Economic forecast are a vital input into the decision making of central banks and governments as well as private sector agents. They can be used to estimate the economic impact of a policy change, to determine if an investment will pay off, or to gauge how fast a recession is likely to hit. A wide range of methods are used to produce economic forecasts, from simple judgmental approaches that rely on the expert’s judgement to adjust a suite of models to more sophisticated dynamic stochastic general equilibrium (DSGE) models that use modern economic theory.
The global economy is expected to remain subdued, reflecting a slowdown in the world’s major economies, as rising trade barriers sap growth and stoke inflationary pressures. In the Middle East and North Africa, a gradual recovery is projected, reflecting an assumed stabilization of armed conflicts, higher oil production and higher exports of services and commodities. However, the outlook is precarious: a re-escalation of global trade tensions, higher oil prices or tighter financial conditions could stifle activity.
The complexities of assessing forecasting accuracy are even more pronounced when the evaluation involves simultaneous predictions of several different economic variables. A given model may incorporate assumptions about the value of various variables—such as gross domestic product, one or more interest rates, employment and price inflation—and the relative quantitative accuracy of competing forecasts is measured in terms of how well each individual prediction performs. Yet these relationships are typically mutually conditioned, so that a method that achieves a high degree of predictive accuracy in predicting one variable might seem to be inaccurate or unreliable when applied to the prediction of another.