966
Serial number: 01/2022
Author(s): Ilkin Huseynov, Nazrin Ramazanova, Hikmat Valirzayev
Language: English
Date: 2022
Abstract: This study examines whether payment systems data can be useful in tracking economic activity in Azerbaijan. We utilize transactional payment systems data at the sectoral level and we employ Dynamic Factor Model (DFM) and Machine Learning (ML) techniques to nowcast quarter-over-quarter and year-over-year nominal GDP. We compare nowcasting performance of these models against the benchmark model in terms of out-of-sample RMSE at three different horizons during the quarter. The results suggest that payment system data along with ML and DFM models have higher predictability than benchmark model and can lower nowcast errors significantly. Although our payment time series are still too short to obtain statistically robust results, the findings indicate that variables at a higher frequency such payments systems data can be helpful to assess the current state of the economy and have the potential to provide a faster estimate of the economic activity.
Key words: payment data, nowcasting, ML, DFM
JEL classification: C32, C38, C52, C53, E42
