Of the panoply of measures proposed by the European Commission to establish a common regulatory framework for financial services after 1992, the draft Capital Adequacy Directive is the most controversial. It proposes an EC-wide minimum initial capital requirement for non-bank institutions as well as for banks, though the requirement for non-banks would be lower. Its aims are to ensure non-banks are adequately protected against risk and to facilitate `home-country' control over financial firms.

These proposals have elicited considerable disagreement among member states over the application of capital adequacy requirements to non-banks. This prompted CEPR to collaborate with the Commission of the European Communities in holding a Seminar in London on 5 February on `Capital Adequacy Requirements in the Financial Services Sector'. It was organized by Paul Grout and Colin Mayer, Co-Directors of the Centre's Applied Microeconomics programme, with funding from the Nuffield Foundation, as part of its support for CEPR's programme of research workshops.
The Seminar opened with a presentation by Pen Kent (Bank of England) on `Capital Requirements in Regulation'. He defined two purposes of regulation: to provide reasonable protection for users of financial services and to contain the danger that the failure of firms could propagate a more general crisis. In securities at least, Kent noted, there are strong incentives for firms to hold capital well in excess of regulatory requirements. Regulation is still required, however, since competitive pressures may erode self-imposed standards, and capital holdings that seem sensible after a period of financial disturbance may come to seem onerous as memories fade. Though on average securities firms hold three or more times more capital than rules might dictate, not all firms have proper backing. Moreover, while alternatives to capital requirements insurance requirements and compensation schemes for investors and clients would provide reasonable protection for users, their contribution to assurance against systemic risk would be minor or even negative.
The UK system of capital adequacy regulation distinguishes among the types of risk predominant in different financial services. Banking requirements focus on credit risk, The Securities Association's rules mostly address position and settlement risk, while the Investment Management Regulatory Organisation (IMRO) deals predominantly with operating risks and risks arising from the investment management process. These authorities also adopt differing definitions of capital. Kent argued that while these functional divisions have considerable merit, they are increasingly difficult to apply as firms undertake a variety of different kinds of business.
In designing any regime, regulators are faced with difficult dilemmas. There is an argument that regulation should err on the side of caution, but the internationalization of financial markets makes it essential that excessive requirements do not drive business elsewhere. The `level playing field' argument within the EC, that financial institutions should compete effectively on an equal basis, does not imply that they should all face the same regulations, according to Kent. It is not clear that a bank, which raises an important part of its funds as retail deposits, should be subject to the same requirements as another firm, which relies more on commercial paper or short-term debt. He therefore criticized the draft Capital Adequacy Directive for seeking to make the same rules on market risk applicable to the investment business of banks and non-banks alike, with potentially serious effects on EC non-banks' competit iveness worldwide. This had only been avoided in the case of the EC's banking regulations, Kent noted, because of the background of worldwide regulatory convergence under the auspices of the Bank for International Settlements.
The presentation by Colin Mayer (City University Business School and CEPR) on `Capital Adequacy for Investment Managers' derived from work with Julian Franks (London Business School). In the United Kingdom, IMRO imposes both a flat-rate requirement of £5,000 and either a 6 week or 13 week proportion of annual expenditure, depending on whether investment managers hold some client assets in their own positions. The Commission proposes an even sharper distinction based on holdings of client assets, with an absolute requirement of either 25,000 or 250,000 ecu. While the UK system demands little capital of small firms, even if they control clients' assets, Mayer noted, the EC proposals would impose substantial barriers to entry of small firms.
In his research with Julian Franks, Mayer had investigated whether capital requirements were needed for investment managers. They had found little evidence of systemic risk in investment management, though managers which took investment positions on their own account were more exposed to the solvency of other firms. A more important class of market failure for investment managers stemmed from imperfect information and the resulting inability of investors to judge the quality of firms. Of these risks, fraud, theft and irregular dealing seemed the most important. Purely financial risks were less troublesome and can be significantly diminished if investment managers' own positions are placed in separate legal entities. Large execution errors could occur but were not common, and tended to be unrelated to the financial performance of investment management firms.
The traditional reasons for imposing capital requirements on banks are therefore largely absent for investment managers, Mayer argued, for whom there was a very poor relation between capital and quality. Capital requirements were not well suited to addressing failures stemming from imperfect information, which were better dealt with by measures to improve screening and monitoring of investment managers: `fit and proper' tests and `conduct of business' rules, both of which are features of UK regulation and EC proposals. Mayer argued that capital adequacy requirements for investment managers should be withdrawn once more appropriate regulations were in place: separation of own positions into different entities governed by different rules; the use of private auditors, private and mutual insurance; and most importantly, requiring the use of custodians to operate client accounts.
Quite appropriately, the EC proposes modest capital requirements for firms employing custodians, but Mayer argued that, because of the poor relation between capital and quality for investment managers, it would be very difficult to define appropriate capital requirements for firms that do not use custodians. In contrast, UK regulations provide inadequate protection to investors in firms which do not employ custodians and may impose unnecessarily high requirements on those that do.
The Seminar also heard two other presentations. Paul Marsh (LBS) discussed `Capital Adequacy for Brokers and Dealers', explaining the evolution of requirements for positional risk into the present UK regulations. He also examined theoretical analyses of rules for market-makers and brokers in the light of evidence from market-makers' books. Geoffrey Fitchew (Commission of the European Communities) concluded the Seminar with a wide-ranging discussion of current regulatory structures in EC member countries and the policy problems of competitive deregulation and regulatory harmonization.