Estimates of monthly GDP for the euro area : an application of the regression approach to interpolation

We discuss here a method for identifying relationships between high-frequency and lowfrequency data based on a dynamic regression technique. This allows users to estimate a quarterly analogue to an underlying monthly regression equation. The resulting equation which may be non-linear in the variables of interest, can be used to interpolate a variable which is observed at only low frequencies. Some practical aspects of implementation are discussed and the method is applied to produce estimates of monthly GDP for the Euro Area. Combined with appropriate forecasting techniques it can also be used to 'now-cast' quarterly GDP growth, producing estimates of quarterly growth about one month ahead of official data for both areas.