Central banks increasingly need to use business intelligence (BI) systems to collect, manage and analyze data in order to inform policy decisions. This report presents the results of a survey conducted by the BIS’s Irving Fisher Committee on Central Bank Statistics (IFC) in 2019.
Its main findings can be summarized as follows:
- BI has become an integral element supporting the statistical function in the vast majority of central banks, especially to analyze, present and integrate statistical data. Other use cases include data elaboration, big data analytics and statistical quality processes.
- Reflecting their various needs, the range of BI tools central banks use is quite large. The main categories are data visualization instruments, design report instruments, developed analytic tools and online analytical processing (OLAP) technology, and analytic and strategic dashboards. In practice, central banks are using multiple BI tools that complement each other. The degree of satisfaction with the use of BI tools is quite high.
- The related technology stack comprises packaged solutions available “off the shelf”, with a strong preference expressed for the business analytics service provided by Microsoft and SAS. In addition, central banks also rely on the extensive use of statistical packages/programming languages for conducting specific BI work: the most used are MATLAB, Stata and R – with a high demand for Python in data science and big data-related projects.
- Visualization is an important issue when making sense of data, and is particularly key for central banks’ communication. From this perspective, a large majority of central banks have developed an internet-based dissemination strategy; only one quarter of central banks rely exclusively on printed documents for dissemination/publication purposes.
- Internet-based data dissemination offers users the possibility to interact with central banks’ publications in a dynamic/interactive way. As a result, two thirds of central banks use a mix of both static and dynamic graphical content when presenting information on their websites.
- JavaScript Graphing Libraries are the most popular means for generating graphical content for web publication. Other tools, such as the spreadsheet and business analytics services provided by Microsoft, are also frequently used to create graphical content. In addition, a significant number of central banks have developed internal, “customized” solutions for their graph production.