Macro-prudential supervision in Europe
Speech by Vítor Constâncio, Vice-President of the ECB, at the ECB-CFS-CEPR conference on “Macro-prudential regulation as an approach to containing systemic risk: economic foundations, diagnostic tools and policy instruments” at Frankfurt am Main, 27 September 2010
Ladies and gentlemen,
It is a great pleasure to host this conference dinner on “Macro-prudential regulation as an approach to containing systemic risk”. I would like to thank the organisers for this very important and timely event, namely Xavier Freixas (University of Pompeu Fabra) for the CEPR, Jan-Pieter Krahnen (University of Frankfurt) for the Center for Financial Studies and Philipp Hartmann and Stefano Corradin from the ECB.
As the President recalled this morning in his opening remarks, the financial crisis has revealed how the financial and economic community has neglected a system-wide perspective on financial stability. A purely micro-prudential approach to financial stability ignores the fact that the joint actions of individual intermediaries can have an adverse effect on the system as a whole. A simple micro-prudential approach to financial supervision is therefore subject to a “fallacy of composition” as the soundness of an individual institution cannot be equated to the soundness of the whole financial system.
A decisive step towards a truly macro-prudential setup is needed. Indeed, this is under way, not only in Europe, but also in other constituencies around the globe. Tonight I would like to discuss three key issues pertaining to the conduct of macro-prudential policy in the European Union (EU): (i) the role of the European Systemic Risk Board (ESRB) and the European Central Bank (ECB) in macro-prudential supervision; (ii) the role of research and analytical tools in the conduct of macro-prudential supervision; and (iii) the efforts of the European System of Central Banks (ESCB) to undertake and stimulate further research in support of macro-prudential policies.
I. The ESRB and the role of the ECB in macro-prudential supervision
Within a relatively short time-frame the EU has agreed on a broad reform of financial supervision and established a new body for macro-prudential oversight. Political breakthrough has been achieved this month: both the European Parliament and the European Council agreed on the new EU regulations establishing the ESRB, entrusting the ECB with the support of the ESRB, as well as the three European Supervisory Authorities (ESAs), to supervise the financial system at large. This means that the new institutions can start operating in January 2011.
Under the new framework, the ESRB will be responsible for the macro-prudential oversight of the financial system within the EU, while the ESAs will be assigned micro-prudential responsibilities. The ECB will play a role in the new framework by providing to the ESRB analytical, statistical, administrative and logistical support. National central banks and supervisors will also provide technical advice, which will constitute an important input into the work of the ESRB. The ECB is already preparing to support the work of the ESRB. First, a Preparatory Secretariat has recently been established to help develop the ESRB steering process, as well as the policy framework for the preparation of warnings and recommendations. The Preparatory Secretariat is also making organisational and logistical arrangements for the ESRB and its sub-structures. Second, the ECB is enhancing its work on financial stability and systemic risk in order to prepare for the future analytical and statistical support it will provide to the ESRB.
The ECB will not be alone in supporting the ESRB. Other important member institutions of the ESRB include EU national central banks, the three new ESAs and the European Commission. By nature, the ESRB will be a collective, independent body. Each member institution will have the crucial task of contributing – from their respective perspective – to the assessment of risks which could propagate through the EU in times of crisis.
The ESRB has three main objectives. First, it should monitor, identify and prioritise systemic risks to financial stability. The development of a suitable framework of analysis is key to the achievement of this task. To this end, the ESRB will: (i) engage in a dialogue with the financial community in order to gather market intelligence; (ii) maintain contacts with other partners outside of Europe (such as the Financial Stability Oversight Council, which has been set up in the meantime in the United States); and (iii) exchange ideas with the academic community.
The second objective is to translate risk assessments into actions by the relevant authorities. The ESRB will provide specific guidance on how to react to systemic risks once they have been identified, by issuing warnings as well as recommendations for remedial action in response to those risks. Therefore, the lack of effectiveness in following up policy concerns on risks with consistent actions, which was one of the previous weaknesses of the financial stability framework, will be addressed. The ESRB, European and national authorities will act together so that risk warnings will not go unheeded as they have been some times in the past. I believe that authoritative and action-oriented statements on what should be done to mitigate systemic risk can make a real impact.
The third objective is to improve the interaction between micro and macro-prudential analysis so as to enhance the effectiveness of systemic risk assessment. To this end, the ESRB will serve as a forum for the exchange of information between the national central banks and the supervisory authorities. We need an environment in which views are shared in a frank and candid way. A common language and supranational coordination between macro-prudential and micro-prudential supervisors is required. In addition, if we are able to combine the expertise and the knowledge of all our members, we will be able to forge new policy tools to address systemic risks.
New official bodies are also being created in other countries and regions of the world to improve systemic oversight and coordinate macro-prudential intervention, namely in the UK and US. A comparison of the institutional models which have been adopted highlights at least one main commonality between the EU, the UK and the US: central banks play prominent roles in all three frameworks. This follows from their important role in safeguarding financial system stability, which they have demonstrated very clearly during this crisis.
A few main differences also come to the fore, some of which could pose challenges to the European setup. First, in the United Kingdom and the United States, the involvement of the central bank in micro-prudential supervision has been strengthened (although in the UK this is via a separate body). In the EU framework, the ECB has a key role, in particular as a provider of analytical and statistical support to the ESRB, but is not directly involved in micro-prudential supervision. Second, in both the UK and the US, the focus of the macro-prudential body is on regulatory and supervisory policies, including the extent of the regulatory perimeter. Although this is not clarified explicitly in the ESRB legislative framework, it is expected that the ESRB will also play a role in this domain by issuing appropriate recommendations and opinions. Third, the role of central banks in crisis management and resolution differs: while the Federal Reserve System and Bank of England can respond to an emerging crisis using traditional liquidity support and regulatory tools, and possibly also as the lead resolution authority, the ESRB does not have any explicit tasks and powers relating to crisis management, apart from the power to advise the European Council on the existence of an emergency situation. Going forward, the quality of ESRB risk analysis, as well as the timeliness and appropriateness of its recommendations, is crucial to overcome some of the challenges stemming from its legal and institutional framework.
II. The role of research and analytical tools in the conduct of macro-prudential supervision
Macro-prudential supervision aims to prevent the emergence of systemic risk and to mitigate its effects. Systemic risk can be described as the risk that financial instability impairs the functioning of the financial system to the point where economic growth and welfare materially suffer. Over the years there have been some differences over the concept of financial stability and systemic risk. As I wrote in 2006: “ …….two main meanings stand out. One, more fundamental, follows Mishkin (1991)  and defines financial stability as a situation where the financial system is able to ensure in a lasting way, and without major disruptions, an efficient allocation of savings to investment opportunities. In another sense, financial stability refers to the absence of a major misalignment of asset prices (bubbles) that can threaten future disruption of markets and the real economy. Both meanings are of course connected as they point to the same notion of smooth functioning of financial institutions and markets.”  Early approaches to the problem connected the concept with the avoidance. Without excluding the issue of excessive volatility of asset prices, recent established concepts are more encompassing and less focused on what would be a very challenging goal to achieve.
At the ECB, we have identified three forms of systemic risks: i) the risk that widespread imbalances that have built up over time unravel abruptly, ii) contagion risk, and iii) the risk of macroeconomic shocks causing simultaneous problems in the financial system. Analytical tools and models are available for the identification and assessment of each of these risks. 
The first type of risk is related to the spontaneous pro-cyclical tendencies of the financial sector to increase leverage and risk-taking in booms and to manifest excessive risk aversion in downturns. The second, stems from the network effects created by the high degree of interconnectedness of the financial system. These first two sources of systemic risk depend endogenously from the sector itself and are connected to well known market imperfections that include information frictions, wrong incentive mechanisms, principal-agent issues or coordination failures.
In any case, macro-prudential supervision will have to address two main sets of questions: first, how to identify, measure and assess the impact of sources of systemic risk: second, devise and calibrate the type of policy instruments that should be used to avoid the materialization of risks or mitigate their consequences.
To accomplish both tasks, we need a lot of research to develop and perfect an appropriate set of analytical tools and models.
Considering the first task, it is evident that an understanding of current systemic risks is at the centre of macro-prudential supervision. A forward looking identification and assessment of systemic risks requires a broad and deep information basis, aided by a wide range of tools to process the relevant data. Important ingredients for robust policy decisions include market intelligence, plain data analysis by inspection of indicators and market developments, more advanced data analysis based on sophisticated analytical approaches and tools and a sound judgement in bringing these different elements together.
In accordance with this conceptual framework, the macro-prudential analysis of the ESRB could be based on two steps, namely risk surveillance and risk assessment. The first component – risk surveillance – consists of monitoring markets, financial institutions and the economy at large, with the goal of detecting potential vulnerabilities. It is important at this initial stage to cast the net as wide as possible to ensure that no risk is overlooked. Several analytical tools and models can be used for this type of analysis. Early warning models and indicators use information in current data to identify early on where imbalances may be building up. For example, the private credit-to-GDP-ratio, property price changes, bank leverage and financial institutions’ maturity mismatches are indicators to consider here. Financial stability indicators can also be useful, as they capture the current state of system stability. Such indicators permit, for instance, an assessment of the severity of current conditions.
In building up an effective risk surveillance process, a comprehensive information base is of the essence. An important element of the ECB’s preparatory work consists of ensuring comprehensive statistical databases for macro-prudential purposes that are available at the level of the EU. These will be established in close collaboration with the national authorities and the future ESAs, enabling particularly risk-focused analyses with adequate detail and scope in terms of, for example, country granularity, financial sector coverage and data frequency. At the same time, information gathered via market intelligence is considered to be of critical value in supporting the risk surveillance process, for example in its ability to capture market developments and perceptions not yet apparent in the most recent data. In this regard, it is envisaged that the ESRB risk surveillance work will be supported by an enhancement of market intelligence activities with an EU-wide reach.
The second step of the ESRB’s macro-prudential process – risk assessment – concerns the evaluation of the possible severity of the materialisation of the risks identified, as well as an evaluation of the ability of the financial system to absorb shocks. Both the likelihood and the severity of a specific risk will be considered in constructing an overall ranking of risks. The expert judgment of the ESRB would draw upon such a ranking in issuing warnings or policy recommendations. Here macro stress-testing models can be used to assess the resilience of the financial system by using some of the risks identified in the surveillance phase of the analysis as input. Contagion and spillover models can also be used to evaluate the impact of specific failures within the financial system. For instance, they can be used to assess the transmission of instability among financial intermediaries and markets owing to contractual and business links.
Let me briefly discuss one trend that we have observed in new analytical developments. A host of new quantitative measures of systemic risk have recently been proposed.  These measures can be useful for macro-prudential oversight and policy discussions, because they provide “thermometers” or “barometers” of widespread risks and stresses in the system. More specifically, the 2009 Geneva Report on the World Economy entitled “The Fundamental Principles of Financial Regulation” argued that the internalisation of risk externalities that financial intermediaries impose on each other requires regulatory capital to be linked to comprehensive measures of this form of systemic risk.  The same approach was also recommended in the collective work of the New York University “Restoring Financial Stability” edited by Viral Acharya and Matthew Richardson.  The researchers are interested in the extent to which an institution imposes a negative externality on the system, which also captures an institution’s propensity to be undercapitalised when the system as a whole is hit by a shock.  In his keynote speech today, Professor Engle presented one important contribution – as part of a larger systemic risk ranking exercise undertaken by faculty members at New York University – towards making such systemic risk measures operational.  In addition, the two papers in the session on “Measuring systemic risk” make valuable advances in this area.
I welcome these and related research efforts, which seek to build reliable indexes of the systemic importance of financial intermediaries. These firm-specific measures should eventually allow us to steer potential macro-prudential policy instruments in the right direction. Nevertheless, it is my impression that, in order to be used in policy decisions, research also needs to become broader and deeper. For example, systemic risk indicators based only on equity returns may be subject to episodes of mis-pricing and ignore the intricate capital structure of financial intermediaries, the liquidity risk arising from the maturity mismatch between assets and liabilities. Also, such indicators do not take as the main gauge the overall effect on the economy at large.
As I mentioned before, the second main task of macro-prudential supervision will be to apply the appropriate policy instruments to contain systemic risk or mitigate its effects. In the case of the ESRB it is true that no competences were given to it in the domain of deciding or implementing policy instruments. Nevertheless, to fulfil its mandate, the recommendations that will be issued by the ESRB must be in many cases sufficiently concrete and indicate what type of policy instruments should be applied by the authorities that are the addressees of such recommendations. In some cases, these recommendations will refer to the use of already existent regulatory competences but in other cases they will require changes in regulation. This considerably extends the scope of analysis and research that must be developed to assist the ESRB in its task. The members of the ESRB will all contribute with their knowledge and expertise. We will nevertheless need more contributions from economic research to better assess the impact of the risks and the effectiveness of policies to neutralize them.
The general overhaul of regulation that has been recently decided or is being prepared has far- reaching implications from the point of view of macro-prudential policy. The increase of capital and liquidity requirements, the limits to leverage, the new framework to deal with derivatives, the measures to address the problem of institutions “too big to fail” either through higher loss absorption capacity or better resolution schemes − are all measures that contain important elements responding to macro-prudential concerns. The implications of all these measures will have to be followed very carefully and attention must be directed to detect new risks that may emerge from the reaction to the new regulations. For instance, the increase in the cost of capital may lead to strategies by financial institutions to more risky activities and projects or to the creation of new types of non-regulated entities to conduct financial business. Information gathering and good analytical tools will be required to react to such possible developments.
Besides the more structural and permanent regulation of the financial sector, other measures will be of a time-varying nature, indexed to the economic cycle, like capital buffers or provisions directed to smooth out the pro-cyclical effects of financial sector behaviour. Sometimes, the mentioning of macro-prudential policy wrongly evokes only this type of measures. Their aim is to increase the resilience of financial institutions to the fluctuations of the cycle and at the same time to guarantee a more stable flow of credit and financial services to the economy. We could say that this type of policies intends to protect the banks from the cycle and to shelter the cycle from the banks’ behaviour. There are still many questions surrounding the proper understanding of the use such policies and their calibration. To mention some examples: what is the relative effectiveness against pro-cyclicality of measures like capital buffers, dynamic provisions or accounting impairments subject to the expected loss model?; what should be the calibration and the triggers that determined the use of some of these instruments over the cycle?; is there justification for liquidity buffers that vary also with the cycle?
These and others questions need to be answered and for that we need the contribution of economic research to develop the methods and models that allow policy makers to take the appropriate decisions.
III. Mars – an ESCB macro-prudential research network
This brings me to our own research efforts in the areas of systemic risk and macro-prudential supervision. A little while ago, the ESCB – this means the 27 EU national central banks and the ECB – agreed to establish an internal network called “Mars” for macro-prudential research. The objective of Mars is to develop core conceptual frameworks, models and tools that will conduct research with a view to improving macro-prudential supervision in the EU.
I should like to make a point of encouraging researchers from outside of the ESCB to also work in these fields and to contribute papers to open events and conferences such as this one. In the field of macro-prudential supervision we are in a much less comfortable position than in the monetary policy field, for example, in particular with regard to widely accepted conceptual frameworks and scientific foundations. Thus far new developments in the academic sector do not seem to have advanced fast enough or to have covered all of the issues in a way that can support us with the launch of the new macro-prudential bodies now or at the latest next year. Substantial knowledge gaps remain, and I think that research plays an important role in this regard. Through Mars, we hope to fill some of these gaps ourselves and act as a catalyst for more and better-targeted academic research in this field. The network has been charged with conducting research in three areas.
The first Mars work area deals with the development of aggregate theoretical and empirical models that link realistic characterisations of widespread financial instability with the performance of the economy. These macro-financial models should aim to capture the two-way relationship between the financial system and the economy at large, a feature that is, for example, largely absent from standard stress-testing models. We need a better understanding of how financial instability can be represented in aggregate economic models. We also need to further clarify the transmission channels of financial instability to the aggregate level, in particular the role that is played by non-linearities, amplification, and feedback effects. Explicit models of the endogenous build-up and unravelling of imbalances, for example, of the Minsky-Kindleberger type, have also received insufficient attention despite their potential to capture relevant features of financial instability. I appreciate recent work in this area, part of which will also be discussed tomorrow. In particular, Professor John Geanakoplos - tomorrow's keynote speaker - has researched "the leverage cycle" since more than a decade. The objective of developing such models is not just to answer fundamental research questions, but they also have the potential to improve existing operational analytical tools, such as macro stress-testing, simulation and forecasting models. For example, it is likely that macroeconomic models that truly capture situations of widespread financial instability would have performed better in predicting the sharp downward revisions of GDP observed during the crisis. Such models will also be extremely helpful in formulating conceptual frameworks for policy-making. This conference showcases a number of promising advances in this regard. 
The second work area on which the Mars network will focus consists of fostering faster progress on early warning models and systemic risk indicators. Although predicting financial crises is a formidable task, in this area macro-prudential supervisors at least have the broadest operational analytical basis available. However, in addition to the need for measures to assess systemic risk externalities for the purposes of financial intermediary capital regulation, important areas of work remain, including for example devising metrics of overall systemic risk, improving the forecasting performance of tools for specific markets and intermediaries, achieving an optimal structure for a comprehensive system of early warning tools covering all relevant countries and sectors, as well as the aggregation of heterogeneous tools that may be used in countries with different financial structures. For instance, while considerable research efforts in this area are already under way, it is my impression that there is still no widely accepted indicator or quantitative diagnostic framework that measures overall systemic risk. Although notable progress has been achieved recently, reflected in some of the contributions presented at this conference, even the most sophisticated tools available today rely only on relatively narrow definitions of a systemic event and are therefore rather partial in nature. More comprehensive yet still tractable frameworks for identifying systemic instability, as well as early warning models for the full picture on systemic risk remain to be developed. Notice the sharp contrast with the monetary policy field, where the main objective variable can be captured through consumer price indices that are relatively mature statistically and economically.
The third area in which the Mars network seeks to achieve progress is the assessment of contagion risks. While there is already a substantial amount of academic and central bank research on bank and financial market contagion, open issues of some importance include cross-border risks within Europe and the transmission of instability through non-bank financial intermediaries, such as insurers, re-insurers, hedge funds, brokers/dealers and private equity firms. However, research on these issues is held back by data limitations, which we hope will be relaxed over time. In addition, the better capturing of feedback and amplification effects through the endogenous risk behaviour of financial agents may substantially improve available contagion simulation tools.
The main work programme of Mars, in particular its objectives and central research questions, has been defined, and research on a large number of individual projects is under way. It is envisaged that the network will operate for a period of two years. It is led by Philipp Hartmann of the ECB, supported by coordinators from some EU national central banks, including the Banque de France, the Banca d’Italia and Česká národní banka, and two chief academic consultants, Xavier Freixas from the Pompeu Fabra University in Barcelona in the finance field and Carmen Reinhart from the University of Maryland in the macro field. We plan to publish a report on the main outcomes in 2012. However, individual research papers will be circulated in the meantime and we may also choose to report on important specific results when this is warranted. Once again, I should like to invite the academic community to join us in our research efforts by focusing on the areas I have outlined, as well as to contribute papers to any related events.
IV. Concluding remarks
To conclude, the agreement on a new macro-prudential policy framework in Europe is a necessary and important step forward. Now we must strive to make it work the way it is intended. The ECB is committed to providing the necessary support to the ESRB, as envisaged in the legislation, and to collaborating with the institutions and individuals involved. Even though currently available research and analytical tools are in no way sufficient, they nevertheless constitute an important element in the future work of macro-prudential bodies in general and the ESRB in particular. We therefore encourage the research communities in both academic and policy-oriented institutions to contribute to the necessary developments and to work with us where possible. This conference is an important step in that direction.
Mishkin, Frederic S. (1991) “Anatomy of financial crisis” NBER Working Paper n. 3934
Vítor Constâncio (2006) “Finance and Regulation” Meeting of the European Association for Banking and Financial History, Lisbon 26-27 May 2006 (available in http://www.bportugal.pt )
For in-depth discussions of concepts, tools and processes in the areas of systemic risk and macro-prudential supervision, see Special Feature B “The concept of systemic risk” in Financial Stability Review, ECB, December 2009, and Special Feature B “Analytical models and tools for the identification and assessment of systemic risks” in Financial Stability Review, ECB, June 2010.
The ECB contributed to developing such methodologies in the early to mid-2000s, but at that time the interest in them was rather measured (see, for example, Hartmann, P., Straetmans, S. and de Vries, C.G. (2004), “Asset Market Linkages in Crisis Periods”, Review of Economics and Statistics, 86(1), pp. 313-326, or Hartmann, P., Straetmans, S. and de Vries, C.G. (2006), “Banking system risk: a cross-Atlantic perspective”, NBER Working Paper Series, No. 11698, October. Since then, a variety of such measures have also been presented in the ECB’s Financial Stability Review.
 See Brunnermeier, M., Crockett, A., Goodhart, C., Persaud, A. D. and Shin, H. (2009), “The Fundamental Principles of Financial Regulation”, Geneva Report on the World Economy, No. 11, July.
Viral V. Acharya and Matthew Richardson (editors) “Restoring Financial Stability: how to repair a failed system” John Wiley & Sons, 2009
See Acharya, V.V., Pederson, L.H. Philippon, T. and Richardson, M. (2010), “Measuring Systemic Risk”, Working Paper Series, SSRN, May available at http://ssrn.com/abstract=1573171
See Brownlees, C.T. and Engle, R.F. (2010), “Volatility, Correlation and Tails for Systemic Risk Measurement”, Working Paper Series, SSRN, May available at http://ssrn.com/abstract=1611229
See, for example, Brunnermeier, M.K. and Sannikov, Y. (2010), “A Macroeconomic Model with a Financial Sector”, mimeo, Princeton University, May, available at http://www.princeton.edu/~markus/research/papers/macro_finance.pdf, or Jeanne, O. and Korinek, A. (2010), “Managing credit booms and busts: A Pigouvian taxation approach”, Working Paper Series, Peterson Institute of International Economics, No. 10-12, September, available at http://www.piie.com/publications/wp/wp10-12.pdf