An economic forecast uses statistical methods to predict future behavior. The techniques are often applied in the context of monetary policy, setting state and local budgets, and financial management. Economic forecasting relies on historical empirical regularities and theoretical insights.
The concept of output most frequently forecasted is gross domestic product (GDP), which measures the monetary value of all finished goods and services produced within a country’s borders. GDP is measured by deflating nominal GDP to produce real GDP. A large literature has shown that nonlinear models may yield better forecasts compared to those using simple linear frameworks. Several studies have also shown that many of the largest forecast errors in some series occur around the beginning or end of recessions, suggesting that linear time series models fail to capture important nonlinearities.
A wide range of mathematical and statistical models are used to make economic forecasts. The methods include time-series analyses, model selection, and error estimation and correction. In some cases, models are trained on previous data to provide a probability distribution of future behavior, and the model’s output is then used to predict new data points. In other cases, a hypothesis is tested to find a relationship between current and past variables.
A majority of respondents say that conditions are worse now than six months ago, and the vast majority expect that they will get worse in the next half-year. While some respondents are a bit more optimistic than last December, overall, they are less likely to expect improving conditions globally or in their own countries.
About the Author
