National Accounts in a world of Naturally Occurring Data
Published on Tuesday, October 4, 2022 | Updated on Tuesday, October 4, 2022
Big Data techniques used
National Accounts in a world of Naturally Occurring Data
Summary
Presentation for the Conference on Non-traditional Data, Machine Learning and Natural Language Processing in Macroeconomics hosted by Sveriges Riksbank and jointly organized with the Federal Reserve Board, the Bank of Italy, and the Bank of Canada.
Key points
- Key points:
- This paper provides the first proof of concept that naturally occurring transaction data, can be harnessed to produce high quality consumption survey.
- A simple aggregation of the data can result in good proxies of National Accounts levels and growth.
- The real time and high-definition components of the data show a powerful economic tool to track the economy in real time not only at aggregate levels but also zooming out across characteristics, gender, income categories and geographical details.
- The combination of this information constitutes the basis for better diagnoses and furthermore for the design of smart policies.
Geographies
- Geography Tags
- Spain
Topics
- Topic Tags
- Banks
- Consumption
Authors
Gergely Buda
Vasco M. Carvalho
University of Cambridge - External partner
Stephen Hansen
University College London - External partner
Alvaro Ortiz
BBVA Research - Head of Analysis with Big Data
Tomasa Rodrigo
BBVA Research - Lead Economist
José V. Rodríguez Mora
CUNEF - External partner
Documents and files
Presentation (PDF)
Naturally-Ocurring_Riksbank-Conference_October-2022-.pdf
English - October 4, 2022