Statistics to deliver price stability and mitigate systemic risk

Speech by Mario Draghi, President of the ECB
at the Sixth ECB Statistics Conference, Frankfurt am Main, 17 April 2012

Ladies and gentlemen,

on behalf of the Executive Board I would like to welcome you to our sixth biennial European Central Bank (ECB) statistics conference. I am very pleased to see that the event has once more attracted many distinguished participants.

Let me take the opportunity to share with you some thoughts on the topics on the agenda.

As policy-makers in the fields of monetary policy and financial stability, we are deeply dependent on the work being done to maintain a stream of timely, relevant and reliable data.

Today and tomorrow you will have the opportunity to discuss ways of further strengthening the information basis for the fulfilment of the mandates given to the ECB and the European Systemic Risk Board (ESRB).

Key challenges

The financial crisis has dramatically increased the need for very timely granular data. And it has led to a rethinking of a number of organisational and conceptual aspects in statistics.

To give just one example, the traditional monetary statistics derived from individual banks’ balance sheets are insufficient to gather information on financial groups that operate across borders and through non-bank subsidiaries. This is because the monetary statistics cover only the euro area and do not focus on the various kinds of risk to which banks are exposed and the adequacy of their capital in view of these risks. This means that it is vital to collect data consolidated at the level of banking groups or insurance groups.

Some of these data are already available at the national level for micro-supervisory purposes, but they are not always comparable across countries. The ECB, the ESRB and the three new European supervisory authorities (ESAs) are all working to fill remaining data gaps and to meet the challenge of timely, frequent and high-quality data for their macro-prudential functions.

Moreover, there is an increasing need to consider measures of dispersion not just averages, and to consider micro data as well as macro aggregates. For financial stability purposes, in particular, a wider variety of granular data and indicators are needed, and more emphasis needs to be put on data “outliers” and so-called “tail risks”.

As we have learned, systemic risk may stem from any part of the financial system, including the shadow banking sector. That is why we also need more data to monitor unregulated financial institutions and markets.

In recent years, significant steps have been taken to identify the data gaps and to fill them by gathering more and more detailed quantitative information. But we are just at the beginning of this process, and I know that statisticians are all too well aware of the fact that the data perimeter may change again in the future.

It is vital that the work done at the European Union level is consistent with the worldwide dimension. In 2009, the G20 endorsed a report on “The Financial Crisis and Information Gaps”. That report contains 20 high-level recommendations on the measurement of risks in the financial sector, international financial linkages and the communication of official statistics.

Quite some progress has been made in the implementation of these recommendations. For example, accounts for economic sectors, statistics on securities, and the Principal Global Indicators website supported by seven international organisations, including the ECB. But it is still necessary to continue with these efforts, in particular on quarterly government finance statistics and on the FSB common template for systemically important global financial institutions. And the transposition of the new rules into all national legal systems needs to proceed swiftly.

The role of statisticians

Going beyond the pure data challenges, the statistical community is still coming to grips with the fact that their work is now more than ever in the spotlight of public debate. “Statistical errors” were once something in which only specialists were really interested. Now, they can receive headline coverage, sometimes even more so than everything else coming out of both official and non-official statistical sources.

Public scrutiny is good for statistics but it also brings new challenges for statisticians. Alongside their more traditional role of collecting data and assisting in the correct interpretation of data, statisticians also need now to be more involved in the communication of results – to the markets, to the media and to the wider public. The role of dashboards and the narrative use of data are examples of tools where the input of statisticians is clearly essential.

Selected ECB statistical initiatives to fill data gaps

Together with the rest of the Eurosystem, the ECB is undertaking a number of important statistical initiatives to support fulfilment of the ECB’s mandate to deliver price stability and the ESRB’s mandate to mitigate systemic risk. The initiatives will also help other policy-makers and decision-makers.

Let me briefly mention three of these initiatives:

First, the Eurosystem is developing a Securities Holdings Statistics database. The financial crisis has highlighted the lack of transparency in a number of financial markets as well as difficulties in assessing financial exposures and linkages between institutions. Securities holdings represent a field where exposures are often concentrated, and a lack of sufficiently comprehensive, consistent and granular information has been identified.

The aim of developing a detailed dataset on securities holdings is to combine holdings of large players in the euro area with information on the individual issuers across the world. This will provide a potentially very useful tool for both monetary policy and macro-prudential analysis.

Second, the Eurosystem has been extending its database to provide more detailed information on the activities of financial institutions and instruments. This covers more detailed balance sheet information on banks but also on investment funds, insurance corporations and pension funds. Complex securitisation operations, which played a major role in the financial crisis, are reflected in new data on the activity of financial vehicles resident in the euro area.

Information on who is lending to whom in the euro area economy can help us to understand the process of financial intermediation better. It is now published on a quarterly basis in the matrix of the integrated sector accounts. The provision of quarterly information on cross-border holdings of financial assets and liabilities by euro area residents allows users to detect cross-border exposures.

Third, the Eurosystem is amending its statistical legal acts to align its economic and financial statistics with the recently revised international statistical standards. These revised standards aim at achieving consistency between domestic sector accounts and statistics for the rest of the world. This consistency will give users a much more complete view of inter-linkages between the rest of the world and the domestic sectors of the economy.

Reporting burdens

The ECB is well aware that closing information gaps requires significant efforts by all parties involved – the statisticians, the reporting agents and the data users. It is also clear now more than ever that a lack of good quality information on which to base decisions can be far more costly to society than the collection of new, additional data.

In any case, all data collections by the Eurosystem are subject to a prior, strict “merits and cost” procedure. Importantly too, there is a strong emphasis on ensuring that there are no double collections of the same information by different statistical authorities and that, as far as possible, existing information serves multiple purposes.

These data collections must be detailed and comprehensive enough to meet various current and potentially future statistical needs. When banks or other agents report the raw data, the compilation of aggregates and distribution measures for monetary policy, macro-prudential and other statistical purposes can be done by the respective statisticians and be used for several purposes. The reporting agents thus can be spared these compilation exercises – saving them resources and cost.

Of course, this requires effective coordination between the various statistical authorities to ensure their data needs are matched. As an example, the ECB is working with the ESRB and the ESAs aiming to achieve the highest possible coordination and synergies with regard to data for supervisory purposes, macro-prudential analysis and monetary policy.

A good example of this is the recent collaboration of the European System of Central Banks and the European Banking Authority (EBA) through the Joint Expert Group on Reconciliation of credit institutions’ statistical and supervisory reporting requirements. The second edition of the group’s report has just been published on the websites of the ECB and the EBA.

The problems faced in such initiatives are often very complex, and progress takes time. For example, as supervisory data on consolidated group structures are currently not fully harmonised within the European Union, efforts are needed to make these data comparable across countries and time. The quality and timeliness of such data requires efficient data transmission and processing between different organisations.

Confidentiality

One of the thorniest issues in such coordinated activities concerns the confidentiality of individual data, and supervisory data in particular.

The protection of confidentiality is sacrosanct for all statistical activities. It is enshrined in the relevant legal texts of the ECB and it is central in standard practices for all official statistics. Where “Chinese walls” need to exist, they are strictly respected. Reporting agents should know and should have the right to know where the confidential information they are providing is to be used.

But a big effort is needed to ensure the harmonisation of divergent practices in the protection of data and sometimes conflicting legal texts in the different fields of competence and in different countries.

Restrictive “blanket” confidentiality rules that cover all types of information, irrespective of whether there are real risks of a breach of confidentiality, are often counterproductive. They end up not only being damaging to information sharing and disclosure, but also to the real protection of confidentiality where and when this is needed. These rules confuse what is truly confidential and what needs strong protection with what is possibly market sensitive or what is simply inconvenient.

So here is yet another area of activity on which statisticians and legal experts need to focus. They need to ensure that legal frameworks and existing practices find the right balance between the need for confidentiality and the possibility of data sharing.

Conclusion

Let me conclude. As I hope all of my remarks have illustrated, statistics is a core activity of policy-making, and central banking in particular. Right now, it is also a highly dynamic activity.

I personally attach great importance to progress in the areas I have mentioned. I am convinced that your discussions today and tomorrow will offer valuable contributions on how best to face the challenges and how to take advantage of the opportunities.

Producing high-quality statistics on the financial sector remains a core task of a central bank. It is a central input not only for decision-making processes but also for the communication of our decisions, and thus, for the credibility of our actions.

In light of these challenges, the central bank statisticians of the future will have to serve both areas – the pursuit of price stability as well as the support of financial stability.

I wish you a very fruitful conference and very much look forward to the results of your deliberations.

Thank you for your attention.