The Usefulness of High-Frequency Alternative Data to Obtain Nowcasts for Japan’s GDP: Evidence from Credit Card Data

High-frequency alternative data, with their nearly real-time availability, provide timely information to track economic activity in real time. This study extends the standard dynamic factor model-based GDP nowcasting for Japan by introducing credit card data and examines how credit card data can be effectively employed in real-time nowcasting. The empirical results of a simulated real-time exercise suggest that credit card data, which accurately capture consumer spending on services, the area that was most heavily affected by the Covid-19 pandemic in Japan, in real time, carry important information and therefore improve the performance of nowcasting relative to nowcasts based on traditional monthly data only, especially at the early stage of forecasting when traditional data, which are published with a significant lag, are not yet available for the relevant quarter.