Institutional Structure of Financial Regulation

Theories and International Experiences

Robin Hui Huang, Dirk Schoenmaker, Robin Hui Huang, Dirk Schoenmaker

• Publisher: Routledge
• Year: 2014
• ISBN:
  9781317910633
• Edition: 1
• Topic: Business
• Subtopic: Finance
• Index: Business

In light of on-going global financial crises, the institutional structure of financial regulation is currently a subject of significant academic and practical interest. The financial crisis has called into question the adequacy of financial regulation at the national and supranational levels, and has instigated financial regulatory reforms in major markets overseas. This has included the enactment of the Dodd-Frank Act in the US, and the programme to split the Financial Services Authority in the UK.

This book examines the institutional structure reform of financial regulation from a comparative perspective, exploring both fundamental theories and international experiences. The book explores the three main institutional structures of financial regulation in the world; the sectors-based model, adopted in the US, Mainland China and Hong Kong; the twin-peaks model with Australia and the Netherlands as its pioneers; and the single-regulator model as represented by the former Financial Services Authority in the UK and the Financial Services Agency in Japan.

The book contains contributions from renowned experts in the field of financial regulation including Douglas Arner, Jeffrey Carmichael, Robin Hui Huang, Dirk Schoenmaker, and Michael Taylor, and will be of interest to students and researchers of banking and finance law, and comparative economics.

Part I

Fundamental theories

1 Regulatory reform after the financial crisis

Twin Peaks revisited

Michael W. Taylor

1.1 Introduction

Prior to the publication of my 1995 paper Twin Peaks: A Regulatory Structure for the New Century,¹ the institutional structure of regulation received little attention in either academic or policy circles. To the extent that issues of institutional structure were noticed at all, they were viewed as at best a second order issue that made little practical difference to the quality or effectiveness of supervision. However, the publication of Twin Peaks gave rise to a vigorous debate in the UK around issues of regulatory structure that was soon replicated in other countries. In the latter half of the 1990s this policy debate began to be mirrored in practical developments and a strong trend towards institutional reform gathered pace in many advanced and developing economies. Australia, the United Kingdom, the Netherlands, Germany, Ireland, Korea, Japan and several other countries introduced or announced changes to their structures of regulation in the years that followed.

Institutional developments in turn have encouraged a growing literature on the leading policy issues connected with these reforms. At the core of this literature has been the examination of several closely related questions: should the central bank be directly involved in banking supervision and regulation? Should it be responsible for the regulation of nonbank financial intermediaries? What are the pros and cons of creating a unified regulatory agency – responsible for banking, securities and insurance supervision – outside the central bank? Should prudential (safety and soundness) and consumer protection regulation be combined within the same agency or are they best carried out separately?

The last issue has been one of the central points of contention between proponents of a Twin Peaks structure – in which prudential and consumer protection regulation are assigned to separate agencies – and the adherents of a single unified regulator. Until the crisis, it had appeared that this issue had been decisively settled in favour of the single regulator model. The UK’s Financial Services Authority (FSA) inspired a wave of countries to adopt a similar institutional structure and gathered many plaudits in the years before the financial crisis, including from such influential bodies as the International Monetary Fund.² By contrast only relatively few countries – Australia and the Netherlands in particular – quixotically went down the Twin Peaks route.³

Since the Global Financial Crisis first broke in 2007 there has been a reevaluation of the single regulator model, combined with a switch in the focus of the debate. Post-crisis the most pressing issues have appeared to arise from the interaction between prudential regulation (especially of banks and other systemically important firms) and the central bank’s roles as lender of last resort and guardian of financial stability, rather than between prudential and consumer protection regulation. Moreover, the increased emphasis now being given to ‘macro-prudential’ regulation – essentially a set of tools that aim to prevent the build-up of excessive risks across the financial system as a whole – has raised fresh issues about how this new form of regulation should be managed, and what relationship it needs to have to the institution-specific ‘microprudential’ regulation of individual firms.

One consequence of the resulting reassessment has been to reawaken interest in the Twin Peaks model of institutional structure. A significant factor in this development has been the decision of the current UK government to dismantle the Financial Services Authority, which was the inspiration for the adoption of a single unified regulator in several countries, and to replace it with a Twin Peaks structure. Two new agencies, the Prudential Regulation Authority (PRA) and the Financial Conduct Authority (FCA) replaced the FSA in April 2013. The FCA has inherited many of the FSA’s former functions, with the exception of the prudential regulation of banks, insurance companies and large securities firms, which has become the responsibility of the PRA. While the FCA remains a standalone agency, the PRA is a ‘subsidiary’ of the Bank of England, thus recreating the close connection between the central bank and the prudential regulator of banks that existed pre-1997.

In the context of these developments it is timely to revisit the case for Twin Peaks and to consider what lessons the financial crisis holds for the institutional structure of regulation. Before entering into the case for the Twin Peaks I will first consider the issue of why institutional structure should matter at all: in other words, the extent to which the intense debate of the past decade and a half addressed genuine issues or was mere displacement activity that substituted for real reform. Having examined this issue, the chapter then sets out the main justifications for the Twin Peaks structure that were set out in my 1995 paper. Finally, I review four lessons for the institutional structure of regulation suggested by the financial crisis.

1.2 (Why) does institutional structure matter?

Until the mid-1990s the institutional structure of regulation was not considered to be a worthy subject of academic study. It appeared to fall between a number of different disciplines: not susceptible to treatment by the techniques of modern economics (which to its detriment generally neglected the impact of different institutional arrangements); too much a matter of policy to greatly interest lawyers; and too esoteric even for academics in the feld of public administration. From the point of view of many policy-makers and those charged with the day-to-day task of supervision and regulation such matters also seemed a distraction from the more urgent tasks at hand. What mattered, it was said, was making sure that regulation was done properly: the name on the brass plate on the building in which the regulators were located was of minor importance.

Institutional reform has also been tainted by the suspicion that it remains the last refuge of politicians who are keen to be seen to ‘do something’ in the wake of a financial crisis. Too often, changing institutional structures by creating new agencies, or by consolidating or abolishing existing ones, gives the appearance of learning the lessons of regulatory failure without necessarily changing actual regulatory practices. The focus on agency consolidation in the Obama administration’s initial proposals for regulatory reform led one leading economics commentator to dismiss them as an example of the ‘characteristic, but futile, response’ to a financial crisis which ‘is to move the regulatory deckchairs on the deck of the Titanic’.⁴ Critics of the UK government’s reorganization of regulation have similarly claimed that it is driven more by politics – squarely placing the blame for the financial crisis on the decisions of its predecessor by dismantling an agency it had created – than by an urgent need to repair a flawed model.⁵

Clearly, any particular institutional structure cannot by itself ensure high standards of supervision and regulation. In an IMF Working Paper, ‘Issues in the Unification of Financial Sector Supervision’ my co-author Richard Abrams and I attempted to identify the main elements of an effective regulatory regime.⁶ We proposed five essential factors:

1. Clarity of aims and objectives. Clear objectives assist the management of a regulatory agency in the allocation of resources and in determining the policy response to a particular problem; they can also provide a break on regulation expanding beyond the minimum necessary to correct market failures.

2. Independence and accountability. A regulatory agency must be able to make decisions that belong to its sphere of competence without undue outside interference, whether by ministers, parliamentarians or industry leaders. Among the factors that best protect agency independence are safeguards against the arbitrary removal of senior management, and budgetary autonomy in the sense of an earmarked source of funding for the agency. However, agency independence should also be balanced by accountability arrangements to ensure that the agency is answerable for its policies and actions to government and legislature, the source of its powers. Some form of formal consultative mechanism with the regulated industry should also be considered.

3. Adequate resources. The regulatory agency needs to have sufficient resources to recruit, train and retain a cadre of professional staff, and to ensure timely and effective data collection and processing.

4. Effective enforcement powers. A regulatory agency must possess powers to require information from regulated firms, to assess the competence and probity of senior management and owners of regulated firms, to take appropriate graduated measures in the event of failure to comply with regulatory rules, including having the ultimate power to intervene in a firm’s operations if necessary and to revoke licences. The effectiveness of enforcement powers depends on the ability to amend them quickly: for this reason only the broad outlines of the agency’s powers should be set out in legislation, leaving the detailed rules to be issued by the agency itself.

5. Comprehensiveness of regulation. A regulatory system needs to be free of gaps so that no particular types of activity or firm escape regulation simply because there is doubt over which agency is responsible for regulating them. Regulators must also have the flexibility to respond to market innovations.

These five factors were not intended to be an exhaustive list of the characteristics of an effective regulatory regime, but they were intended to put the discussion of regulatory structure into a broader context. Whether regulation should be conducted by a single agency or a multiplicity of agencies, and whether multiple agencies should be organized according to objectives (the Twin Peaks approach) or according to the traditional distinction between different industry segments (banking, insurance, securities) are essentially subsidiary issues to the more fundamental problems of ensuring clarity of objectives, independence and accountability, adequate resources, effective enforcement powers and the comprehensiveness of regulation.

Nonetheless, the institutional structure of regulatory agencies is not a mere distraction from more urgent issues as is sometimes suggested. If the regulatory structure is not a sufficient condition for ensuring effective regulation then it is surely a necessary one. This is for three main reasons.

First of all, the type of regulator...