Economic forecast is the effort to predict future values of economic variables, based on patterns observed in their past behavior. These patterns can be either statistically identified (e.g., regression analysis) or intuited from their relations to other economic variables such as interest rates or consumer delinquency rates. This methodology usually requires an immense amount of acquired knowledge about the representation of those behavioral patterns and a substantial amount of faith that these represent enduring trends.
The prevailing assumption of economists and policymakers is that there are identifiable patterns in economic performance, and therefore that it is possible to manage the economy through targeted policies at appropriate times. In support of this view, there has long been a strong desire to develop methods for predicting the economy in real time. Consequently, there are many different approaches to economic forecasting. Some of the more popular approaches include estimating models that use historical data to predict future values, such as linear time series models or non-linear factor models; and using judgmental survey and model-based forecasts in combination to produce a forecast. Most countries and some regions have surveys of professional forecasters, for example the ECB Survey of Forecasters, which has been conducted since 1999. It provides forecasts from around 50 top forecasters from banks, manufacturing industries, brokerage firms and insurance companies. These forecasts are combined to create a consensus. Another set of real-time forecasts is the Greenbook, which is prepared as discussion material for each meeting of the Federal Open Market Committee.