems , 1982 .

[36]  C. Granger,et al.  Forecasting Economic Time Series. , 1988 .

[37]  Emanuel Parzen,et al.  ON SPECTRAL ANALYSIS WITH MISSING OBSERVATIONS AND AMPLITUDE MODULATION , 1962 .

[38]  Richard H. Jones,et al.  SPECTRAL ANALYSIS WITH REGULARLY MISSED OBSERVATIONS , 1962 .

[39]  Cheng Hsiao,et al.  Autoregressive Modeling of Canadian Money and Income Data , 1979 .