In practical econometric forecasting exercises, incomplete data on current and immediate past values of endogenous variables are available. This paper considers various approaches to this 'ragged edge' problem, including the common device of treating as 'temporarily exogenous' an endogenous variable whose value is known, by deleting it from the set of endogenous variables for whose forecast values the model is solved and suppressing the corresponding structural equation. It is seen that this forecast can be adjusted to coincide with the optimal forecast. The initial discussion concerns the textbook linear simultaneous equation model; extensions to non-linear dynamic models are described.
KEY WORDS Macroeconomic models Forecasts Incomplete information Exogenization
In practical forecasting exercises with econometric models, there is not a clean break between the sample period for which data exist and the forecast period for which they do not. Rather the data have a 'ragged edge', because official statistical agencies produce data at different intervals and with different delays, so that at any time we have 'current' observations on some variables but not on others. Thus there is a gradual transition between the period for which complete data are available and the period for which none are available, and this paper is concerned with forecasting during this transition period. Since there is always some delay in the production of data, forecasting exercises begin by forecasting the present and the recent past, but we are concerned with the consequences of varying delays, and wherever the transition period is located, the practical question is how to make efficient use of the partial data that are available. It is a kind of missing data problem, but the difference is that the missing data will eventually be found, so that standard methods of forecast evaluation can be applied.
An econometric model-based forecast requires various data inputs-projections of exogenous
This paper is a revised version of ESRC Macroeconomic Modelling Bureau Discussion Paper No. 5 (March 1985), which was written during a visit to the