It is a particular pleasure to be presenting this report on Capital Adequacy Standards for Securities Firms to the annual conference of lOSC0 here in Santiago de Chile. I can trace my own association with this organisation right back to the first meeting of its predecessor organisation, the loser American Conference, which I think I am right in saying took place In Caracas In 1974. had the privilege of participating In a number of the Inter American conferences, in South and North America, but have never previously had the opportunity to come to Chile. It is a great pleasure to do so now, and I would like to add my voice to those which have already congratulated the organisers on the success of this meeting.

Working Party n° 3 Is one of those originally set up when the Technical Committee first met, in London in July 1987. The work we are doing i5 the first of its kind, in that no other organisation, at least to my knowledge, has set out to analyse the rules set by securities regulators for the maintenance of capital, and to suggest international standards on which such rules should be based. I do not thins 1 need stress the importance of capital standards In today's world. The onward march of international securities business, and the presence of the financial multinationals in the markets of so many parts of the world leads to exposures which firms find increasingly hard even to measure, still less control. And the markets are in a state of volatility which at least for the time being looks like staying with us. Under these conditions, firms dealing in both national and international markets need to be reassured that those with whom they will find themselves in counterparty relationships have sufficient capital for the business they are undertaking.

Before going on to introduce the four papers which our Working Party has submitted to this conference, I will mention two other important factors which we have to take into account. The first is the extent to which the traditional barriers between different kinds of financial institution are breaking down - indeed, in many countries, have broken down. delegation continues to make the differences between banks, insurance companies, mortgage loan companies and securities houses increasingly difficult to define. Many countries have abandoned the separations still maintained in the U.S. through the Glass-Steagall Act and the Bank Holding Company Act. These trends do not make life easier for regulators and supervisors. The problems of supervising financial conglomerates is having to be carefully thought about. There i6 a very particular problem of finding capital standards which can be applied by banking and securities regulators to entities which ray fall under the tutelage of either kind of regulator but which compete directly or the same business. I will cone back to the subject of banks and securities firms little later.

The second factor which influences the work we are doing in Working Party n 3 is the introduction of rules on capital adequacy in the European Community. This measure, referred to as the Capital Adequacy Directive, or CAD, is being introduced as part of he drive towards the completion of the internal market in the E.C. in 1992. Since five of the twelve member countries of the Technical Committee are also members of he E.C. (and Sweden and Switzerland are to some extent linking themselves to the EC legislation), we obviously have to keep a close eye on the shape the CAD Is taking. Indeed, one of the earliest objectives we set ourselves vas to seek to avoid a serious divergence between the rules being set in the E.C., and those being set elsewhere amongst the developed capital markets of the world.

Last time many of us net vas at this same meeting last year in another beautiful city La Serenissima - in other words, Venice. At that meeting I presented to the Plenary session the first report from Working Party n° 3. It vas a relatively brief document, but we had laboured long over it. Our report contained an analysis at the conceptual level of the approach of the securities regulator towards setting capital requirements. It emphasised the need for the regulator to assess as closely as possible the risks to which the securities firm was subject, and to tailor the requirements to those risks. In our report we described the nature of the risk which a firm runs, and suggested that the capital a firm should have available should have a number of components. When I made my address to the Plenary session in Venice said that if our first report was accepted, it would be our intention to develop our analysis of those capital components at a more detailed level. That we have been doing, and our analyses are the subject of the four papers which have been circulated you. I want to give you the briefest summary of those papers now, and then to scribe the other work on which we have been engaged.

We decided that we should examine in some depth the requirements in the four countries which were the original members of our Working party, party France, Japan, e U.K. and the U.S. We wanted to examine the scope for harmonisation of the various elements of the requirements. We took as our subjects the equity position risk requirements, debt position risk requirements, the base requirement and the minimum requirement for capital, and finally the definition of capital. I will take then in turn.

The risk which the equity position risk requirements is designed to cover is t he possible adverse change in the price of equity securities in which the firm holds a principal trading position. Such a risk arises as soon as a transaction has been made. The requirements are commonly referred to as "haircuts'', because their effect to discount (or shave an amount from) the latest market value of equity positions held, to cover the possibility of a future adverse price movement. And nor e my reference to the latest market value. For it is a central feature of our recommendation that all securities held are "marked to the market" every day. The concern is the potential price volatility of the positions held, since the assets are assumed to be held for trading purposes.

The rules we describe In the paper relate to fires which take positions, and hence have exposures, in the equity market. They do not therefore apply to florins which restrict themselves (or are required to do so) to agency brooking.

Our paper is based upon a close analysis of the rules as they apply in the four countries I have mentioned, which make up the original membership of our forking Party. In each of these countries a carefully structured approach has been taken, so that the capital requirements do match the extent of the risks. Thus, in addition to the mark to market principle, which I have mentioned, and the use of the haircut technique, our report shows how the four countries distinguish between the differing degrees of liquidity, which bears upon the prospective volatility, of different classes of equity. Me show how the regulators penalize positions which involve an unusual degree of concentration in the securities of a particular issuer; and where a single position constitutes an unduly large element In the fire's trading portfolio.

An important part of our analysis bears upon the way regulators accomodate hedging techniques. Hedging may take three forms. There i6 for a start the management of the firms own boot. There are quite considerable disparities in the way regulators make allowance for off-setting long and short book positions. Then there Is the use of derivative markets to hedge individual positions. Finally, there are the very large arbitrage positions taken by firms, where a basket of stocks is assembled which is designed to replicate the performance of the index, and which is offset by a position in the index future or option. In these maneuvers the 'gross positions nay be very large, but the risks are low, and so are the potential returns.

Towards the end of our report on equity position risk we put down the broad objectives which guide the regulators In the four countries under review. They are worth reciting here. They are that the requirements

1. should reduce the risk of firms failing, owing suns to clients or counterparties
2. should generally reflect the risks, so that fires are encouraged to use risk reduction techniques (because they will be treated less severely)
3. should not affect market practice perversely
4. should not be too costly to comply with nor too complex.

Our second paper is a comparison of position risk requirements for debt securities. his one, it must be said, is more difficult to read, at least for anyone who does not find it rewarding to immerse himself from the beginning in tables of statistics. But the principles are basically the same as those which guide the same four countries (for once again we have restricted our comparison to France, Japan, the .K. and the U.S.) in their approach to equity position risk. Thus, positions are recognised in principle on a trade date basis (although the U.S. starts from settlement date); positions are market to market and the aggregated amount of calculated risk is deducted from the firm's capital.

In three of the four countries surveyed, risk weights are determined according to the type of issuer and to the maturity of the security. In general, the assumption used which is of course borne out by experience) is that the longer the maturity, the higher the volatility, and hence, the risk weight is set accordingly higher. Risk weights also take account of the relative liquidity of the bonds.

As might be expected, the rules governing position risk for bonds are sophisticated and complex. Account is taken - and here again, one is talking of trading, not agency, fires - of off-sets between, for example, long and short positions in the same sub-category of maturity. All four countries have rules which recognise the ability of the firm to reduce its exposure (to risk) by the use of derivative markets. And In countries where firms undertake interest rate swaps, the rules treat the swaps for the purposes of market risk as notional government bonds.

We take considerable heart from our analysis of the capital rules applying to bonds. We think there is reason to suppose that harmonisation of risk weights, and of the treatment of offsets and hedging, would be possible to achieve. We believe that more detailed work again should be done towards that end.

We found ourselves on rather more difficult ground when it comes to studying the base requirement and the minimum requirement for capital. Our report on this component of capital discloses that there are differing philosophical as well as technical approaches amongst the securities regulators themselves, and that when one starts to compare the regime applied by the banking supervisor with that of the securities ,regulator, the differences are even greater.

The essential point about base or minimum requirements is that each nay be defined as being not directly linked to either market risk or counterparty risk. In this respect, the capital requirement does not follow the precept of our original concept paper, that the capital requirements should be related to the assessment of risk, although some countries require a cushion of capital which reflects the size of a firm (on the basis of its expenditure or its volume of business).

The approach towards minimum capital is thoroughly pragmatic. Any firm, in whatever kind of economic activity, needs starting capital, and only owners' capital can be ultimately counted as capital. Even where the requirements are expressed as a base, i.e. an amount determinable other than by reference to a fixed sum, there is frequently a minimum below which a firms capital is not permitted to fall (without, that is, the intervention of the authorities). Depending on local conditions, or on the type of business to be undertaken, the minimum level nay be set relatively high, so that the very threshold keeps out financially inadequate fires (as the case, for example, of members of some futures markets); or it may be set at a relatively low level, in order to allow the easy entry of new firms to the market, but in which case the requirements are made to rise as the level of business (however measured) increases.

This last approach is to apply a base requirement for capital. Such a rule takes as its starting point that there are certain business risks which are not susceptible to measurement in the same way that, for example, price volatilities in equities and bonds nay be measured and provided against. These risks would include the possibility of a general reduction in the volume of business, which would put strain on a firm's capital position if were not required to maintain a cushion of capital, which would still be held even if all the specific trading and other risk had been removed.

Our paper shows some considerable diversity in the way the base requirement is determined. Some countries relate it to the level of the firm's expenditures over a given period; some to the value of assets under management; others to the number of professionally qualified staff; and yet others to a volume of business requirement. Two leading countries, for example, relate the base requirement to a percentage of client's open positions, or client-related receivables.

On the basis of the report we have made on this subject, the search for harmonisation of regulation In respect of either minimum or base requirements for capital looks beset with problems. These I have described In terns of the very different approaches among securities regulators. And if alignment with the rules set by banking supervisors is sought, the problems are even greater. There is no practical possibility of applying a base requirement to the entire activities of a bank; and although there is a move in a number of countries to allow banks to "strip out" their securities trading activities from their conventional banking business (in order to apply lower capital weights than the general credit risk weight of 8% set by the Committee on Banking Supervision in Basle), there is resistance to accepting a base requirement as applicable to the trading book alone.

That the base requirement is a potential source of competitive distortion between banks and non-banks is undeniable. We are not at that point, so long as the banks are required to apply the 8% rule, and the securities firms are required to have both risk-based capital and a base capital. But if the risk-weights for banks and non-banks are aligned, the base requirement could constitute a severe competitive disadvantage for non-banks.

These rather pessimistic remarks bring me to our fourth paper, which is on the Definition of Capital. You will find set out here a very lucid exposition of the need for a proper definition of what is to be allowed as capital, and a description of the different forms in which capital may be provided. If it were simply a question of finding common ground amongst securities regulators, one would be entitled to feel confident on the basis of our analysis of reaching substantial harmonisation. But here again, considerations of competitive equality between banks and non-banks arise. There are two specific areas which give rise to contention between the two types of regulator. They are the use of subordinated debt as capital, and the use of guarantees. The position is that in France, Japan (for the branches of foreign firms), the U.K. and the U.S.,- securities regulators allow the use of subordinated loans to make up part of the capital of the firm. Generally, the amount of subordinated debt permitted is greater and the term required is shorter than the banking supervisors permit. But there are good reasons for the practice of the securities supervisors. By the use of subordinated loans, securities firms are enabled to meet their fluctuating, risk-based requirements, and if a firm were to encounter difficulties, the funds representing the subordinated loans could be used to meet the firm's obligations to customers and counterparties. Subject to certain conditions, as suggested in our paper, securities regulators believe that subordinated loans can provide as much comfort as is required, both to the lender and to the regulator.

Bank regulators do not allow banks to include guarantees as capital at all. Guarantees are accepted by some regulators, but they do not represent actual funds, and there is always the possibility of the failure of a guarantor. Guarantees therefore present difficult problems of alignment.

Before turning to other matters I must draw your attention to a small but important editorial change to the paper on the definition of capital. This consists in the deletion of the words "net worth" at the end of the first paragraph on page 12 of the paper. The reason for this deletion is to avoid any inference being drawn that the mark to market principle is applicable to the liabilities side of the balance sheet. I hope you will allow me to make that amendment to the text.

I have mentioned several times in the course of these remarks the problems of aligning the capital requirements applicable to non-bank securities with those which are applied to banks. The differences were very briefly touched upon in an annex to our original concept paper, and it was always our intention to develop the subject further. We have in fact been doing so, fairly intensively, even since the meeting in Venice last year. That work has been going along in parallel with the preparation of the four reports which I am presenting to you today.

We began by enlarging the size of our Working Party, to include representatives of Germany, Holland, Sweden and Switzerland, all countries in which banks play a significant role in the securities market. Over the time that we have been working on the subject, pressure from firms has continued to increase to find a basis for convergence between the rules applying to banks and those applying to securities firms. We have prepared a substantial analysis of the techniques and objectives of the two types of regulator. Our report on that subject has not yet been submitted to the Technical Committee, and hence is not really within my scope today, but I thought I should let you know of its existence.

I mentioned pressure from firms. The pressure manifests itself also on the banking regulators, who have always accepted that the 8% credit risk weight applied through the Basle Accord, reached in 1988, was not a refined way of looking at market risk. In fact the committee of Banking Supervisors at Basle has had for some time sub-committees looking at three types of market risk: interest-rate risk (which includes position risk for debt securities); equity position risk; and foreign exchange risk. At a meeting held in Basle in September between banking supervisors and securities regulators, consideration was given to the possibility of a common system for measuring risk, and a minimum set of capital adequacy requirements which could be applied to the securities business of bank and non-bank securities companies. There was enough general consensus to warrant further work being done on risk measurement of debt and equity positions. e are left, even if agreement on those is found, with the difficult matters have described, in the application of the base requirement, and the definition if capital. Those problems will need further careful study, before we can expect :o reach agreement.

June 1990