Export Nowcasting with AIS and Foot Traffic Data

In recent years, nowcasting using alternative data has become increasingly active, however, there are relatively few cases focusing on individual companies or combining high-frequency, fine-grained multiple alternative data sources. In this paper, we performed nowcasting of the export volume of major Japanese automakers by utilizing AIS data from export ports and foot traffic data from assembly plants. We demonstrate that using alternative data enables the capture of export trends more quickly and accurately than existing statistics. Furthermore, we show that combining foot traffic data with AIS data tends to improve the accuracy of extrapolative predictions, highlighting the advantages of combining multiple alternative data sources. This paper takes advantage of the fine-grained nature of alternative data by focusing on individual companies’ exports and reveals new possibilities in export forecasting by simultaneously utilizing AIS data and foot traffic data.