The Bank of Finland constructed a credit data system and launched a machine learning project to revamp the quality control of statistical data.

The Eurosystem's largest ongoing statistics project is the establishment of a common analytical credit database.

The Bank of Finland developed information systems necessary for the collection and processing of national credit data. Credit institutions reported their first granular credit data in October–November 2018. The Bank of Finland reported the data to the European Central Bank (ECB) in December 2018.

The Bank of Finland has contributed actively to the development of a pan-European Integrated Reporting Framework. Finland has long experience of combining several reporting frameworks into an integrated reporting framework. The scope of the data content to be incorporated was assessed in cooperation with banks in 2018.

The Bank of Finland launched a project for the development of payment statistics. The reform of payment statistics reflects the new Eurosystem requirements and national needs. Statistical reporting in accordance with the new framework will start in 2021, as a result of which we will obtain more detailed statistical information on the major changes taking place in the field of payments.

Machine learning enhances quality control

At present, the Bank of Finland already collects large volumes of granular securities and credit data. But we also need new methods and practices for ensuring the quality of the data and for processing the information for decision-making purposes.

The Bank of Finland launched a project that utilises machine learning. The quality control of statistical information is traditionally based on the rules defined by experts, whereas in machine learning, programs adjust themselves independently, based on the data. The purpose of using machine learning is to identify more effectively interdependencies between variables and deviations in the data. The results of the development project will be finalised in 2019.

The Statistics unit is actively analysing new areas in which to apply the experiences gained in the project.

Dynamic visualisation and data collection

The Bank of Finland launched for the users of banking statistics a new type of dynamic visual tool in November 2018. It provides new information on loans and deposits by showing more detailed information on the special features of the economy.

The Bank of Finland illustrates its statistics also with videos and animations and provides information on topical issues via an image carousel.

In 2018, the Bank of Finland launched a data collection targeted at crowdfunding entities and peer-to-peer lenders and published new data on the issue. Crowdfunding and peer-to-peer lending are not yet significant in volume in Finland, but are rapidly growing phenomena.

The Bank of Finland produced also for its website new information on private equity funds and started to publish information on credit institution market shares.

Share article