Claudia Buch: Welcome speech at the G20 Workshop on Data Sharing

Welcome speech by Prof Claudia Buch, Deputy President of the Deutsche Bundesbank, at the G20 Workshop on Data Sharing, Frankfurt am Main, 31 January 2017.

Ladies and Gentlemen

A warm welcome to all of you participating in the G20 Workshop on Data Sharing. The workshop is part of the second phase of the Data Gaps Initiative and co-organised by the IMF, Eurostat and the Deutsche Bundesbank.

I am very happy to welcome participants from G20 and non-G20 countries, from international organisations, central banks, statistical offices and finance ministries. Many of you have travelled long distances, which shows how important data sharing is for our analytical work. 

As you know, Germany fully supports the second phase of the G20 Data Gaps Initiative (DGI-2) and will fulfil all recommendations within the agreed timeframe (FSB and IMF 2016). The initiative provides key statistical inputs for analysing the economy and financial stability. The initiative is well on the way to realising its objectives within the envisaged five-year horizon. This has been achieved through excellent cooperation between international agencies and G20 economies, implementation of thematic workshops - like today's - and concrete timetables.

This workshop's topic is Recommendation #20 on promoting data sharing. We believe this recommendation has a lot of potential. Aggregate datasets provide only limited information on the drivers and effects of changing structures in the real economy and the financial sector. The recommendation is therefore explicitly concerned with sharing of granular data. This is necessary to understand global developments and in particular differences across countries.

Take the decline in cross-border banking activities since the global financial crisis as an example. This development raises many questions:

  • Which banks have withdrawn from foreign markets?
  • Has the withdrawal been driven by regulation, by macroeconomic factors or by bank-specific factors?
  • Are these patterns similar across countries?
  • What have been the effects on bank risk and lending?

These questions cannot be answered without combining datasets from different sources:

  • Bank income statements and balance sheets
  • Reports on external positions
  • International data

Combining datasets, looking beyond aggregate statistics into heterogeneous developments - all this is required to transform "data" into "knowledge". Yet, legal constraints make it difficult, or often practically impossible, to link micro datasets from different jurisdictions, even for research and financial stability analysis. Better accessibility and sharing of granular data would open up new possibilities for analysis by providing new insights into the effects of policies.

Let me give you an example. In order to analyse issues pertaining to global banks with granular data, the International Banking Research Network was founded. This network brings together researchers from around the world. They use micro-level data to better understand the consequences of global banking: country teams work on the same topic in parallel, using their distinct, bank-level data sets to help develop comparable cross-country evidence. Although the underlying regulatory data cannot be made public, the results of the common analysis can be shared to inform the policy dialogue.

One example is the latest research effort on the impact of changes in prudential policy instruments on lending of global banks (Buch/Goldberg 2016 and Buch/Bussière/Goldberg 2016). This project uses a new database that focuses on changes in the intensity of the usage of several widely used prudential tools, taking into account both macroprudential and microprudential objectives. It finds that international spillovers vary across prudential instruments and are heterogeneous across banks. Bank-specific factors like balance sheet conditions and business models drive, both the amplitude and direction of spillovers to lending growth rates. Spillovers into lending growth from prudential tightening are positive in some cases, and negative in others. Spillovers are most likely to occur via the affiliates of foreign banks hosted in a country, although some evidence exists of inward transmission through home-country global banks. This would be in line with an attempt of domestic banks to preserve market shares and to protect the domestic market from international policy spillovers. 

This example shows: The availability and flexible use of granular data early on is an integral part of any structured process of policy evaluation. Other examples are easy to find: one can think of information on channels of transmission of monetary policy across different institutions, sectoral and firm-level drivers of productivity, the build-up of risks to financial stability and the causal impact of regulatory and monetary policies. 

To put it in a nutshell, we need to implement a new paradigm in statistics, moving away from traditional, aggregate statistics towards the flexible use of (standardised) micro data. Micro data are collected only once and can be used for multiple purposes, which has the added advantage of a lower statistical reporting burden (Deutsche Bundesbank 2016). In order to develop this benefit immediately, we need a close cooperation between bankers and statisticians.

We, therefore, welcome the work done so far. Let me name just a few milestones that have been reached over the last year in the G20 context.

  • After the regional conferences, a Global Conference of the G20 Data Gaps Initiative was held in Basel in June 2016. This conference agreed on concrete action plans for implementing the DGI-2 recommendations.
  • The decision was taken to establish an informal G20 working group on data sharing, chaired by the IMF and Eurostat in cooperation with the Deutsche Bundesbank.
  • Building on this, a meeting was held in Washington D.C. in November of last year to take stock of G20 countries' experiences. First ideas were developed for a common terminology for granular and micro data, and suggestions were made on how to balance confidentiality and user needs in times of the internet.
  •  All these elements built the basis for the draft recommendations on data sharing and accessibility, which will be discussed in detail at this workshop. 

This workshop offers the possibility for an in-depth discussion of the draft recommendations, together with best practices. The final version should be agreed by the DGI contact group and the IAG by the end of February 2017. This document will serve as a deliverable for the meeting of the G20 Finance Ministers and Central Bank Governors in March 2017 to be reflected in the Communiqué. Furthermore, the workshop should debate the next concrete steps to be taken after adoption of this Communiqué.

I wish you fruitful and stimulating discussions and every success developing your ideas. 


Buch, C M, and L S Goldberg (2016), "Cross-Border Prudential Policy Spillovers: How Much? How Important? Evidence from the International Banking Research Network", NBER Working Paper no. 22874.

Buch, C M, M Bussière, and L S Goldberg (2016), "Prudential policies crossing borders: Evidence from the International Banking Research Network", VOX, CEPR's Policy Portal, 09 December 2016.

Deutsche Bundesbank (2016), Microdata - paradigm shift in central banks' statistics, Annual Report 2015, Frankfurt a.M., pp 47-59, February 2016.

Financial Stability Board (FSB) and International Monetary Fund (IMF) (2016), The Financial Crisis and Information Gaps, Second Phase of the G-20 Data Gaps Initiative (DGI-2) - First Progress Report, September 2016.