Big data and machine learning in central banking

BIS Working Papers  |  No 930  | 
04 March 2021

Summary

Focus

The world is changing and so is the way it is measured. For decades, policymakers and the private sector have relied on data released by official statistical institutions to assess the state of the economy. Collecting these data requires substantial effort and publication often happens following a lag of several months, even years. However, the last years have seen explosive growth in the amount of readily available data, as well as in the technology and software used to analyse it. These developments have spurred central banks' interest in big data and machine learning.

Contribution

This paper provides an overview on the use of big data and machine learning in the central bank community. It leverages on a survey conducted in 2020 among the members of the Irving Fischer Committee. The survey contains responses from 52 central banks from all regions of the world and examines how they define and use big data, as well as which opportunities and challenges they see.

Findings

The analysis highlights four main insights. First, central banks define big data in an encompassing way that includes unstructured non-traditional as well as structured data sets. Second, central banks' interest in big data and machine learning has markedly increased over the last years: around 80% of central banks discuss the topic of big data formally within their institution, up from 30% in 2015. Third, the vast majority of central banks are now conducting projects that involve big data. Institutions use big data and machine learning for economic research, in the areas of financial stability and monetary policy, as well as for suptech and regtech applications. And fourth, the advent of big data poses new challenges, among them data quality, legal aspects around privacy, algorithmic fairness and confidentiality, as well as budget constraints. Cooperation among public authorities could relax the constraints on collecting, storing and analysing big data.


Abstract

This paper reviews the use of big data and machine learning in central banking, leveraging on a recent survey conducted among the members of the Irving Fischer Committee (IFC). The majority of central banks discuss the topic of big data formally within their institution. Big data is used with machine learning applications in a variety of areas, including research, monetary policy and financial stability. Central banks also report using big data for supervision and regulation (suptech and regtech applications). Data quality, sampling and representativeness are major challenges for central banks, and so is legal uncertainty around data privacy and confidentiality. Several institutions report constraints in setting up an adequate IT infrastructure and in developing the necessary human capital. Cooperation among public authorities could improve central banks' ability to collect, store and analyse big data.

JEL codes: G17, G18, G23, G32

Keywords: big data, central banks, machine learning, artificial intelligence, data science