An economic forecast is a prediction of the performance of a nation’s economy. While the term economic forecast can encompass a wide variety of predictions, the specific economic variables that are most often predicted are real GDP and other indicators of national income and expenditures (such as exports, imports, government expenditures). The methods used to predict these parameters may vary, but almost all economic forecasts utilize some form of macroeconometric model.
In most cases, the prediction of these variables is based on a combination of theory and judgment. Generally, economic theory provides the broad outline of what is expected to happen and judgment determines how that theory should be applied in particular circumstances. For example, the forecaster might decide that a situation is unique enough that it requires a deviation from the usual statistical time-series method.
A variety of models are used to generate economic forecasts and these have been extensively studied in recent times. The most common are factor models and multivariate nonlinear models. Both types have been shown to perform well, though the results from a single model or type of model cannot be assumed to apply to all situations. For this reason, many forecasters use a combination of models and judgmental survey-based forecasts to produce their economic forecast. These are known as combination forecasts and the evidence supporting their efficacy is growing. See, for example, Clark and McCracken (2010), Assenmacher-Wesche and Pesaran (2008), Kuzin et al. (2013), and Aiolfi et al. (2011).
About the Author
