Statistics as an indispensable signpost
Speech by Danièle Nouy, Chair of the Supervisory Board of the ECB, Ninth ECB Statistics Conference, "20 years of ESCB statistics: What's next?", Frankfurt, 11 July 2018
In the early days of statistics, there were some who warned that “statistics are no substitute for judgement”. Well, then as now, statistics might not be a substitute, but they are an indispensable signpost.
We need a signpost because our judgement is often biased. Think of all the people who are afraid of flying and prefer to drive instead. In reality, travelling by road puts you in much greater danger than travelling by air. The statistics are very clear on this. So people who read this signpost are better able to judge the risks. And this is just one of many examples of biased judgement.
We banking supervisors are aware of these biases, of course. That’s why we value good data so highly and rely on them so much. Good data help us to rein in errors that might arise when exercising our judgement. Without good data and sound statistics, we would not be able to do our job.
Statistics and banking supervision
We rely on supervisory data in all aspects of our daily work. We rely on them to create risk profiles of banks and we rely on them to analyse a bank’s liquidity and its capital. Data underpin our judgement, our decisions and our actions.
So it was vital to have good data from day one of European banking supervision. And thanks to all the excellent work done in advance, we did. This was one of the benefits of being part of the ECB and having access to the shared services, including statistics.
But data on banks are not only of interest for supervisors. Many people would like to gain deeper insights. However, data on individual banks are mostly very sensitive, and for good reason. This is why we always maintain the highest standards of data confidentiality. But we can, and do, aggregate data and share them in this high-level format. As an additional service to the general public and to the banks themselves, since 2016 ECB Banking Supervision has been publishing extensive supervisory banking statistics on its website. These include some Pillar 3 indicators, meaning the disclosure requirements in the Basel Committee on Banking Supervision (BCBS) framework aimed at promoting market discipline. The ECB Annual Report on supervisory activities also presents the most important statistics for the periods under review.
And the banks themselves also need good data. We always emphasise this when discussing governance and risk management in banks. Good data help banks to take good decisions; bad data leave them with blind spots, which might prompt bad decisions. Against that backdrop, many banks still need to improve their data frameworks. The BCBS Principles for effective risk data aggregation and risk reporting – BCBS 239 for the experts – provide relevant guidance.
In this regard, the data we collect from banks gives us deeper insights; it takes us to another level, if you will. Because the quality of the data we receive reveals much about the banks’ ability to comply with BCBS 239. And this, in turn, reveals much about the bank’s ability to manage its risks. Poor data are a poor basis for managing risks.
Indeed, the quality of the data as such opens a window into the state of a bank. We even use a data quality index and data quality dashboards to understand the weaknesses in a bank’s reporting and draw our conclusions. After all, assessing the quality of data provided by banks is part of our Supervisory Review and Evaluation Process.
All in all, there is no doubt that, for many reasons and from many angles, data help us to do our job well. That said, we have to think about the future. There are challenges ahead, and there are things we still need to work on. Let us take a look at some of them.
What will be relevant in the future?
First of all, the world does not stand still. The business of banking is constantly evolving: new technologies are invented, new products are developed, new competitors appear and new risks emerge. And as the world changes, our data needs change too. At the same time, technology helps us to enhance our analyses. Big data, for instance, can help us to improve our risk assessment of banks. Statistics departments need to remain flexible to be able to cater for these changes and stay ahead of the curve.
Another point is the indisputable burden that reporting puts on banks. We acknowledge that banks have to carry this burden and we try to alleviate it through various initiatives. Three relevant actions here are to standardise, integrate and harmonise reporting.
For standardisation, one goal should be to have unique identifiers for banks and their counterparts. This would help to further consolidate reporting and lighten the burden on banks. The Legal Entity Identifier, LEI for short, has been a good first step, and banks are already benefiting from applying it.
But we also must work to integrate and harmonise reporting. At the European level, banks still have to deal with a huge number of reporting obligations. Data are collected through different reports for both supervisory and statistical purposes, based on different legal acts. At the same time, reporting obligations differ across countries, making life for cross-border banks rather difficult, as you can imagine. As we aim for a truly European banking market, this is something we need to address.
And European banking supervision is doing just that. We will, for instance, set up a database containing all the recurring requests for supervisory data which are addressed to banks under our direct supervision. This database will allow us to monitor the overall reporting burden on these banks. It will also allow us to provide banks with a preliminary overview of data requests for the year ahead. And it will allow us to present the ECB’s Supervisory Board with an annual dashboard showing all bank-specific and horizontal data requests. All this will help us to streamline and harmonise the data requests and thus simplify reporting for the banks.
And that’s not all. There is also the BIRD, the Banks’ Integrated Reporting Dictionary. The BIRD is a voluntary tool that helps banks to organise their data by providing a common data dictionary and common transformation rules. This makes it much easier for software companies and banks to implement reporting requirements in a uniform manner. And it helps banks to report data to the required level of granularity.
From an institutional point of view, it might help to formalise and expand the role of the European Banking Authority, the EBA, in the field of supervisory data. The EBA could, for instance, integrate supervisory reporting and disclosures under Pillar 3 of the Basel Accords. It could then develop and maintain a relevant hub of bank data collected from regulatory reporting.
This would help banks as they would only need to report the respective data once. And it would help to set the right incentives for improving data quality; this is one of the insights from the EBA’s transparency exercises. It would also help supervisors and market participants because they would have easier access to the data. The United States could serve as a role model here.
Ladies and gentlemen,
Data are at the heart of banking supervision. To do our job, we supervisors need high-quality data, and we need them in good time. The statistical departments throughout the Eurosystem ensure that these needs are fulfilled. They have done so for the past 20 years – or for the past four years in the case of banking supervision.
In a sense, statisticians are the silent heroes of the story. They tend to work backstage, but without them nothing would happen on stage – nothing good, at least. So, to all of you who work on collecting and providing the data we need: thank you!
Yours is a difficult job. Technology keeps changing, and that has an impact on how data are collected, analysed and used. At the same time, data need to keep changing, and you have to adapt to that as well. This conference marked the 20th birthday of Eurosystem statistics by looking to the future. It shed light on all the new developments and how to deal with them. I am sure that you will be well able to handle them efficiently and will continue to deliver high-quality data for the decades to come.
Thank you for your attention.