On the use of high-frequency economic information to anticipate the current quarter GDP: A study case for Mexico

摘要

A method for anticipating quarterly estimates of GDP and its main components in advance of the official publication is presented in this study. Using high-frequency information, time series equations and regression analysis, the model predicts the quarterly GDP based on three different approaches: production side, expenditure side, and principal components extracted from a set of strategic indicators. This model, considered a purely econometric system with no personal data adjustment, allows us to forecast the current quarter GDP and its price deflator using monthly information on economic activity, financial markets, futures prices and forwards, and expectations.