1.01  It is said that ‘generals fight the last war’. Regulators can do the same. In the wake of the crisis, regulators and politicians are currently fighting on a number of battlefronts. New policies are rapidly proliferating from the international committees of regulators, from international organizations such as the OECD and IMF, from national governments, and from the EU. New provisions are being brought in, regulatory structures are being reconfigured, and new regulators are being created.

1.02  The question is whether the financial regulators are building the regulatory equivalent of the Maginot Line, or whether they are devising strategies that will enable them to anticipate future crises, and sufficient resilience to withstand them when anticipation fails. For we know there will be another crisis, but do not know when it will come or the form it will take.

1.03  There is no doubt that policymakers recognize some of these difficulties. In particular, the need to build a greater capacity for surveillance has been recognized in the creation of organizations charged with macro-prudential surveillance at the global, EU, and in some cases national level. However, regulators failed not just because they did not look hard enough at what was happening in the markets. It (p. 4) was also that their cognitive understandings of the way markets operated, and the way markets and regulation interacted, were flawed. Conventional wisdom has been overthrown by the power of events, prompting a paradigm shift in some aspects of financial regulatory policy. However, just what paradigm should replace the old is still unclear, although the contours of the new conventional wisdom are becoming clearer. Part of the challenge for regulators going forward is whether they can put in place ways to challenge conventional wisdoms, including their own, before another crisis does it for them.

1.04  The crisis thus created a series of cognitive as well as financial shocks. This chapter analyses the cognitive shifts prompted by the crisis and the impact they are having on the organization of global and EU financial regulation. The chapter focuses on regulators’ capacities for anticipation rather than resilience per se, and argues that for these capacities to be developed, the current mechanisms by which the financial regulators learn of their own and each others’ performance need to be quite fundamentally reoriented. In addition, regulators need to build in mechanisms for cognitive challenge. As senior economists advised the Queen, the crisis was caused by a ‘collective failure of imagination’.¹ But despite the various changes being introduced, it is not clear whether some of the main lessons have really been learnt. There is some recognition by regulators of the complexity, fragmentation, interdependencies, and dynamic adaptability of the system they are attempting to regulate. However, rather than move to build flexibility and scope for variety and learning into the regime, the policy dynamic at the moment is to retreat into hierarchical regulation that seeks to control both markets and other regulatory actors through detailed rules, and within the EU, harmonized control. The dynamics driving these policy processes are understandable, but there is a significant risk of introducing new rigidities into the system. Instead, reflexive learning and dynamic responsiveness to the regime’s own performance need to become the central principles on which the global and EU regulatory regimes operate.

1.05  The chapter argues that in order to develop structures and strategies for dynamic responsiveness we need first to understand the changes in the organizational structures of global and EU financial regulation and the cognitive shifts which underlie them. With respect to financial markets, the assumption that the ‘sum of the whole is greater than its parts’ has been replaced by the recognition that ‘the whole is only as strong as its weakest link’.² It is increasingly recognized in some quarters that this aphorism applies as much to the regulatory system as it does to the markets. (p. 5) No longer is it seen as sufficient to focus on the monitoring and supervision either of individual institutions or of individual countries; what is important is the operation of the system as a whole.

1.06  Linked to these cognitive changes is a normative reassessment of what regulation should be trying to achieve. Although there is broad consensus on the normative goals of financial regulation, namely financial stability, investor protection, and prevention of market abuse, there is far less agreement as to just what these mean in different instances,³ and even less as to what measures should be taken to attain them. There are also conflicting policy dynamics, with centripetal pressures to move regulation to the international level being countered by centrifugal pressures for unilateralism either by individual countries or the EU.

1.07  The chapter groups these developments into four main areas: surveillance, resilience, stability, and regime management. It then considers the developments in the institutional structures at the international level, focusing on the re-formulated, and newly invigorated, Financial Stability Board (FSB), the IMF, the EU and the multi-lateral arrangements for coordination through colleges of supervisors at the global and EU levels. In each case it considers current mechanisms of observation, communication, and enforcement mechanisms within different parts of the regulatory regime, focusing primarily on the changing role of the FSB. The discussion then turns to examine in more depth the difficult issues of regime management, coordination, and learning. The pace of policy development is so great and the proliferation of proposals currently emanating across the regime so numerous that there are significant risks of creating further tensions and internal conflicts if their development is insufficiently coordinated.⁴ Coordination is not simply a matter of technical agreement, however; the tensions and conflicts are often deeply rooted, raising significant issues of regime management and enhancing the need to develop capacities for reflexive observation and dynamic adaptation. So the final section of the chapter considers how the system’s existing and potential capacities for reflexive learning and dynamic responsiveness can be strengthened. It focuses on two elements of that challenge: enhancing information and knowledge, and developing mechanisms of challenge.

1.08  It is fashionable now in financial regulation circles to emphasize the complexity and adaptability of the financial system and the unpredictability of regulation.⁵ However, more generic analyses of regulation and governance have been emphasizing these themes for some time.⁶ Regulatory systems and the systems they attempt to regulate are often complex and dynamic, with significant interdependencies existing within and between them.⁷ Power and knowledge are fragmented between different actors, with significant implications for the construction and operation of regulatory regimes. The performance of regulatory governance is also often disaggregated into a number of different functions that are dispersed between a number of actors at the international, regional, national, and sub-national level who co-exist in a range of different relationships. Complex interactions and interdependencies exist between social actors, and between social actors and government in the process of regulation, some of which regulation itself creates.⁸ Those being regulated have significant operational autonomy, and their response to regulation is often unpredictable. Those attempting to regulate others thus face significant challenges. Some of these are unique to the particular task or task environment—such as managing particular risks or structuring certain markets; others, such attempting to manage behaviour, are generic across regulatory regimes.

1.09  Regulation thus is a messy, complex, and largely imperfect process. In order to have some hope of regulating effectively the regulatory system collectively needs the capacity to regulate dynamically. Regulatory capacity, it is suggested, is a composite of resources and motivations.⁹ Resources comprise: information, knowledge and expertise, financial resources, organizational capacity, power or strategic position, and authority and legitimacy. However, resources are only one aspect of capacity; what is also important is an understanding of motivations of those in possession of resources, their ‘motivational postures’ or ‘attitudinal settings’, which may be based in norms of self-interest, appropriateness, or both.¹⁰ Moreover, where either (p. 7) or both regulatory actors and capacities are dispersed, then there needs to be some way in which they are coordinated to achieve the particular outcome or outcomes sought, either by the whole or each of its participants.

1.10  The challenges facing the global financial regulatory regime are thus not unique in all respects, but they are severe.¹¹ The regime comprises non-state and state actors operating at the global level, at the regional level, at the national level, and within individual firms. The focus here is not on the relationship between regulators (state or non-state) and firms.¹² Rather, it focuses on the interactions between regulators themselves within the regime.¹³

1.11  At the global level, the regime is characterized by a multiple of international regulatory committees (IRCs) and other organizations, each with different memberships, legal bases, mandates, and powers.¹⁴ There are individual committees of securities regulators, notably the International Organization of Securities Commissioners (IOSCO), central bankers from the G20 countries (pre-crisis, from the G8) who are members of standard setting committees of the Bank of International Settlements (eg the Basle Committee on Banking Supervision (BCBS)), and insurance regulators (International Association of Insurance Supervisors (IAIS). These coordinate in the development of principles for financial conglomerates in the Joint Forum. In addition, the International Accounting Standards Board (IASB), a group of professional experts, sets accounting rules, the International Financial Reporting Standards, which are used by most major economies apart from the United States. The international financial institutions (IFIs), the World Bank, and the International Monetary Fund (IMF) play a role in that they monitor the implementation of some of the IRCs’ Principles within individual countries as part of their broader Financial Services Assessment Process (FSAP) activities. Although the IFIs are the only organizations in this global regulatory matrix which have a legal underpinning, the FSAP process is conducted under the IMF’s non-binding technical assistance programme, and so technically takes the form of guidance and assistance. All global regulation is thus based on soft law. Finally, these bodies are (p. 8) all members, together with G20 countries, of the Financial Stability Board (FSB). The FSB originated in 1999 when, following the rescue of a hedge fund, Long Term Capital Management, the Financial Stability Forum (FSF) was created to coordinate these international regulatory committees, the IFIs and the OECD, and to bring G7 finance ministers closer to the global standard-setting processes.¹⁵ As a consequence of the crisis, the membership of the FSF and the Basle committees BCBS was broadened to comprise the G20 countries, and the FSF reconstituted as the FSB and given a mandate by the G20 to implement its policies on financial regulation.

1.12  At the European level, in addition to the usual law-making institutions, there are separate committees of finance ministers that advise the Commission on securities, banking, and insurance legislation. In the period immediately preceding the crisis, three ‘level 3’ committees of regulators also operated in each of these areas: the Committee of European Securities Regulators (CESR), the Committee of European Insurance and Occupational Pension Supervisors (CEIOPS), and the Committee of European Banking Supervisors (CEBS).¹⁶ Their role was to advise and facilitate consultations on draft legislative proposals and to develop technical guidance on the implementation of legislative provisions. Prior to the crisis, the European Central Bank has to date played almost no role in financial regulation. As discussed below, this structure was radically altered with the introduction of the European System of Financial Supervision in January 2011.

1.13  The financial crisis has prompted significant changes in the organizational structures of financial regulation at the global and EU levels. Underlying these changes is a complex set of policy and institutional dynamics. The political talk is of harmonization, coordination, and cooperation, but on key issues we see Balkanization as countries go it alone.¹⁷ Centripetal pressures that push regulation towards certain central points are being counteracted by significant centrifugal pressures which are pushing it back towards national governments.

(p. 9) 1.14  There are a number of centripetal pressures for international harmonization. The markets are peripatetic and can easily engage in regulatory arbitrage, so countries that introduce tougher regulation can suffer ‘first mover disadvantage’ as financial business moves elsewhere. Moreover, the crisis demonstrated that the globalized markets create significant interdependencies between regulators, with each relying on the other for it to be able to achieve its objectives, both in normal times and in times of crisis. They also create interdependencies between governments’ fiscal positions. In particular, within the EU the principle of home country control, combined with the global nature of financial markets, makes one country significantly dependent on the quality of regulation in another. Through the passporting regime, one country’s weak regulation can be brought into another country through a cross-border bank.¹⁸ At the same time there is less willingness to trust other national governments to deliver regulatory regimes that are robust enough to be relied upon. Harmonization, supported by mechanisms of enforcement, can increase control by a group of countries over another’s regulatory regime and thus reduce the vulnerabilities that such interdependencies create. Further, as the handling of the crisis demonstrated, unilateral actions by one country can have negative spillover effects on others, which coordinated action can avoid. For example, when Ireland introduced a full deposit guarantee for all its banks in September 2008 there was an immediate flight of capital from the United Kingdom to Ireland, prompting the United Kingdom and then the EU to raise deposit guarantee limits within a matter of days.

1.15  So on the one hand, although normative consensus may be hard to achieve, regulators and governments recognize that international harmonization can be in their own interests for a number of reasons. These ‘bottom up’ pressures accord with attempts from international bodies and the EU authorities to enhance the control that they exert over the financial regulatory regime, or at least parts of it. All of these factors together exert centripetal pressures, pushing regulation to the regional and international level.

1.16  On the other hand, centrifugal pressures are also strong. National governments have become acutely aware of the potential cost that the financial system can impose on their own taxpayers. As has been acutely observed, banks are global in life but national in death.¹⁹ There are clear signs that this ‘mortality mismatch’, as I call it, is exerting a fundamental influence on policy-making, prompting policies of ‘fiscal protectionism’. There is a recognition that international harmonization, or at least co-ordination, is necessary to manage financial institutions in life (eg through colleges of supervisors) and when they are critically ill (through (p. 10) cross-border crisis management procedures). However, the process of managing their death (powers and procedures for the resolution of failing banks) and dealing with its consequences (funding bail-outs and/or deposit guarantees) remains national, at least for the moment. In 2007–09, governments in the United States, the United Kingdom, and the rest of the EU were forced to inject US$4.89 trillion directly into banks and other financial institutions, equivalent to 6 per cent GDP in each country/region and to issue guarantees on bank borrowing and bank assets that, if called upon, would equate to US$14 trillion gross: the equivalent of 50 per cent of the GDP in each country/region.²⁰ Governments are understandably concerned to ensure that they will not have to do this again and are putting in measures to prevent such calls on their budget deficits, often unilaterally, though talk of a common resolution fund for EU banks is growing. Whilst it remains national governments and their taxpayers that have to pick up the bill for financial failure, those governments will be concerned to protect their fiscal position and to retain the right and ability to do so.

1.17  There is thus a fundamental tension between the centripetal forces which push agenda-setting to the international level, and centrifugal forces which push it out to national governments as countries ‘go it alone’. Managing this tension is not straightforward. Part of the reason for the Commission’s proposal for an EU-wide resolution fund for banks (as distinct from that for sovereign debt), for example, is to minimize the incentive effects for EU Member States to act unilaterally in a crisis to protect their own taxpayers on the insolvency of a cross-border bank.²¹

1.18  The organizational restructuring, it is suggested, is integrally linked both to these conflicting forces and to some of the significant cognitive shifts in the understanding of the nature of the markets, the nature of the problems, and the solutions that need to be imposed. The plethora of current policy initiatives is potentially over-whelming, but in an attempt to impose at least analytical coherence on them they are divided here into four interlinked groups: surveillance, resilience, stability, and regime management.

1.19  Getting information, and just as importantly, making sense of it, is a critical element of any regulatory system and has been shown to be deeply lacking in financial regulation.²² As the FSB/IMF report on information gaps observed, ‘[T]he recent (p. 11) crisis has reaffirmed an old lesson—good data and good analysis are the lifeblood of effective surveillance and policy responses at both the national and international levels’.²³ With respect to institutions, supervisors need to know more about their liabilities, their consolidated position, and the nature of their maturity mismatch. With respect to the system, they are seeking to understand the build-up of risk within the system, the nature of the inter-linkages between participants in the markets and between different types of risk, and the nature of maturity mismatches that exist on banks’ balance sheets and within and between currencies. In particular, the challenge is to understand the interrelationship between macro-economic developments at the global and national level, their relationship with movements in the financial markets, and their impacts on the stability of individual financial institutions, and in turn of those institutions on financial stability—the ‘macro-micro’ link.²⁴

1.20  The crisis showed that these links had been either ill-understood or significantly underemphasized. For example, the links between aggregate leverage, valuation, and liquidity, particularly in a mark-to-market environment created by accounting rules and margin requirements, had simply not been recognized prior to the crisis.²⁵

1.21  The crisis also revealed significant blind spots in the regulator’s vision. There was very little data on inter-institutional, intra-group, and cross-border exposures.²⁶ Some countries had only patchy information on their payment and settlement systems.²⁷ Particular activities of some financial institutions were firmly in the spotlight, but others, notably hedge funds but also the activities of banks themselves, operated very much beyond the regulators’ gaze. The ‘shadow banking system’ should be more accurately described as the ‘invisible banking system’. ‘Over the counter’ dealings in derivatives equate to ‘under the radar’ dealings, so little is known by regulators about the details of these deals or of the inter-linkages and risk concentrations that are thereby created.

1.22  Visibility is now sought over all activities of all participants, in a relentless search for ‘synoptic legibility’.²⁸ However, whilst improved legibility is necessary, what is also important is the development of a cognitive framework in which to make sense of the information collected and convert it from data to knowledge.²⁹ Here the central (p. 12) challenge at the moment is to understand the exact nature of the links between macro-prudential developments and the adjustments that need to be made to the micro-prudential supervision of individual financial institutions. Surveillance is important, but until those links are better understood it remains only a partial basis for dynamic adjustment of the regulatory regime to the changes in risk profiles of products, markets, and actors that comprise the financial system.

1.23  Secondly, regulators are struggling to find ways to enhance the resilience of the financial system. This is not a new strategy in the regulatory governance of risk. Indeed, some argue that resilience rather than anticipation should be the dominant strategy for responding to uncertainty.³⁰ In engineering terms, and particularly when dealing with complex systems, regulators should ensure that systems have built-in buffers between parts that may fail to prevent a failure in one part affecting other parts, and ‘redundancies’, controls which come into operation when others have failed.³¹ What is new to financial regulation post-crisis is the conscious focus on bank recovery and resolution, and the relationship of resolution and supervision.

1.24  The rationale for regulating banks has always focused on the potential for the failure of one institution to have systemic consequences, but the channels for that systemic crisis have traditionally been seen to be inter-linkages through the payment system and inter-bank market, and through contagion effects arising from bank runs caused by loss of investor confidence.³² The regulatory system has traditionally built in resilience through the lender of last resort facility, and failing that, through the contained collapse or managed takeover of the bank. The crisis showed that in many countries there was no further backstop plan. In some cases, for example the United Kingdom, putting these structures in place had been on policymakers’ ‘to do’ lists, but in the benign macro-economic environment that prevailed, they were not seen as particularly urgent.³³

1.25  Building in resilience to the financial system is proving an extremely challenging task. At one level, the battle is relatively straightforward. It involves enhancing the types of resiliency structures and mechanisms that already exist, or creating new techniques which follow the same logic and understandings of how the financial system works as operated in the past. These include the introduction (p. 13) of special legal regimes for the failure of banks,^34,35 requirements, initiated in the United Kingdom, taken up by the BCBS and more recently by the FSB, for firms to produce ‘living wills’ or ‘resolution plans’,³⁶ and initiatives to enhance the cross-border crisis management and the resolution of cross-border banks.³⁷ This work is critical, but so is recognition that individual resolution plans may only work if banks fail one at a time. But when systemic crises hit, several fail together. Whilst core operational functions may be maintained by the resolution authorities, if markets are frozen there are then no buyers for the bank’s assets, good or bad.

1.26  The hardest part of the battle is coming to grips with the realization that the system simply did not operate in the way that regulators, banks, and economists had thought it did. If you do not understand how the system works, it is very hard to build in mechanisms either for managing risk or for ensuring the system’s resilience when those risks crystallize. Whilst regulators and others realize a new cognitive framework is needed, they are struggling to develop one.

1.27  For example, in the initial post-crisis phase, the nature and characteristics of systemic risk, which had previously been almost unquestioned, were found to be matters of considerable doubt and debates began as to which institutions or practices are systemically important³⁸ and how systemic significance is created.³⁹ The identification of the ‘global SIFIs’ has brought some closure to that debate, but it may be disrupted again when the next crisis strikes.

1.28  Linked to these cognitive shifts is the realization that the technologies of risk management that both regulators and financial institutions had used were deeply inadequate, if not fundamentally flawed. They also gave regulators and firms a highly distorted view of the nature and distribution of risks in the financial system. There is widespread recognition that liquidity risk was underestimated,⁴⁰ as was leverage: risk-adjusted leverage measures failed to capture the multiples of exposure created (p. 14) by pooling and tranching structured credit instruments, or the compounding of that embedded leverage through re-securitization.⁴¹ Further, there is dawning awareness that although regulators and risk managers can tend to put risks into ‘buckets’: market, credit, liquidity, and so on, risks can quickly overspill from one to the other: liquidity risk quickly evolved into market and credit risk, as lack of liquidity caused a drop in market value and prevented firms from rolling over short-term borrowing.⁴²

1.29  Risk models were also found to be deeply flawed. Value at Risk (VaR) models, the darling of risk managers and banking regulators since the early 1990s, were found to have procyclical effects. As all follow the same models, there is a herd effect, further enhancing volatility. As a result there are now calls for VaR models to ‘see through the cycle’⁴³ and a pull away from their entrenchment in the Basle capital requirements. Model-based risk assessments were also based on limited historical data, often restricted to the ‘Golden Decade’ of the last ten years that had seen particularly benign economic conditions.⁴⁴ Even worse, in the case of residential mortgage-backed securities, they were based on no empirical data at all.⁴⁵ Correlation risks were not understood, again particularly with respect to credit derivatives obligations.⁴⁶ Stress tests were woefully inadequate. Even in late 2007, before the most intense periods of the crisis, the Chief Financial Officer of Goldman Sachs, David Viniar, was reported as commenting that events that were in most models assumed to happen only once in several billion years (once every 6 × 10¹²⁴ lives of the universe in fact) were happening several days in a row. Models will always be wrong to an extent, but as Andrew Haldane commented, the models failed Keynes’ test, which is that it is better to be roughly right than precisely wrong. ‘With hindsight, these models were both very precise and very wrong’.⁴⁷

1.30  Moreover, it has been recognized that regulation does not operate as a neutral instrument, but can create negative feedback loops, amplifying the very risks they are meant to be controlling. The potential procyclical effects of Basle II have long been noted.⁴⁸ Moreover, the failure of the leverage measures (where they were applied) to capture the full extent of leverage or the multiples of exposure incentivized (p. 15) greater leverage and did nothing to abate the risk concentrations that were accumulating. The problems did not only lie with capital rules for banks. Accounting rules on loan loss provisioning based on incurred losses, and in particular mark-to-market accounting requirements, have been shown to have profoundly procyclical effects.⁴⁹ Private sector risk management techniques had the same impact, enhancing the links between valuation techniques, leverage, and asset prices. Triggers in debt or over-the-counter (OTC) derivative contracts and haircuts on financing transactions based on market valuations or credit ratings added liquidity during the boom but drained it out when conditions were stressed, exacerbating deleveraging and asset sales in a vicious downward spiral.⁵⁰

1.31  Thirdly, there is a focus on developing strategies for macro-economic stability. Here too there has been a significant cognitive shift. In particular, there is a recognition of the inter-linking and inter-dependencies between the stability of macro-economy, individual financial institutions, and the network effects of the financial system as a whole. The mechanisms of these inter-linkages are still only partly understood, but there is broad agreement emerging on four issues.⁵¹ First, the stability of the financial system is linked to stability of the macro-economic system, and vice versa (the macro-micro issue). Secondly, that ensuring the stability of one does not ensure the stability of the whole (the inter-connectedness issue). Thirdly, though more tentatively, that financial stability is not a ‘banks only’ issue but is affected by the activities of all market players, including those not normally within the remit of banking supervisors, such as hedge funds, OTC dealings, insurance and reinsurance companies, and mutual funds (the ‘network’ issue).⁵²

1.32  There is far less consensus on the policy implications that should flow from these diagnoses, however. There are significant divides on key issues, for example whether banks should be broken up and how;⁵³ whether monetary authorities or financial supervisors can spot ‘bubbles’ and have the political authority to pierce them;⁵⁴ (p. 16) what the relationship is between monetary policy and financial supervision;⁵⁵ or whether trade imbalances can and should be addressed.⁵⁶

1.33  Fourthly, there is a concern to put in place institutional structures which can develop and deliver a regulatory system that will ensure that these battles are successfully fought at every level, from international committees, through regional and national systems of regulation, down to firms’ internal governance structures. Here too there has been a cognitive reframing, but this time as to the needs, capacities, and resources of different actors in the regulatory system, the nature of their inter-linkages and interdependencies, and the negative feedback loops and externalities that thereby can be created.

1.34  As explained above, regulatory capacity is a combination of the possession of certain key resources and the ability and willingness to deploy those resources in pursuit of a certain set of normative goals. Critical resources are information, organizational capacity, expertise, financial resources, strategic position, and legitimacy and authority. The crisis demonstrated that regulators did not have adequate information nor did they have a means of making sense of what they had. They did not have sufficient technologies to manage and regulate risks. They did not have the organizational capacities to coordinate and perform regulation in conjunction with other national authorities or overseas regulators. The strategic position of those purporting to perform regulation, over firms or other regulators, was often weak.⁵⁷ At the international level, and indeed at the EU level, there are significant questions of the legitimacy and authority of some those purporting to manage the regulatory regime by those they hope to manage.

1.35  There are four discernible themes in the current organizational realignments, which are focused on in more detail below. First, there are ‘re-centrings’ occurring: changes in the distribution of regulatory powers and functions between actors within the system and attempts by various actors in the system to pull more regulatory powers back to themselves. These are happening at the national level as national governments reorganize their national systems of regulation.⁵⁸ They are evident at the regional level, as the EU radically restructures its regulatory regime and the balance of responsibilities and powers between the EU and Member States. They are occurring at the international level, as participation is broadened and functions redistributed between the ‘siloed’ international committees and the coordinating (p. 17) role of the Financial Services Board. They are also visible within firms themselves, as regulators focus in far more detail on the internal governance structures of firms, and on the role of shareholders within them.⁵⁹

1.36  Secondly, there is a re-evaluation of the regulatory capacities of different participants within the regime, including regulated firms and other market actors. In particular, there is a growing recognition that regulators do not have the resources necessary by way of information, tools and technologies, organizational capacities, leverage, and, at the international level in particular, legitimacy and authority to perform regulation effectively. However, this is also matched by a recognition that firms and markets do not possess these resources either.⁶⁰ In many areas, self-regulation is politically dead, or at least in intensive care.⁶¹ The OTC derivative markets, credit rating agencies, and hedge funds are all finding the regulators’ spotlights turned firmly on them, though often for very different reasons.⁶² Firms’ internal governance structures and remuneration policies are also firmly in the regulators’ sights.⁶³

1.37  Thirdly, and linked to this, are changes in patterns of enrolment, in particular the role of credit rating agencies. Much attention has been given to the reliance that (p. 18) investors placed on ratings and there have been calls from regulators for investors to perform their own due diligence.⁶⁴ However, the regulatory system has been just as negligent. In significant areas, credit ratings are hardwired into the regulatory regime, often acting as thresholds or triggers for regulatory action.⁶⁵ The standardized approach introduced under Pillar 1 of Basle II relies on credit ratings of borrowers assigned by ‘external credit assessment institutions’ (ECAIs) where these are available to compute banks’ regulatory capital for credit risk.⁶⁶ Central banks have relied on credit ratings to determine what they will accept as collateral.⁶⁷

1.38  Enrolment can confer benefits, extending regulatory capacity. However, as the crisis has demonstrated, enrolment can enhance regulatory capacity but it also creates significant dependencies and vulnerabilities.⁶⁸ Moreover, it can create negative feedback loops, as illustrated in the effects of the mark-to-market accounting rules noted above. It can also create opportunities for gaming the regulatory rules—as where banks guaranteed the liabilities of their SPVs, which gained a high credit rating as a consequence. Banks then bought the commercial paper of their special purpose vehicles, relying on the high credit rating to reduce their capital requirements.⁶⁹ There is thus a significant re-evaluation of the nature and extent of this particular enrolment relationship. Credit rating agencies are now to be regulated within the EU.⁷⁰ Central banks have indicated that they will no longer rely on credit ratings agencies to determine what collateral they will accept.⁷¹ Under the revised Basle II requirements on securitization, exposure has been amended to include requirements to ensure that banks perform their own due diligence and do not simply rely on credit ratings given by the agencies.⁷²

(p. 19) 1.39  Fourthly, new actors are being created, existing ones reconstituted, and relationships between regulators are being rearticulated and reformed, as detailed in the next section. There is a notable thread running throughout these restructurings, which is that there needs to be a greater focus on managing the regime as a whole. However, as discussed in the next section, this management process is fraught with difficulties, all of which have a bearing on the ability of the regime and its different components to develop capacities for reflexive self-observation (capacity to know how the regime itself is performing) and dynamic responsiveness (ability to adapt rapidly to changes in the market and its own performance).

1.40  The shape of the new regulatory structures is by now settling into shape, though there is still a good degree of flux as each carves out its new institutional position relative to others within the institutional structure as a whole. At the global level there have been two key developments in the structure and management of the regulatory system: the extension and enhancement of the central coordinating body, the FSB, and the enhanced monitoring role of the IMF. At the EU level, the new regulatory structure bears marks of both evolution and revolution, and is likely to restructure the relationship between the EU and Member States in the area of financial regulation quite fundamentally. This section considers each of these in turn, asking whether and how these developments are likely to enhance the capacity of the global and EU regulatory regimes for reflexive learning and responsive adaptation.

1.41  Potentially the most significant change in the organizational structures of global financial regulation has been the reconstitution of the FSF into the FSB.⁷³ Its membership has widened to include the G20 countries and the European Commission. Its role is also changing from one of a loose coordinator to putative regime manager and regulator.⁷⁴ There have also been suggestions that it may break out of the usual communicative circle of regulators and international bodies and engage directly with financial institutions.⁷⁵

(p. 20) 1.42  The transformation of the FSF from a loose coordinator to regime manager and even direct regulator is not a straightforward one, though there has been a clear consolidation of the FSB’s position since 2009. The FSF, the FSB’s predecessor, was itself born out of a crisis, the near collapse and rescue of Long Term Capital Management (LTCM). Although seen as a significant crisis at the time, LTCM was a minor local difficulty in comparison with the events of 2007–09. The Forum was a delicately struck balance between the interests of national governments, national regulatory authorities, central banks, the international financial institutions (the IMF and the World Bank), the existing international committees of regulators, and various other inter-governmental bodies, notably the EU, ECB, and the OECD.⁷⁶ Its creation was in part a further step in the search for coordination between the existing international committees of securities, banking, insurance, and accounting regulators and other global actors. However, by introducing financial ministers directly into the structure it also forced politicians and regulators to confront each other at the international level.⁷⁷

1.43  The FSB’s Charter, agreed in 2009, clearly establishes it as the regime manager for the G20, and provides the FSB with a basis for building its own institutional position within the international architecture. It gives the FSB greater legitimacy as an overall coordinator and even standard-setter, at least for the G20 countries, but complicates its dynamics. It is not yet clear whether the FSB is an arm of the G20 (useful for the FSB in that it enhances its leverage) or a separate institution with its own institutional identity, position, and agenda (useful for it if it wants to apply its existing agenda to non-G20 countries or develop a distinct agenda of its own for G20 and/or non-G20 countries with respect to issues which the G20 has not asked it to address, or indeed if the FSB wants to sanction or even expel a member). Moreover, political tensions clearly exist between the members, though at present these are not paralysing. The FSB comprises politicians, financial regulators, and central banks from the G20, international organizations (World Bank, IMF, OECD), the EU Commission and ECB, pre-existing IRCs, and the IASB. The membership of the FSB was a political decision taken in the heat of the crisis, though, not the result of careful planning and consideration of what should be the criteria for membership. Notably, there is no consonance between the identity of G20 countries and the identity of countries which are of global financial systemic significance. Some G20 members are hosts to significant financial markets, others, such as Argentina, have almost none. There are other significant financial centres that are not members, such as Singapore. Further, whatever its composition, like other international organizations there will be tensions between members as each (p. 21) national government pursues its own national agenda and interests. In addition, the financial markets take a different form in each country, and have different lobbying powers, both of which shape the policy agendas each state is pursuing at the international level.

1.44  Finance ministers introduce some semblance of democratic engagement into the global regulatory regime and are shaping the global regulatory agenda in a way that they have arguably not done since the decision to set up the first Basle Capital Accord.⁷⁸ However, this can produce tensions with the international committees of regulators who find their technocratic world now politicized in unpredictable, and for some, such as the IASB, uncomfortable ways. Indeed the IASB chairman stated he considered resigning in protest against the political pressure being place on the IASB by the EU and in particular French President Nicolas Sarkozy, to amend IAS 39, which provides rules on fair value accounting for financial instruments.⁷⁹ Furthermore the presence of both international organizations and a selective group of their members can create uneasy dynamics. The de Larosiere Report, for example, indicated frustration that some of the EU Member States have a separate voice at the international level and can use these international fora to oppose and outmanoeuvre it. It argued that only the EU should be represented in international organizations, including the FSB,⁸⁰ though this is not a suggestion that has been adopted with the alacrity of some of its other recommendations. Others have criticized the over-representation of EU Member States, though for different reasons.⁸¹

1.45  The FSB has to manage these tensions. Its reformulation as an independent actor is one that is in its infancy, and the process of developing its own identity as an organization, distinct from the aggregate views and interests of its members, will evolve over time. It has clearly evolved considerably over the last two years, as discussed below, and this evolutionary dynamic is likely to continue. As the analysis of policy shifts above suggests, it is expected to develop a role in systemic surveillance, together with the BIS and the IMF. It is also developing a greater role as a standard-setter, formulating principles for regulators to implement. It is also actively developing a role in promoting and overseeing the implementation of international financial regulation by national governments, both G20 and non-G20 Member States, and broadening participation in its deliberations to facilitate that role.

1.46  The international financial regulatory committees have been enthusiastic standard-setters since their various inceptions. The principles and rules they issued were directed at Member State regulators and focused both on the organization and operation of regulation by national governments and their regulatory authorities, including central banks and on the regulation of market actors. The FSF was content largely to leave the function of standard-setting to the existing IRCs, principally IOSCO, the BCBS, the IAIS, and the IASB. It confined itself to compiling a compendium of a selected set of their principles to be monitored under the FSAP process. However, since its reconstitution in 2008 it has become a standard-setter itself. It issued two sets of Principles in its first six months (on remuneration and cross-border crisis management), more than it did in the last 10 years in its old formulation as the FSF.⁸² It is has also issued a set of ‘key attributes’ for resolution regimes and at the time of writing is consulting on principles for underwriting practices for residential mortgages.⁸³

1.47  It may be thought that in the new settlement it is the FSB which is the lead principles provider, but the dynamics of principles-development are worth examining more closely. For example, the role of the FSB in issuing the remuneration principles could be viewed as simply a pragmatic allocation of functions between the different organizations that are involved in the international financial regulatory regime. They could have been issued by any or all of the other sectoral regulatory organizations, but as they apply to all financial institutions it makes sense for the FSB to issue them. However, it is suggested that there is a deeper significance to this development. It is important to recognize that principles are not simply regulatory instruments used in an attempt to affect behaviour. First, they can have symbolic significance: they are used to establish their authors’ own institutional position within the regulatory regime. Secondly, they have a broader functional role: not just to regulate the behaviour of market actors but also that of other national regulators. Thirdly, the increased focus on implementation means that principles are increasingly being used as benchmarks of performance against which national regulatory regimes are assessed, and thus as criteria of accountability.⁸⁴ Consequently, the development of the FSB’s role as standard-setter is a step towards establishing its position as overall coordinator, manager, and enforcer of the international regulatory regime. The FSB has also begun to adopt a clearer coordinating and project management role in the development of principles by the other IRCs. For example, in late 2011 it initiated a project on shadow banking with work streams distributed between BCBS, IOSCO, (p. 23) and FSB, with BCBS and IOSCO charged with developing policy recommendations on securitization in the shadow banking sector for FSB approval.⁸⁵ This is the first time that such a coordinated standard setting process between the IRCs has been introduced, and the positioning of the FSB as project manager indicates the shifting tectonics of the global regulatory landscape.

1.48  However, there are limits to the extent to which the FSB is ‘head of global’. The process of developing principles, for example, is not necessarily a linear, hierarchical, ‘flow down from the top’ process, but more complex. The members of the IRCs are national regulators who have considerable rule-making autonomy in their own jurisdictions (at least outside the EU). The development of principles on remuneration provides a good early post-crisis illustration of the dynamics of ‘principles production’ in this polycentric system of financial regulation, and of the challenges facing the FSB if it wants to become ‘head of global’. Despite the rhetoric of the need for greater international coordination and harmonization, the United Kingdom’s FSA was a ‘first mover’, declaring that it was prepared to act unilaterally. It issued its draft code on remuneration in February 2009. This was followed by the communiqué from the G20 that principles governing remuneration should be developed. As a consequence, the FSF (as it still was) issued its Principles on Sound Compensation Practices in April 2009, followed by Implementation Standards for the Principles in September 2009. Two of the international committees of regulators, IOSCO (securities) and the BCBS (banking) then began to develop proposals to implement these Principles. Separately, regulators in Australia, Switzerland, France, and the Netherlands published principles on remuneration practices, broadly following the FSF’s, and the United States indicated its intention to do so. Meanwhile, various initiatives were emanating from the EU. The CEBS issued the final version of its high-level principles on remuneration just after the FSB in April 2009, and the Commission published a recommendation,⁸⁶ and draft amendments, to the Capital Requirements Directive (CRD). These were formalized and sent to the EU Parliament and Council in July 2009.⁸⁷ None of these principles aligned exactly with one another.⁸⁸

1.49  So the production of the remuneration principles was not the simple ‘flow from the top’ process that it might appear. Principles were being produced simultaneously and in parallel with only patchy coordination. The process demonstrates the (p. 24) ‘multi-authorship’ of principles, and the multiple roles that individual regulators play in each of these rule-writing fora: the FSA, for example, is both contributing author of and subject to the principles emanating from the FSB, IOSCO, BCBS, and the European regulators. However, it may be more influential in some of these fora than others with respect to different issues, and so, like others, will seek to exploit the variable geometry of the global regulatory structure to its best advantage. The process also illustrates the difficulties of system management and the lack of coordination that can characterize norm formation in polycentric systems, as each regulator wants to develop its own norms to suit its own local conditions and priorities.

1.50  Setting principles may be a necessary part of regulation but it is by no means sufficient. As all regulators know, the greater challenge is ensuring they are complied with. Implementation has traditionally been left to the IMF FSAP process, discussed below, but the FSB and the IRCs are now paying much more attention to monitoring the implementation of their own principles. Greater transparency may also be creeping in, for implementation of international standards by national governments is no longer seen by the international financial regulatory community to be a matter that can remain private to that government. Rather, it is information that all need to know, as the financial stability of one country can be dependent on the regulatory performance of another.

1.51  From a global, system-wide perspective, it is thus significant that the G20 Member States have now agreed to submit their national financial regulatory regimes to the FSAP process, and for those reviews to be more in depth. There is also renewed impetus from the IMF for all systemically significant countries to be reviewed regularly, not just those in the G20. However, the drive for all members to be reviewed and for reports to be published will require some changes in the attitudes of national governments to the review processes: some have already demonstrated that they are more open to criticism than others. Three of the new FSB members had not previously agreed to being assessed: Indonesia, China, and the United States. Reviews of them are currently underway. Not all have previously agreed to have their reports published (eg Brazil and Russia for its 2008 report). Ten countries had only agreed to have their reports published in summary form, including Germany, Mexico, and Japan. All are now required to agree to publication.⁸⁹

1.52  Under the FSAP processes, the FSF’s role was limited to assembling the compendium of principles against compliance was to be assessed as part of the wider FSAP review. In contrast, the FSB is now being positioned by the G20, and is positioning itself, as a central actor in ensuring that standards are implemented. To achieve (p. 25) this, the FSB has set up two systems of peer review. The first is a country-based review of implementation of FSAP recommendations to G20 member countries, and second on a thematic basis, for monitoring implementation of the FSB’s standards, again by G20 countries.

1.53  Under the FSAP follow-up review process, the initial focus of the FSB is on jurisdictions that could pose a risk to financial stability because of their importance in the financial system and their weak adherence to the relevant standards (though note that non-G20 countries can also be systemically significant). By early 2012 it had completed country peer reviews of Italy, Spain and Australia, Canada and Switzerland. Under the thematic review process, by this time it had also conducted reviews of risk disclosure practices, deposit insurance systems, and a follow-up review of compensation practices.

1.54  The added value of the peer review process, the FSB hopes, will come from ‘the cross-sector, cross-functional, system-wide perspective brought by its members’ and from dialogue with peers’.⁹⁰ Peer review reports and any commentaries provided by the reviewed jurisdictions will be published. Following publication of the report, jurisdictions’ implementation of agreed actions will be monitored by the FSB and, if implementation lags, capacity-building mechanisms will be made available through technical assistance and peer pressure may be applied.⁹¹

1.55  There is an ambiguity in the role of peer reviews, however, as to whether they are primarily aimed at providing technical assistance or whether they are aimed at ensuring compliance. The dynamic of the peer process can be such that countries are reluctant to criticize each other for they know that they have their own weaknesses. Nonetheless, the FSB’s stance at present is tending towards the latter. However, the FSB has no direct means of promoting or enforcing compliance. It is nonetheless developing a ‘toolbox’ of positive and negative measures that could be deployed.⁹² This comprises a combination of carrots and sticks (and some orange-painted, carrot-shaped sticks). At the softer end are ‘compliance’-based strategies such as advice and technical assistance. Moving up the sanctions pyramid,⁹³ borrowing from IOSCO, the FSB suggests that those who are not adherent should sign a multilateral Memoranda of Understanding (MoU) on information sharing and cooperation as a condition of membership of various international bodies or their working groups (but note that it does not include membership of itself in its list of examples, raising questions as to the extent that the FSB as an organization separate from the G20 can determine its own (p. 26) membership).⁹⁴ It suggests that national regulators could vary their stance to financial institutions that are based in non-compliant countries, for example in making decisions relating to market access, or enhancing their supervision of such institutions in a number of ways, including higher capital requirements. At the top of the pyramid it suggests that the IFIs could consider withholding financial assistance from such countries. The toolbox thus contains a wide range of tools, but the FSB can exercise very few of these itself, other than ‘naming and shaming’. Whether in practice there is the political will to adopt these measures when they are needed (and whether those who are named will in fact be shamed) remains to be seen.

1.56  There is also the question of what use is made of the information collected on implementation: whether FSB is there to collate and disseminate information, as it has begun to do through its Implementation Monitoring Network with respect to FSAPs, or to take a more active role in analysing and acting on it.

1.57  The second, more recent peer review mechanism, is directed at the latter function. It is the Coordination Framework for Implementation Monitoring (CFIM), set up in 2011 in collaboration with the other IRCs.⁹⁵ The objectives of the FSB’s CFIM are to ensure that implementation monitoring processes are comprehensive, rigorous, and timely, and promote the overall coherence and consistency of implementation across sectors and functions; to generate comprehensive and consistent information on the nature and pace of implementation of agreed G20/FSB recommendations so that it can be reported to the G20 and to the public; identify and help to reduce impediments to, and gaps in, implementation by providing political impetus and leveraging peer pressure; and identify useful lessons from experience on the effectiveness of the policy reforms and the associated standards as well as on any unintended consequences.⁹⁶ Priority areas for review are decided by the FSB Plenary, and initially are to focus on Basel II, II.5 and Basel III implementation through the BCBS’s Standards Implementation Group; a Peer Review Council for monitoring consistent implementation of global SIFI (G-SIFI) policies; a coordination group on OTC derivatives reforms; and ongoing monitoring on the implementation of compensation principles and standards. In addition to detailed monitoring, the FSB Secretariat, in consultation with FSB members, will provide an annual scoreboard to the G20 Leaders tracking progress across the full (p. 27) range of reforms.⁹⁷ In addition, the FSB will monitor implementation of all G20/FSB recommendations through the Implementation Monitoring Network, which will be upgraded for this purpose and will remain the FSB information hub on implementation progress.⁹⁸

1.58  The Framework is intended to add an important coordination and analytical function to existing systems for information gathering and for implementation reviews by the international regulatory committees. Critically, it puts the FSB as the head of a line reporting process up to the G20. To this extent, it is a mechanism which not only performs the stated function of improving implementation but it also reinforces the FSB’s role as system coordinator.

1.59  Over the last two years (2009–11) the FSB has thus taken some significant steps to define and reinforce its position as ‘head of global’. So far, it has used its position as the arm of the G20 to bolster and justify that role. However, this reliance on its G20 mandate, whilst currently empowering the FSB, may over time prove to be a source of weakness. The Framework document makes it clear that, at least for the moment, the FSB’s mandate stops at the limits of G20 policy. It has no role in monitoring and reporting on other standards of the IRCs nor of FSAP reviews to the extent they cover issues outside the G20 mandate.⁹⁹ The Framework thus establishes a clear demarcation to the limits of the FSB’s institutional position as leader and coordinator of the international regulatory regime. It is likely that the FSB will continue to build itself a role as central coordinator and ‘head of global’ over the next few years based on the position the G20 has given it, but it will be interesting to see the extent to which it can build this position on its own account, expanding its role in practice beyond that bestowed on it by its Charter.

1.60  Indeed, there is a tension, currently implicit, between the FSB’s role as agent of the G20 and the FSB’s role as coordinator of the global regulatory structure independent of the G20 (notwithstanding its membership). As the initial post-crisis appetite and energy of the G20 governments to address and agree on issues of financial regulatory policy wanes, the FSB may well find itself wanting to adopt policies on its own initiative and, critically, with respect to non-G20 countries.¹⁰⁰ There may be important functional reasons for it to do so. Critically, there is not a perfect overlap between G20 membership and global systemic importance. It is notable that the IMF considers four G20 jurisdictions not to have systemically important (p. 28) financial systems and it considers a further ten non-G20 countries to be systemically significant. It is therefore introducing mandatory FSAPs for them.¹⁰¹

1.61  It will be interesting to see whether the FSB broadens its role to follow up on these non-G20 FSAPs. More generally, it would be unfortunate, to say the least, if the current constitution of the FSB meant that it was unable to fulfil its role of maintaining global financial stability because it is able to focus only on those countries which happen to be members of the G20 and not others which are systemically significant. If it is to extend its remit to cover non-G20 countries, it will have to adjust the basis on which it claims that it is legitimate for it to do so, and not tie its legitimacy so closely to the G20 mandate. It has recently broadened consultation beyond G20 countries, which could expand its legitimacy base. For example, in 2011 it set up six Regional Consultative Groups comprising FSB member and non-member countries to exchange views on vulnerabilities affecting financial systems and on initiatives to promote financial stability. However, to what extent it can establish itself as ‘head of global’ rather than ‘arm of the G20’ remains to be seen.

1.62  In short, the FSB has a delicate institutional balance to strike. The more the FSB ties the legitimation of its position as ‘head of global’ to its role as an arm of the G20, useful in these early days, the less it may find it can act independently of it, which could hamper the FSB in the future.

1.63  The FSB is not the only body which has been shifting its institutional position in the architecture of global financial regulation. The IMF has also been repositioning itself. The IMF is not a new actor in the financial regulatory regime. It was a founding member of the FSF. As noted above, with the World Bank it has been monitoring countries’ compliance with certain principles of regulation issued by the international committees of regulators as part of its FSAP since 2002. Consistent with the IMF’s mandate, these reports were focused on individual country compliance. They did not consider the systemic implications of one country’s regulatory regime.¹⁰² The IMF also performed macro-economic surveillance. However, in a reflection of the pre-crisis cognitive and institutional divide between macro-economic surveillance and micro-prudential supervision, there were no mechanisms in place for their reports to be integrated into the processes of banking supervision.

(p. 29) 1.64  In the wake of the crisis, the expectations of the role the IMF should play in the global financial system have increased. In particular, the FSAP process has been enhanced and in September 2010 the IMF Executive Board agreed to make it mandatory for 25 jurisdictions with systemically important financial sectors to undergo financial stability assessments under the FSAP at least every five years as part of the legally based bilateral surveillance activities of the IMF under Article IV.¹⁰³

1.65  This is a significant shift, as previously FSAPs had been a voluntary form of technical assistance, and indeed remain voluntary for other countries. The crisis made it clear that the FSAPs needed to be enhanced, and that there also needed to be much closer integration between the FSAP and the IMF’s other Article IV assessments.¹⁰⁴ However, it also revealed the gap between the IMF’s legal mandate and the new role expected of it, and in 2010 the IMF had launched a mandate review, arguing that its legal powers and purposes were addressed at ‘matters important to a bygone age’, trade and balance of payments, and based on a bilateral relationship of the IMF with individual members.¹⁰⁵ Ultimately, the solution was finessed through an expansive interpretation of the IMF’s mandate and an agreement that financial stability assessments are an important part of the bilateral relationship of countries with the IMF, obviating the need for legal changes.

1.66  Other issues still remain, however. The IMF has only limited and episodic access to supervisory data (eg in the context of FSAPs), and members have in the past declined to provide systemically relevant information on grounds of confidentiality.¹⁰⁶ Further, the Fund has no legal authority to require confidential data on entities such as large complex financial institutions¹⁰⁷ Finally, there is a resourcing issue. The IMF is introducing a strong risk-based element to the assessments, moving to having shorter, ‘modular’ assessments on particular issues and greater use of cross-country thematic reviews to facilitate comparability.¹⁰⁸ In part this is to facilitate better management of resources, but nonetheless (and illustrating one of the inevitable consequences of risk-based monitoring systems), non-systemically important countries are concerned that they will be ‘poor relations’ and receive less technical assistance as a result of the increased focus on systemic jurisdictions.¹⁰⁹

(p. 30) 1.67  The IMF’s mandate review also illustrates the uneven role that law plays in the international financial regulatory system and its implications for regulatory capacity. The FSB has been reconstituted without need for any legal debate or treaty changes, but in contrast does not have hard-edged legal powers that it can deploy. The IMF has some considerable legal powers, but in order to fulfil its new role it has to engage in a series of creative interpretations in order to avoid the difficult process of amendment to its Articles.

1.68  The radical restructuring of the institutional structures of EU financial regulation is an excellent example of how not to waste a good crisis. Indeed the Larosiere Report, on whose recommendations the restructuring is based, admitted that its proposals were more about the enhancement of the European single market than they were a response to the financial crisis. Nevertheless, it did argue that uncoordinated crisis management action had led to negative spillover effects which needed to be addressed.¹¹⁰

1.69  The authors of the new EU regime have extensive ambitions to establish the EU authorities as the central regime managers of the network of financial supervision within the EU, but are approaching the task in a very different way than the FSB or even IMF. This is partly because they can: the EU has the institutional and legal structures that it can deploy to implement its strategies. Its authors are searching for control, and for the elimination of discretion, variety, and difference in financial supervision across Member States.

1.70  The legislative reforms introduced in 2011 are discussed elsewhere in this book, but briefly, they established a European level body charged with overseeing risk in the financial system as a whole, the European Systemic Risk Board (ESRB) and create a European System of Financial Supervision (ESFS).¹¹¹ The ESRB is to monitor and assess potential threats to financial stability that arise from macro-economic developments and from developments within the financial system as a whole (‘macro-prudential supervision’) and, where necessary, issue recommendations for action to deal with these risks. The three new, legally constituted, European Supervisory Authorities (ESAs) replace the previous Level 3 committees and have powers to issue binding technical standards, subject to a fairly complex rule-making procedure. A central aim of the creation of the new system is to create a single rulebook for the whole of the EU that is consistently interpreted and applied throughout the EU by all Member State financial regulators. This process is to be strengthened by introducing more directly applicable rules at the EU level wherever possible.

(p. 31) 1.71  As with the FSB and IMF, a key tool is to be the peer review process, discussed in Niamh Moloney’s chapter in this book.¹¹² Peer reviews were introduced by CESR in 2003, and institutionalized through the creation of a Review Panel. CEBS followed suit in 2007 and CEIOPS in 2008. The reviews were largely formal assessments of powers and practices, with an orientation to promoting best practice. It is clear that the significance of the peer review process is intended to increase and that they will form a central mechanism of coordination, not simply mechanisms of providing technical assistance. The ESAs are legally required to conduct regular peer reviews, though national authorities are not formally bound to follow the ESAs’ recommendations but are to ‘endeavour’ to do so.¹¹³ Again, exactly where on the spectrum between information sharing and technical assistance at one end, and enforcement on the other, remains to be seen.

1.72  The crisis has created a series of ‘cognitive shocks’ for financial regulators, prompting a series of policy changes and, of main focus here, of organizational restructurings and realignments. There are clearly a number of challenges going forward, but the remainder of the discussion focuses only on three: system coordination, reflexive self-observation or ‘learning’, and responsiveness.

1.73  One of the key challenges of polycentric regulation, particularly when it is multi-level, is coordination or ‘orchestration’.¹¹⁴ In the financial sphere, the question is how best to manage a complex, interlocking network of regulatory actors and their relationship with an even more complex, dynamic, and interlocking financial system. Both the FSB at the global level and the EU authorities at the regional level are attempting to become ‘regime managers’, and in some respects are being positioned in this role by other actors in the regulatory arena. However, this is a far from straightforward trajectory, particularly for the FSB which, unlike the EU authorities, cannot rely on a complex institutional infrastructure and legal fiat to accomplish this task. But with respect to both the FSB and the new EU regulators, the issue of ‘who governs’ remains contested and is underlain with a number of tensions and conflicts.

(p. 32) 1.74  There are four main types of institutional structures for control: hierarchy, community, markets/competition, and managed networks.¹¹⁵ The FSB is attempting to use three of these—hierarchy through the promulgation of rules and a system for their enforcement; community in attempting to deploy peer pressure, and peer cooperation through support and capacity building; and network management strategies in that it is attempting to facilitate concerted actions by developing or steering processes that either encourage negotiations and interactions or foster the conditions for collective behaviour by building levels of consensus to points that allow for action on a given issue, for example as in cross-border crisis management.

1.75  The EU is opting only for one: hierarchy. Although there are indications that the system should in practice be managed on a ‘hub and spoke’ basis, the ESAs’ powers to settle disputes among national financial supervisors, to impose temporary bans on risky financial products and activities, and to directly prevent or remedy any breaches of EU law by financial institutions should national authorities fail to act, mean that the ‘hub’ has overall powers of control to ensure the spokes do not fly off. The only strategy not being used strongly by either is markets: there is little faith at present in the idea that coordination can be achieved ‘through the “invisible hand” of the self interest of participants’ who are willing to exchange resources and conclude agreements in order to attain mutually beneficial solutions and higher levels of collective welfare.¹¹⁶ The crisis has shown the limits of that approach as a strategy for management of an international regulatory regime.

1.76  A difficulty is that each of these modes of control has quite distinct logics that are hard to combine, and often a demanding set of institutional preconditions for them to operate successfully which are simply not present. Hierarchies are top-down institutional and control structures in which a central control body lays down rules and policies that provide direction to the network of inferior institutions. This works best if there is a high degree of organizational integration from top to bottom, there is a clear and commonly understood mandate, and the control body has the authority, tools, and capacity to organize the network.¹¹⁷ Hierarchical controls work less well where networks are loosely constituted and where the member organizations are numerous, independent, divergent in characteristics, and oriented to different objectives.¹¹⁸ Hierarchies also depend on recognition by those within them that there is an apex, and that the regulatory process is one of (p. 33) implementing goals formulated ex ante by a central authority. They struggle in a context in which there are a number of constellations of actors operating in the same policy space,¹¹⁹ and moreover where participants see the nature of the task not to be one of rule implementation but as an interactive process of exchanging information about problems, preferences, responses, and trade-off goals and resources.¹²⁰ This is a message that EU policy-makers would do well to reflect on.

1.77  Community controls are based on recognition of a stable group of peers who share a common set of interests, and whose main concern is to remain accepted by the community. Self-interest is either emasculated or is recognized to be dependent on or identical to the community interest. Regulatory strategies based on peer pressure and peer support work less well where participants have conflicting goals, objectives, and interests and have motivations to compete with one another, and where non-participation in the group is not fatal to the survival of the outsider. There are aspects of the international and EU regulatory system where these institutional structures are present, but they are by no means pervasive.

1.78  Network management strategies are directed towards harnessing the different control capacities of the involved actors,¹²¹ or developing processes that change perceptions and allow collective actions to be taken. Network management strategies are arguably better suited to polycentric regimes in which diffused regulators interact and bring different capacities to bear on issues without there being any hierarchical linkages but instead there are wider dispersions of authority to a number of autonomous bodies who may only be loosely coordinated. These strategies involve creating a governance mechanism for the network which establishes agendas rather than common goals; creates communications channels; makes ad hoc arrangements to support collective action; brokers solutions by bringing problems, solutions, and parties together; promotes favourable conditions for joint action; or manages conflicts through mediation and arbitration.¹²² Mediation processes may, indeed, involve the establishing of units dedicated to liaising between different actors with the aim of effecting bridges between parties with different interests and orientations so as to build relationships and inter-organizational trust.¹²³ However, for the network manager, the problems are that network management strategies are messy, complex, fluid, and their outcomes uncertain. There are strong arguments that it is nevertheless strategies of network management that are needed to (p. 34) regulate complex adaptive systems such as these; that strategies have to be adaptable, dynamic, and match variety with variety. However, these are not messages that governments and regulators currently want to hear.

1.79  Despite the differences in their formal organizational structures, both the FSB and the EU authorities face a number of similar challenges in trying to position themselves as hierarchical ‘leaders’. First, both are operating in a crowded policy space. In many respects the FSB is competing with its members as a standard setter, and potentially as the guardian of implementation. Further, its members have significantly different regulatory capacities, enhancing the dependence of some on the FSB, but reducing that of others. The OECD, for example, is here a ‘norm entrepreneur’—a body that issues principles or standards in the hope that others will adopt them and implement them. The OECD has issued its own principles for financial regulation, but they need adopting by the FSB if these principles are to be integrated into the peer review process either of the FSB or of the IMF/World Bank.¹²⁴ In contrast, the EU has far greater regulatory capacity as it has legal powers to act unilaterally, as of course do individual nation states who are members of the FSB.

1.80  However, as a standard-setter the EU is also operating in a crowded policy arena. It is competing with international regulators and, to the extent it has left them any room to create their own rules, with Member States. Some Member States are making as much use of this remaining room for policy independence as they can, whilst they can. Moreover, through their own membership of the international committees of regulators its Member States can leapfrog the EU, exercising influence in the decisions made in other fora which the EU then has to decide whether or not to accept. So although the EU can be a critical and selective adopter of transnational standards, as its approach to the IASB’s rules indicates,¹²⁵ in crafting its rules it is often working on a canvas on which the main outlines have already been drawn by others.

1.81  There are differences between the challenges the FSB and the EU Commission face, however, arising from the differences in the organizational infrastructure in which they are situated and in their regulatory capacities. As discussed above, the (p. 35) FSB is still in an ambiguous institutional position. It may evolve to be the overall coordinator of the international regulatory regime, but it faces the main problems of network management: conflicting objectives and interests of participants; difficulties of surveillance over the regime; the use of regulatory tools with conflicting logics within the regime; differential capacities and resources of participants; lack of its own resources and capacities to effect change or manage the network; and the difficulty of getting recognition from participants that they are in a network that should be managed and that the manager is the FSB. A critical issue will be success in getting members and non-members to implement the principles that the FSB issues. For this, countries and other members of the FSB have to afford it sufficient recognition of its legitimacy and authority to govern. It also has to be able to exercise sufficient leverage through peer pressure or other means to get compliance from members, and more problematically perhaps, from non-members. None of these are straight-forward challenges to meet.

1.82  Further, the FSB (and indeed the EU) still faces the fundamental issue that it is national governments who, at least under present arrangements, underwrite the financial institutions operating in their jurisdiction. The ‘mortality mismatch’ of financial institutions, combined with an understandable policy of fiscal protectionism, means that national governments are deeply unwilling to confer sovereignty to international bodies or to sign up to common standards or modes of supervision if these would compromise their ability to protect their fiscal position in times of crisis.

1.83  The EU authorities also face a number of challenges. As in the case of the international regulators, a critical issue is not just the creation of standards but their implementation. The EU can use legal fiat to achieve its objectives, but it would be a mistake for it to think that the use of legal powers removes the need to actively build legitimacy and authority with respect to Member State regulators. It would also be a mistake to assume that complete harmonization in the interpretation and implementation of its rules is either attainable or desirable, though all the indications are that this is exactly what is being attempted. Although the Authorities now have legal powers to take action against infringers, this should be used as a last resort. The focus should be on building recognition and authority independently through their processes and day-to-day operation. Here the conflicting rhetoric and powers create a further ambiguity, however. The insistence on complete harmonization cannot easily be reconciled with rhetoric that it should be created ‘fully respecting the proportionality and subsidiarity’.¹²⁶

1.84  Moreover, there are conflicting objectives both at the EU level and between the EU and Member States. The structure itself is still ‘siloed’, as at the international level, between securities, banking, and insurance, and with an institutional divide between macro and micro-prudential regulation, reinforcing at the organizational (p. 36) level the cognitive constructions of financial markets which the crisis has shown to be flawed. Ensuring coordination between these different actors is clearly essential, but all too often each organization can remain bounded by its own job description and institutional focus.

1.85  Furthermore, the new EU Authorities have a set of legal powers but very little operational or regulatory capacity to effect change or manage the network of Member State regulators, although the rapid increase in their staffing levels may enhance their capacity in this respect. Nonetheless, they are likely to continue to rely heavily on a small number of Member States for expertise. This ‘asymmetric Europeanization’ is a familiar feature of the operation of other EU regulatory agencies, and there is little reason to think the regime for financial regulation will be any different.¹²⁷ Further, although again the legal structures create a clear legal hierarchy for managing the regime, this does not on its own resolve the problem of how to gain recognition from participants that they should be managed, and that the appropriate managers are the Authorities.

1.86  Moreover, the policy dynamics created by banks’ mortality mismatch and the politics of fiscal protectionism exert a significant centrifugal force at the EU level, just as they do at the international level.¹²⁸ The Larosiere Report’s insistence on harmonization was shot through with exceptions to allow Member States to act in the interests of national financial stability, and these have been embedded into the new regulatory structures and processes. The passporting regime for banks, the central plank of European financial integration, has been put into question, and as noted above has been strongly criticized in some quarters as providing the channels through which financial instability can spread.¹²⁹ The focus now is less on the ability of banks to do cross-border business, and far more on the protection of national deposit holders and the national fiscal position. Without an EU-wide bank resolution fund, however, whilst national taxpayers are in line to bail out national banks, national regulators will want to maintain as much control as possible and it is right that their governments should want them to do so, not least because it is national governments, not EU institutions, that are democratically accountable to those same taxpayers.

1.87  Finally, the EU regime itself sits in an interesting position within the global regime. To the extent that the EU is attempting to create a single system of financial regulation, it creates a potentially ambiguous position for itself in the international regulatory sphere, one which raises interesting questions. If the EU institutions are not just coordinators but are themselves the key regulators in the EU system, should they not be subject to peer review monitoring by the FSB, the BCBS, or by the IMF (p. 37) (albeit under a revised IMF mandate)? Arguably they should, if they are really to be setting the single rulebook and single set of supervisory strategies. This may not be a process the EU institutions would be entirely comfortable with, however. This would create an interesting dynamic between EU institutions and their Member States, who may welcome an FSB/IMF assessment of the ESAs, particularly if they think there are deficiencies in the EU regime. They may yet use their individual membership in the global bodies to push for one.

1.88  In addition to the restructurings outlined above, there has been a move to strengthen multi-lateral regulatory arrangements for the regulation of firms themselves. In the past, the main focus of the international regulatory efforts has been Member States: this has been a ‘government regulator to government regulator’ conversation, not regulator to private market actor.¹³⁰ One of the weaknesses of ‘regulator to regulator’ conversations is that they can be too removed from what is happening in the markets. A key difficulty for regulators is asymmetric information: that although they are in a better position than individual market actors to have aggregate, system-wide information, they lack the detailed, ‘on the ground’ information that the latter can have.¹³¹

1.89  A further difficulty is ensuring adequate information flows between different parts of the regulatory regime. It has been suggested in other contexts that building resilient systems of information and surveillance, particularly in conditions of uncertainty and which are subject to rapid change requires multi-scale networks, in which different types of information is gathered, aggregated, and tested at different points within a network.¹³² The global, including EU, financial regulatory regime has a number of points at which information is currently collected about firms. In addition, national regulators have generally good arrangements in place for exchanging information with respect to individual firms in the context of enforcement processes. IOSCO, in particular, has required all members to agree to multilateral MOU on information sharing.¹³³ However, the crisis demonstrated that the regulatory structures at the EU and global level ensured that, to the extent they occurred at all, these conversations took place within regulatory silos of (p. 38) insurance, securities, and banking, with significant impediments to information flows between them.

1.90  However, as yet there is only a weak linking between these regulatory silos and other parts of the regime at both global and regional (EU) level, and indeed between the national, regional, and global levels. The FSB’s Framework is aimed at resolving part of this problem. Nonetheless, as noted above, it is at present limited to monitoring implementation of G20 mandated policies. Gaps still remain. Three notable examples are the weak linkages between colleges of supervisors at all levels, between colleges and the peer review process, and between both colleges and peer review processes and the standard-setting processes of the IRCs and, potentially, the different European authorities.

1.91  Colleges of supervisors have the potential to become critical nodes in the coordination of both global and EU financial regulation, not least as significant points of information-gathering and dissemination, though again both their dynamics and their position in the institutional topography are not unproblematic. Supervisory colleges are at present the main place where both ‘regulator to regulator’ and ‘regulator to firm’ conversations occur outside the national setting. In formal terms, international colleges of supervisors (ICSs) are permanent, although flexible, structures for cooperation and coordination among the authorities responsible for and involved in the supervision of the different components of global cross-border banking groups. They have also been more recently developed for insurance companies, and for conglomerates.

1.92  Supervisory colleges are intended to provide a forum intended to facilitate sharing of information, views, and assessments among supervisors in order to allow for a more efficient and effective consolidated and solo supervision and timely action; enable supervisors to develop a common understanding of the risk profile of the group as the starting point for risk-based supervision at both group and solo levels; coordinate supervisory review and risk assessment, establishing supervisory plans, arranging any division of tasks and joint onsite visits, thus avoiding duplication of work and reducing the regulatory burden; and coordinate decisions taken by individual authorities. They are also intended to facilitate the development of a consistent interpretation and application of regulatory provisions across the group and build trust between supervisors.¹³⁴ They are meant to work alongside firm-specific cross-border crisis management groups under the FSB’s Principles for Cross-border Cooperation on Crisis Management.¹³⁵

(p. 39) 1.93  The notion of colleges of supervisors for banks pre-dates the crisis, but their crisis gave significant impetus to their development, and they have now been mandated by the G-20 governments for global financial institutions,¹³⁶ and within the EU under the revised Capital Requirements Directive. At the global level, the BCBS has been actively promoting colleges as important not only for consolidated supervision of banks at the microprudential level, but also for the promotion of financial stability at the macroprudential level, and has issued a set of Principles for how they should operate.¹³⁷ Notably, there is also a move to promote joint on-site inspections by members of colleges by all the IRCs.¹³⁸

1.94  Within the EU there is a separate requirement under EU law for EU Member State supervisors to form EU level supervisory colleges.¹³⁹ The EBA is an observer in the EU colleges, and has established a Review Panel to assess the degree of convergence reached by members in the implementation of EU legislation and to monitor convergence in supervisory practices. Reviews are conducted by the Review Panel based on self-assessments provided by its members on a ‘comply or explain’ basis.¹⁴⁰

1.95  A global bank can thus have two colleges, the international and the European, raising further issues of coordination between the colleges. The EU home supervisor is required to ensure that non-EEA supervisors contribute to the EU college’s assessment of major risks within the banking group. However, there is at present no systematized process or structure in which the two colleges can, or are obliged to, confer or exchange information with respect to the same financial institution, or indeed merge to operate as a single college when supervising a cross-border bank; instead, coordination occurs through one or two members of the EU college attending meetings of the international college.¹⁴¹

1.96  On the whole both EU and global colleges have been positively received by the banks and others,¹⁴² although there are several issues to be resolved in the way the colleges operate.¹⁴³ The structure of the college is the responsibility of the home state, and colleges can have a variable geometry. In order to function effectively, (p. 40) colleges require a high degree of understanding and strong levels of mutual trust and confidence between national supervisors. They also require the ability for supervisors to share information, but thus far this has proved to be a significant stumbling block, partly due to issues of trust, but also in large part due legal restrictions in Member States on sharing data. There are also concerns that once given the information will not remain confidential, either due to legal provisions in national freedom of information legislation which may require its disclosure, or due to poor practices on the part of those to whom it has been disclosed. Moreover, although firms may be multi-functional, regulation at the global and EU levels remains siloed into securities, banking, and insurance. In the EU, only the banking and insurance supervisors have coordinated in developing principles.¹⁴⁴ There is even less coordination apparent at the global level.¹⁴⁵ There is the additional question of who should be the prime author of principles for supervisory colleges. The ESAs are not the only actors in the regulatory space, and as noted, the BCBS has set its own principles. The FSB has not led this process. However some, notably the European Banking Federation, argued that it should be the FSB who takes the lead in setting principles for the colleges’ operation, though drawing on EU principles.¹⁴⁶ Again, this illustrates the complexity of the Principles production process within the international regulatory framework.

1.97  There is thus still considerable fluidity as to the role and effectiveness of individual colleges and their coordination. However, supervisory colleges could be critical ‘nodes’ in the system of international financial regulation. It is thus important to ensure, first, that they function effectively, and secondly, that the knowledge they have a development at the local level of individual institutions is captured elsewhere within the system. In particular, it needs to be integrated into other activites such as peer review and standard setting, discussed below.

1.98  The regime for financial regulation at both global and EU levels thus faces a significant set of challenges with respect to how it manages itself, let alone how it attempts to regulate market actors. Orchestrating the terms of cohabitation between participants is difficult where there is no completely agreed score for each to play, little agreement over who should be the conductor, only weak consensus as to whether (p. 41) all need to follow him, and only muted ability of each to hear what those in other sections of the orchestra are playing. Faced with such challenges, issues of how to ensure the system is able to anticipate future crises or challenge conventional wisdoms can appear mere fripperies: nice to have, but there is more important business to be getting on with.

1.99  This reaction is understandable, but short-sighted. Regulators and others are struggling to create a new cognitive framework in which to develop policy responses. Part of that process requires gaining knowledge. Gaining knowledge requires acquiring information and making sense of it through a cognitive framework. However, there is a ‘chicken and egg’ dynamic at work here: the cognitive framework shapes the search for information as well as its interpretation. Like the drunkard who only searches for his lost wallet under a street lamp because that is where the light is, regulators can focus only on information that is easily visible, systematisable, and quantifiable, and not look elsewhere.¹⁴⁷ However, the dangers of using only limited sources of information in managing risks are well noted in the risk management literature,¹⁴⁸ and were manifested in the crisis. In Rumsfeld’s well-worn trichotomy, there are the known knowns, the known unknowns, and the unknown unknowns. In risk management, different strategies are appropriate for each state of knowledge.¹⁴⁹ Regulators can never know all, and therefore never can anticipate everything: that is the nature of risk management. But they can at least improve on their ability to limit the extent of the unknown unknowns and respond accordingly.

1.100  Work therefore needs to focus on two interlinked fronts. First, regulators need to gain information from a wider range of sources than at present and coordinate its collection and analysis, in order to help them overcome problems of scale—how those developing strategies at a systemic level can have sufficient granular knowledge of how different parts of the system are working ‘on the ground’. Secondly, they need to develop mechanisms for cognitive challenge.

1.101  Focusing first on information and problems of scale, work done to date on network management systems suggests that a fruitful coordination strategy for managing complex systems is to develop intermediators, links in the system between its component parts.¹⁵⁰ Creating points for the collection and dissemination of (p. 42) information about the performance of the different parts of the regime is a good place to start. To give two examples, there are two existing mechanisms discussed above that are currently not utilized in this way but that offer some potential: colleges of supervisors and the peer review process.

1.102  Colleges of supervisors have their own challenges, as discussed above, but they are in an extremely good position to gain a global view of a financial institution. That view is of course not perfect, as the information asymmetries that exist between regulator and regulated will always remain. But it is a better vantage point than any national regulator has, or any regional or international committee of regulators. However, at present there are four and potentially up to six different colleges operating (two at the international level for banking, and insurance, plus one (potentially) for securities; and two at the EU level, again for banking, insurance, and potentially, securities), and on a number of occasions with respect to the same financial institution. As noted above, there is as yet little or no coordination between them at either level or between the EU and international level.

1.103  Moreover, there is no clear way in which the reports or observations of the colleges are reviewed or otherwise incorporated into the activities of the international committees of standard-setters. Although the BCBS has stated that supervisory colleges can have a valuable role in supplementing the peer review activities of the BCBS, it is clear that this is seen as a ‘top-down’ process in which colleges promote coherent and consistent implementation of standards across jurisdictions.¹⁵¹ What is not developed is a role for colleges in transferring information ‘upwards’ to the BCBS or ‘outwards’ to the other IRCs about the situation in the jurisdictions of different college members.

1.104  In failing to provide adequate coordination the regulatory regime is missing a significant opportunity to gain valuable information as to what is happening within financial institutions themselves, and thus to address, at least in part, the problem of scale. Either the FSB or the Joint Forum has a potential role to play here on gathering and reviewing information from the international colleges with respect to cross-border financial institutions, particularly where the colleges are ‘siloed’, and providing a point for aggregating and testing that information and then disseminating it around the rest of the regime, and further integrating it into the standard-setting process.

1.105  With respect to the peer review process, the FSB’s newly introduced Framework does recognize the need for reviews to be analytical processes as well as information-gathering exercises. Arguably, what is needed is not just a policy by policy analysis, however, but a further level of cross-policy analysis, examining the interaction of different standards on the ground, bearing in mind the negative endogenous effects created by the interaction of IASB standards and capital rules, for example.

(p. 43) 1.106  Moreover, there is a case for linking these two potential nodes: colleges and peer review processes, and indeed, more radically, for using ‘360 degree’ review in the peer review process. At present a peer review, either via an FSAP or via the ‘Level 3’ committee and ESA processes, consists of a self-assessment and/or a review of national regulators by regulatory officials from other countries, supported by a secretariat from the World Bank/IMF or relevant ESA. This process is an important facet of the regime, as noted above. The peer review process, however, is also overlooking potentially valuable sources of information, notably the supervisory colleges. National regulators can gain a good insight into the performance of each other within the colleges of supervisors. Indeed, early reports on their operation indicated that much of the time was spent in regulators giving technical assistance to other regulators.¹⁵² It is always a sensitive matter to ask one country what it thinks of another’s performance, but not impossible. Integrating an assessment of a regulator’s fellow members in a supervisory college into the peer review process could provide a greater insight into the quality of a regulator than an assessment against the formal indicators which are predominantly used at present. Going further and asking financial institutions themselves for an assessment of their supervisors’ performance is probably a step too far for many national regulators, but it would bring the peer review assessment process closer in line with common management practices within both public and private sector organizations.

1.107  The second task that regulators need to focus on is to create mechanisms for cognitive challenge. Risks are things that may happen in the future. As we have seen, scenario analysis is a key aspect of risk management for financial institutions and is a necessary and inevitable part of regulating risk. Managing risks requires managers to imagine how they may arise and what their consequences may be. But the scenarios that they imagine can be too bounded and conservative. In short, regulators need to be more imaginative. They need to use an enhanced knowledge of the past and the present to build more imagined, and imaginative, futures. Furthermore, rather than searching for harmonization, the regulatory regime needs to build in structures for challenge and experimentation.¹⁵³ Regulatory reviews need to focus not just on impact assessments and sunshine reviews, but on what risks and activities it may be ignoring. This includes developing the equivalent of the ‘near miss’ analysis so important to risk management in a number of domains, including air traffic control, nuclear power, and medicine.¹⁵⁴ But it also requires a cognitive openness to different understandings of markets, risk, and behaviour, that requires (p. 44) gaining information and understandings from multiple sources. The claim is often brought that regulators need to be experts, to know how the markets operate. To this end, the Bank of England used to have ‘grey panthers’, senior bankers who advised officials on the markets. There is no doubt a place for grey panthers, but there is arguably a greater need for mavericks, of any hue. Greater diversity in the membership of key decision-making bodies and in the people that they talk to could help to promote a different way of thinking. There needs to be someone to say when the Emperor has no clothes (albeit that few might listen).

1.108  Regulation is a messy, complex, and often thankless task. In theoretical terms, the regime for financial regulation provides a clear illustration of a complex polycentric regime operating in a complex and dynamic environment where interactions and interdependencies are ill understood, in which power is fragmented and contested and regulatory capacities are highly variable. The putative ‘regime managers’ are attempting to steer the regime or different parts of it, but their role remains contested as existing actors seek to maintain their own policy autonomy. Further, all regulators are in a complex institutional position, in that the crisis has demonstrated to them and the markets that both sets of actors were operating on fundamentally flawed assumptions. The cognitive shocks caused by the crisis have prompted a series of critical self-observations. Whilst this learning process is essential, it has to confront two problems.

1.109  The first is that regulators are learning from a limited range of experiences. The crisis was an experience of major significance, but it took regulators into largely unchartered waters. There had been banking crises before, but none that affected so much of the global financial system at one time and on such a scale. There are well-recognized difficulties for organizations who attempt to learn from limited experiences: the lessons drawn may relate well to this crisis but may not be relevant for the next crisis that will take a different form.¹⁵⁵

1.110  Secondly, regulators have to engage in processes of critical self-reflection if they are to learn, but in so doing they risk jeopardizing their claim to expert authority. Leaders and experts are expected to know the answers. To be constantly questioning what one is doing can look to outsiders as if one is trapped in self-doubt.¹⁵⁶ Regulatory organizations are afforded a significant part of their authority and legitimacy on the basis of their impartiality and expertise. Admitting to mistakes (p. 45) can therefore be self-undermining.¹⁵⁷ This paradoxical position is difficult for an organization to avoid. Traditionally, regulators try to manage the problem through self-effacement and pleas that they are simply implementing decisions made elsewhere. This is a difficult strategy to adopt, however, when they operate in the public gaze and where transparency of their operations and discussions is demanded by their different audiences or legitimacy communities, including other regulators within the regime.¹⁵⁸

1.111  There is no simple answer to the multiple challenges that financial regulators face. The easy solution is to advocate organizational reform, such as a single, global financial regulator. This may have the advantage of making policymakers and regulators look busy, but it can often avoid problems rather than confront them. There is also a limit to what organizational structures can achieve. Coordination is a difficult challenge, but the solution advocated here is not the creation of a single ‘world financial regulator’ or a financial ‘WTO’ although these have been proposed by some.¹⁵⁹ It could be argued that this is the natural solution: that what is also needed is clearer leadership from the top, so that all regulators in the world are operating to a single set of rules and using a single set of supervisory processes. Such a development at the global level is politically hard to envisage at present, but after a few more crises it may be where we eventually end up. It is exactly this which is being attempted at the EU level, though in a very different political and legal context.

1.112  At either level, it is not clear that centralization is always desirable. There are still significant problems of scale to overcome: problems of legibility, oversight and multiple translations of rules as they course through the implementation web do not go away just because there is one regulator rather than several. Moreover, a global and EU financial regulator would still have to rely on national regulatory authorities to implement regulation, and many rules would still have to be at the level of general principles to be able to accommodate local conditions and cultures. Furthermore, the political difficulties of managing an organization whose members are themselves sovereign countries would still remain, with the inevitable issues relating to gaining agreement on rules and sanctioning recalcitrant national governments or their regulators. There would still be significant differences in regulatory capacities between different regulators and different countries meaning that any harmonization at the level of rules (even if they are detailed) is likely to be undone at the level of implementation. Moreover, it is by no means clear that such a system would be (p. 46) any more or less transparent, representative, or ‘democratic’ than the present one. It would be neater to depict on an organizational chart, but that should not be its principal virtue.

1.113  As for harmonization of rules and practices, it certainly makes the rule books look tidy, and cross-border firms like it, or at least say they do—in fact they can benefit from differences between regimes through regulatory arbitrage. However, assuming that harmonization is in fact achieved in practice, the dangers of harmonization, as evidenced in part by the crisis, are that it creates endogenous risk. Just as uniformity of risk models meant all market actors responded in the same way, magnifying the risks they were meant to be mitigating, standardization of regulatory requirements meant to mitigate risks, if flawed, can cause risks to spread far more quickly than a variety of standards.¹⁶⁰

1.114  From a purely functional perspective, the real difficulty faced by anyone trying to redesign the structures of global financial regulation, even assuming they could start with a blank canvas and with no interested parties scrabbling to hold the pen, is that managing global financial markets makes contradictory demands on organizational design. We can draw an analogy with Perrow’s analysis of organizational and technical systems. In his analysis of ‘normal accidents’, systems can be complex or linear, and tightly or loosely coupled. In systems in which there is a high degree of complexity (many parts, operators, procedures) and a high degree of interaction between them, the degree to which catastrophic failure could occur depends on how tightly they are coupled they are: what buffers exist in the form of technological buffers, or time to correct the error, or substitutability of personnel, procedures, systems, or resources. Different organizational structures are likely to be appropriate for different types of systems. Highly centralized organizations may be appropriate for systems which are linear and which have expected, visible interactions with tight coupling between them, such as dams or rail or marine transport. In contrast, decentralized structures are more suited to systems with complex interactions which are loosely coupled, such as research and development, multi-goal regulatory agencies, or mining. However, for highly complex but tightly coupled technical systems, such as a nuclear power or chemical plants, space missions, or aircraft construction or, we could add, financial markets, then there is a contradictory set of demands placed upon organizational structures: organizations with strong central controls are needed, but so is a structure which allows decentralization and decision-making at the local, operator level.¹⁶¹ Whatever the organizational design for their management, however, the bad news is that such systems, he argues, will inevitably experience failures.

(p. 47) 1.115  So the solution proposed here is modest and partial—less likely to grab the headlines but probably more likely to make a difference. It is to reiterate the need for greater coordination, but it is in addition to recognize that going forward, regulators need different types of information and different ways of interpreting it, they need to have a far closer understanding of the complex and dynamic network of financial markets that both regulators, governments, and market actors are continually reconstituting, they need to be open to challenge and they need to be imaginative. That requires two interrelated strategies: creating linkages between the component parts of the organizational structure to enable information to flow through whilst at the same time making it more resilient, and creating scope for greater imagination, experimentation, and challenge: a place for mavericks as well as panthers. Without these, regulators will indeed be trapped fighting the last war.