The world of financial regulation & supervision has become hard-wired for technology innovations and data-driven initiatives, and there were plenty of them on display in Frankfurt at the 28th Eurofiling Conference hosted by the European Central Bank (ECB) last week. Much has changed in the 28 years since the 1st Eurofiling Conference was first initiated as an open meeting place for Regulators, Supervisors, and Entities from the public and private financial sectors.
Supervisors, both pan-European and national – in the European Union (EU), United Kingdom (UK), Israel, and others, modernize their data collection and analytics potential. The times when the usage of technologies like Artificial intelligence (AI), Machine Learning (ML), etc in supervision was the stuff of science fiction has passed and the digital era disrupts the supervisory sector around the globe.
Here are the key tangible takeaways distilled from two days of big thoughts and discussions during the Conference.
In March 2023, Twitter and the business news media exploded because of the Silicon Valley Bank, the second-biggest bank collapse in U.S. history that happened in just 48 hours. Due to the swift reaction of the US federal government, the domino effect was prevented but there are precious lessons left showing that banks should be prepared for much faster runs.
Now as the dust settled, the importance of more real-time data is obvious more than ever and we should take it seriously, Silke Stapel-Weber, Director-General of Statistics at the ECB, mentioned at the ‘Future of Digital Statistical Reporting’ session. Having quarterly and monthly data in place, U.S. regulators were not able to recognize or predict the patterns of the upcoming collapse and respond immediately. Getting back to the EU landscape, the mentioned example shows that real-time data and analytics, its further integration, and on-the-fly management is key to gaining up-to-date insight for efficient supervision, immediate actions, and getting closer to what Gartner analysts have termed “continuous intelligence“.
And rather than leaving it at a pat line, the EU banking supervision sector is continuing to evolve at the same pace as the banking industry by adopting cutting-edge data architecture and technologies, like AI, to enhance the assessment of supervised entities and of the risks across the European banking sector as a whole.
European Union (EU) banks face an extensive range of data reporting obligations, including statistical, resolution, and prudential information specified in various legal frameworks, including the sectoral module of Securities Holdings Statistics (SHS-S), granular credit and credit risk data (AnaCredit), implementing technical standards of the European Banking Authority (EBA), etc.
Being part of a larger European initiative, the Integrated Reporting Framework (IReF) came as an ambitious endeavor to redefine the collection of statistical data from banks in terms of both content and processes. It should be achieved by using a common data dictionary with all the data definitions, enhancing reusability and interoperability of the data, and consolidating all European System of Central Banks (ESCB) statistical requirements for banks in a unique legal act applicable across the euro area.
The source of the image.
Speaking about the IReF, Silke Stapel-Weber has also outlined the importance of cooperation with the regulatory authorities, banking industry, and other stakeholders to evaluate the impact and further enhance the reporting under IReF before its expected go-live in 2027.
As part of the recent developments, ECB updated the overview of the integrated reporting framework (IReF) and conducts a complementary cost-benefit assessment for the implementation of IReF between May 05, 2023 and July 31, 2023. The publication of the results of this assessment is expected toward the end of 2023.
Sharing the results of the analysis on Data Point Modelling (DPM) Refit conducted collaboratively by ECB, EBA, EC, and SRB, collectively called the Expert Group on DPM Refit (EG DPM Refit), Juan Alberto Sánchez, ECB, emphasized that the DPM Refit metamodel showed very good coverage of the IReF metamodel requirements and could serve for hosting it. Some workarounds and small improvements were easily implemented, and others were considered out of the scope of a Data dictionary metamodel, and what is important, none of these are considered a showstopper.
⇲ More about the DPM Refit initiative, recently published DPM standard 2.0 we wrote here.
The recent ISO decision to establish a Working Group, where Michal Piechocki, CEO of BR-AG is taking one of the leading roles, for ISO 5116 strategic review to further enhance and harmonize the DPM Method reflects a nice track of collaboration between involved authorities.
Underlying the importance of dialogue and joint work with the banking industry, Massimo Casa, Banca d’Italia, has also turned attention to the need for further investment in Banks’ Integrated Reporting Dictionary (BIRD), one of the pillars of the ESCB strategy for standardizing banks’ reporting, to reduce the room for interpretation of reporting regulations. It plays a crucial role in maintaining a close connection between the data production of the reporting agents and the data collection of the authorities.
Collaboration is useless without fair dialogue between parties. ‘Good, but not perfect’ – addressing the audience with this message, Noa Granot Reich, Bank of Israel (BoI), highlighted the complexity of change, frequent revisions in the model, the difficulty of connecting DP between versions, lack of tools for data integration and analysis of ever-larger volumes of data reported (XBRL and non-XBRL) are among the main challenges that BoI is struggling with nowadays.
The source of the image.
Responding to the real-time challenges, the ‘experiment phase’ of SupTech usage in the BoI and other Central Banks is quickly transitioning to a ‘new normal’. Designing their SupTech Data Strategy, the national regulator is paying attention to the delineation of the current situation, analysing the steps taken in supervisory entities worldwide, mapping gaps, deciphering potential moves and their prioritization.
⇲ Earlier we wrote more about why SupTech Strategies are gaining traction and the impacts they bring here.
The place of AI at the table of SupTech tools for banking supervision becomes more and more irreplaceable. Understanding this reality, Barend de Beer, South African Reserve Bank (SARB), highlighted the vital meaning of AI algorithms for supervisory and prudential statistics in a way of identifying patterns that humans fail to spot and thereby enhancing the quality of supervision. Additionally, it has the ability to enhance the agility of supervision by promptly notifying supervisors of any anomalies in real time.
Understandably, AI is always linked to data protection and requires a regulator’s ‘privacy-respecting’ mindset regulated by strict laws.
At BR-AG, what truly matters to us is to reimagine the complex, intricate, and time-consuming processes and transform them into the opposite ones. Together with other global tech leaders, we were delighted to present our ATOME Platform to an audience of regulators, showing how it addresses their challenges in an agile and scalable manner.
New supervisory needs and regulatory changes affect data collection and processing each year. Just in the EU banking sector, there are three new frameworks for banks to introduce more than 3 000 new validation rules. Coupled with time-consuming administrative activities and the manual efforts required – the whole change management can be a daunting task for both supervisors and reporting entities.
Pursuing demand for connectivity and ensuring trusted data, the ATOME Platform was designed to help collect, validate, standardise, analyse, and understand the data, making it accessible and interoperable.
Download the presentation, to take a closer look at the ATOME ecosystem.
The source of the image.
If you have questions or comments on any of the topics discussed this year or would like to get a comprehensive ATOME demo – reach out to us:
