Recommendation 1: Insolvency practitioners and governments should ensure that measures taken to improve cooperation, recognition and access will also provide for speed and certainty in international financial insolvencies.

The preliminary steps in an international insolvency procedure include winning recognition for the procedure in foreign jurisdictions where the firm was active, and access for the administrator to the foreign courts concerned. Direct cooperation between the different courts can help.

Insolvency practitioners are already at work developing methods to achieve greater cooperation and more assured recognition and access. Their efforts should take account of the special circumstances and features of international financial insolvencies. Among the measures that could encourage speed and certainty are the following:

• Local financial regulators helping to ensure that foreign insolvency practitioners are supported by a "friend of the court" brief, where possible and appropriate;

• Encouraging prompt action in the courts, appointing an administrator and initiating contact with courts in other jurisdictions;

• Seeking some standardization of the protocols governing the way in which courts and administrators cooperate, a device that is being used in a growing number of international insolvencies;6

• Assigning all international financial insolvencies to a single court, or in some other way establishing a recognized "port of call" and a pool of expertise in the court system; and

• Establishing authoritative legal opinion on what the effects of recognition are likely to be in particular countries for financial insolvencies, even where recognition is left to the discretion of the courts.

There is one other issue that matters: the ring-fencing of assets -- that is, keeping local assets available for settling local-creditor claims first. This principle is increasingly unsuited to modern conditions. As institutions move toward managing their risks on a global basis, it is likely that managing an insolvency on a global rather than a local basis will reduce the chances of a concentration of severe loss that could threaten other institutions by distributing the eventual losses widely. This is not to suggest, however, that local law on separate legal entities and legally enforceable relationships should be disregarded.

Ring-fencing is motivated by a concern that domestic interests will not be adequately protected if global firms are in trouble. Supporters fear a lack of effective supervision by foreign supervisors, and doubt that there will be an equitable settlement of claims in foreign insolvency proceedings. Eliminating ring-fencing will require national authorities to be more confident that the interests of their nationals will be adequately protected by official actions abroad. Thus, a unified approach to international insolvency process may be possible only if it is combined with agreement on accepted standards of home-country consolidated supervision.

Recommendation 2: Legislators should enact laws to provide for speed and certainty on the issues of judicial cooperation, access and recognition in international financial insolvencies, preferably supporting the norms of universality.

One element of the work by INSOL and UNCITRAL explores the possibility of legislation to ensure judicial cooperation, access and recognition. Some countries already have such laws. Others are considering legislation, and more may do so in the future. The advantage of legislation is that it reduces uncertainty about the initial steps in an international insolvency. And it need not remove all national discretion on the conduct of an insolvency: generally speaking, legislation that provides for cooperation in cross-border insolvencies reserves the right to refuse recognition when that would conflict with legitimate public policy.

In the case of financial insolvencies, the first policy consideration should be systemic risk. This tends to strengthen the case in favor of international cooperation rather than against it. Laws that acknowledge this fact would be beneficial; and the norms of universality, which call for a global approach to insolvency, should wherever possible be incorporated into new law for financial insolvencies. Otherwise, the issues for legislation are much the same as those for informal cooperation discussed under Recommendation 1. Even where the law does not produce certainty and speed for international commercial insolvencies in general, it should do so for financial insolvencies.

Recommendation 3: Financial regulators should pursue international contact among regulators of different types and with insolvency practitioners.

Regulators from different countries and sectors meet in several forums: the G-10 Central Bank Governors at the Bank for International Settlements (BIS); the Basle Committee of G-10 bank supervisors at the BIS; the International Organization of Securities Commissioners (IOSCO); and, newest among the groups, the International Association of Insurance Supervisors (IAIS). The Basle Committee and IOSCO have pursued joint discussions on a number of occasions, and the Basle Committee, IOSCO and IAIS have established the Joint Forum to pursue more effective supervision of international financial conglomerates. The various groups can exert considerable influence over national regulatory policy and international private practice, although they have no formal authority. (The counterpart European Union regulatory groups have a more direct role in shaping EU law.)

Regulators may know the leading insolvency practitioners in their own country today. Indeed national law sometimes stipulates that the administrators must be regulators themselves: in the United States, for example, federally insured banks are administered by the Federal Deposit Insurance Corporation when they fail, and stock brokerage firms by a trustee appointed by the Securities Investor Protection Corporation. Also, regulators are increasingly familiar with their direct counterparts in other countries, largely because of the work of the international financial regulatory groups. The fact that the Joint Forum was set up is testimony to the importance which regulators are beginning to attach to contacts across industry lines.

However, this falls far short of the international web of familiarity and trust that is needed. The Barings insolvency was marred by insufficient trust and cooperation among the various regulators and insolvency practitioners involved. Regulators from different industries too often have sharply different legal responsibilities and quite distinct ideas about systemic risk. Bridging these gaps will require more contact across borders and across sectors.

This recommendation reflects the experience of INSOL and UNCITRAL. Providing the opportunity for informal international contacts has been a highly valued achievement of their work to date. Until recently, many insolvency judges had little or no contact with their counterparts in other countries and little knowledge of insolvency practices abroad. The INSOL and UNCITRAL meetings and conferences have helped judges, as well as lawyers and accountants and other practitioners in different countries, to start tackling their common problems.

Recommendation 4: National regulators should work through the international regulatory groups to develop a protocol on the organization, timing and type of assistance they should extend to foreign administrators.

During an international financial insolvency, the regulators who have supervised the failed firm can provide vital help to the administrators concerned. A foreign administrator in a secondary proceeding may need basic information (which a national regulator should have) about the operations of the failed firm such as where its local establishments are, what its legal structure is, the extent of its operations, its governance arrangements, the applicable law and court and, perhaps, where qualified local counsel and other professionals might be found. In establishing their authority, the administrators might often be helped by the friendly intercession of local regulators with the local courts.

The best approach to providing this assistance will vary from country to country, but certain general principles should apply. For example, in the same way that this report calls upon countries to set up a "single port of call" in the legal system for international financial insolvencies, so it should be possible for regulatory agencies within a country to agree among themselves which of them should be the contact point for foreign administrators in the event of the insolvency of a financial institution. This facility must be operational twenty-four hours a day. Perhaps the desirable characteristics for information -- scope, timing and conditionality -- could be agreed. The nature of assistance for foreign administrators, not only with the courts but also on the practical problems of exercising local control, might be worked out. Since adoption of these principles across countries is desirable, the international financial regulatory groups could contribute greatly to their design.

Recommendation 5: National regulators should take the insolvency regime in the home country into account when licensing the operations of foreign financial firms.

Regulators have to consider many applications for licenses from foreign financial firms wanting to open a subsidiary or a branch in their countries. To suggest that they should take into account the insolvency regime in the home country may seem contentious. However, it has recently become common for regulators to consider the quality of home country regulation and to grant a license only when they are satisfied that it is adequate. Moreover, they can and do give consideration to other home country policy issues, such as the prevailing standards on accounting and disclosure. This recommendation would extend that screening to include the home country insolvency regime as well. The factors that regulators might consider are, broadly speaking, the issues raised in this report, such as access to local courts and the ease of achieving recognition, and the rate of progress which is being made in addressing these issues.

Notwithstanding the importance of this issue, however, fair and adequate consideration of insolvency issues will take time: time for minimum standards to be developed and agreed, and for them to be reflected in national supervisory practice and adopted into national law. This recommendation does not provide a new pretext for excluding foreign competitors.

Information

Recommendation 6: International financial firms should aim to have up-to-date information on all significant exposures by transaction, agreement, legal entity and jurisdiction, held centrally and in a readily available, standard form.

When a financial firm becomes insolvent, its assets and liabilities have to be valued in order to establish the scale of competing claims. To do this, the administrator has to examine positions and exposures in some detail. Information may be scattered across different parts of the organization, documentation may vary (even for one type of transaction, even with a single counterparty), transactions records may not be easily related to legal records, and the records themselves may be out of date. The administrator must be prepared to examine the firm's position transaction by transaction -- which can take a long time.

If a failed firm or its parts are to be put up for sale, portfolios have to be valued. This is relatively straightforward for instruments and contracts from a liquid market, such as exchange-traded instruments. But valuing over-the-counter portfolios is much harder because positions have to be run through the data management systems and models of other firms.

That was what happened when Drexel Burnham Lambert failed in 1990. Although its over-the-counter portfolio was uncomplicated by today’s standards, it took a long time to assemble all the information, establish how it was organized and then run the tapes and produce a value in which an interested firm could place some trust. The process would be far more protracted and difficult today for any large internationally active financial firm.

Some over-the-counter transactions (such as long-dated interest-rate swaps) last a long time, and documentation practice may well change materially before they expire. Records must therefore distinguish between documentation of different vintages, whenever these differences have significant implications for value.

Ideally, an internationally active firm should keep exposure data centrally and in a standard up-to-date form. In the event of insolvency, this would help the administrator and other firms value its businesses quickly. Although it is not in the interest of any particular firm to do this itself, each would stand to benefit if all did it.

The goal of this recommendation is information standards for systemically significant financial institutions. Achieving this goal will require collective action by the firms to establish standards, after first studying the level and kind of exposures to be covered, the form and content of data collected, and the feasibility and cost of doing so. National regulators could encourage this process, and might ultimately play an active role in ensuring that the standards are properly applied.

Size of Potential Losses

Recommendation 7: As part of sound credit-risk management, international financial firms should monitor, measure and manage all sizable exposures by legal entity, form of documentation and law of jurisdiction.

Any internationally active firm is likely to be made up of several affiliated legal entities. When such a firm fails, the parent may or may not be insolvent and the insolvency may well affect only some of the subsidiaries. Then the potential loss of any counterparty depends not just on the size of its exposures, but also on the legal entities with which it has positions; the form and enforceability of the documentation underlying the exposure; the jurisdiction in which any disputes about losses have to be resolved; and the internal record-keeping to identify who the counterparties are.

Looking carefully at legal issues is, of course, only part of what is needed to prevent the spread of insolvency via counterparty exposures. For example, the accessibility and certainty of credit enhancements, such as collateral or guarantees, have to be understood and properly evaluated. Credit concentrations, not just to single firms or groups, but also to firms with a particular geographic, industry or business cycle exposure, have to be properly controlled. Sound credit risk management is the first defense against the spread of trouble. For internationally active firms, this requires systems and controls that monitor credit exposure in sufficient detail. As in the previous recommendation, this one seeks the establishment of new industry standards, developed by systemically significant firms and encouraged by regulators.

Recommendation 8: National regulators, sometimes working through international regulatory groups, should encourage the speedy adoption of measures to reduce the risks in clearance and settlement.

When a big firm fails, it is almost certain to leave many transactions still in train with other financial institutions. Depending on how the clearance and settlement process works and the nature of the transactions, counterparties can suffer significant losses. For example, in many stock markets, there is still a delay between delivery of a stock in one direction and payment for it in the other. Delay is also common in foreign exchange markets, and the amounts involved are much larger, so the solvency of the counterparties of an insolvent firm might become an issue. The responsibility to contain this risk is divided between the firms themselves (for whom it is a type of credit risk) and those who manage the clearance, settlement and payment systems.

For at least two decades, regulators, markets and several private sector groups have worked together to strengthen these systems. Despite their achievements, it is hard to be sure that the chances of a systemically significant disruption have fallen. For, as the volume of transactions grows and the financial system becomes more international, the importance of quick and reliable systems for settlement and payment increases.

One recent initiative comes from the G10 Governors Committee on Payments. It has been developing recommendations for deepening and strengthening the international payments system, specifically to reduce the "Herstatt risk" of an incomplete transaction in the foreign exchange market. A group of private foreign-exchange-market participants, the G20, has been working in parallel to tackle this problem through the establishment of a new foreign-exchange clearing bank. Other recent work has covered clearance and settlement systems in securities and options markets, bilaterally and multilaterally. These initiatives have been useful. The recommended improvements should be implemented.

Recommendation 9: In jurisdictions where there is still doubt about the enforceability of close-out netting, legislators should clarify the position as quickly as possible. In over-the-counter markets, it is good risk management practice for firms to group individual transactions between them within bilateral master agreements. Each master agreement is intended to be a legally enforceable agreement that requires payment of the net value of all included transactions in the event of an insolvency. The importance of these agreements in controlling the effects of an insolvency is that, while the gross amounts can be very large indeed, the net amounts may be quite small.

Unless enforceability is clear, there is a danger that an administrator will ignore the net position, preferring to treat an insolvent firm's liabilities as common credits (where the proceeds of liquidation would be apportioned in due course) and its assets as credits on which the full amount should be paid. In that case, the solvent party can be made to pay up, even if the insolvent side does not.

Much work has been done by different industry associations to prevent this "cherry picking" and some jurisdictions now prohibit it by law. But in others, the enforceability of close-out netting agreements is supported only by legal opinion; and in a few places, there is no supporting opinion, precedent or law, so doubt about enforceability is all the greater.

A limitation on the extent of netting is that the separate capital of each legal entity within a large financial group will be subject only to the claims of those creditors with whom that entity has legally binding agreements, not to all creditors of its parent and affiliates. While broader netting might be helpful in reducing systemic risk, this recommendation is only intended to encourage netting between parties in legally binding relationships. It does not advocate disregarding the separate legal status of distinct entities and their assets.

Consistent with this legal framework, in countries where there is doubt about enforceability, legislators should change the law.

Recommendation 10: In countries that draw no clear distinction between house and client assets, the law should be amended.

Some jurisdictions have a statute (or a legal tradition) that recognizes the distinction between house and client assets inside a financial institution. Others do not. So, for example, a US brokerage might execute a transaction on behalf of a client through a subsidiary abroad. If that brokerage then failed, US bankruptcy courts, or the Securities Investor Protection Corporation as trustee, might consider client funds or securities to be exempt from creditor claims. But under foreign law, creditors of the foreign subsidiary might have a claim against those same assets.

As a result, in insolvency proceedings there can be uncertainty about the potential loss to some clients. It is often difficult enough for those in charge of an insolvency to distinguish between own funds, co-mingled funds and client funds. The difficulty is compounded if an insolvency involves a jurisdiction without a tradition of making any distinction.

This has usually been viewed as a consumer protection issue. For example, consumer concerns were the main motive behind a similar recommendation by the Futures Industry Association in 1995 (See appendix II). But the issue is broader and more complex than that. The client could be another financial firm and the amounts involved so large that the law's failure to discriminate between client and own funds could put the counterparty itself in danger.

Recommendation 11: Laws are needed to make it possible for administrators of financial firms to continue operating in the interests of managing market risk.

In a number of countries, the law does not grant an administrator much (if any) authority to continue trading the portfolio of an insolvent institution. The suspension of authority to operate may be particular to certain instruments, discretionary and temporary; or it may be inclusive, automatic and lasting throughout the insolvency; or some mixture of these constraints.

In the insolvency of an internationally active financial firm, the administrator often needs to continue trading. The value of the firm has to be hedged against changes in financial markets to avoid potentially large further losses. Typically, the market risks of its portfolios will be managed continuously before the insolvency occurs, and this process should not be interrupted.

Even if the law does permit trade to continue in principle, that may not ensure that it does so in practice. Potential counterparties will shun an insolvent firm unless they are confident that they will be paid in full in any new transaction. They need assurance that such transactions will take precedence ahead of previous creditors. Finally, there may even be a case for limiting the power of exchanges to stop an insolvent member from trading.

Recommendation 12: Just as widespread use of master agreements for over-the-counter transactions has standardized contract terms and enforceability of netting, netting should be extended across agreements, ultimately through development of a single master agreement covering many different products.

Use of standard documents does much to ensure the clear understanding of contractual obligations in insolvency. At the moment, there are different master agreements in use for different types of over-the-counter business. These have developed in line with each activity, typically to suit its particularities. Traders like this diversity, because it is convenient and familiar.

However, in an insolvency, diversity can exacerbate the losses of counterparties. To take an example, suppose that a counterparty is a net debtor to the insolvent firm under one agreement and a net creditor under a second agreement. That counterparty is liable for the full settlement of its position with the administrator under the first agreement but, at the same time, it should only expect some fraction of its claims to be made good under the second agreement. Just as it is desirable to minimize the potential exposure of counterparties through formal netting arrangements within over-the-counter transaction types, so there is a case for extending netting across them.

The same limitation on the extent of netting discussed in Recommendation 9 applies in this case as well. While worldwide, cross-affiliate master agreements may be the ultimate objective, for the moment cross-agreement netting should be encouraged between parties in legally binding relationships, taking account of the separate legal status of distinct entities and their assets.

The most straightforward route to cross-netting would be through the adoption of a single multi-product master agreement. This would require the various associations either to collaborate in developing a new agreement or to endorse an existing instrument, such as ISDA's agreement which is already widely used.

Facilitating Sales

Recommendation 13: The international regulatory groups should help establish standards on the hand-over of proprietary pricing models by insolvent firms.

Any financial firm that is interested in buying the over-the-counter portfolios of an insolvent firm will almost certainly be unable to value every transaction and position. In many internationally active firms, parts of their over-the-counter portfolios are priced using unique proprietary models. Potential purchasers may have to depend on those models to establish a sensible value.

This is an area that offers more questions than answers. Could the proprietary value of models be protected if the authorities used them on behalf of bidders? What if a firm bought the portfolio, after relying on official assistance in valuation, and then concluded that it had paid too much? Alternatively, should these models be made available to bidders? If so, when and how? Should there be standards for documenting models so that bidders could quickly assess them?

Even if firm answers to these questions are impossible, guidelines would be useful.

Recommendation 14: Regulators should consider sponsoring a forum to explore ways to speed up the process of assigning a portfolio of over-the-counter instruments in case of insolvency.

When a firm has agreed to buy a portfolio of over-the-counter instruments from an administrator, assignment of each component contract has to be negotiated with the third party firm involved. The latter may be reluctant to have one of its contracts assigned if (for example) it was out-of-the-money and might be used to offset some other obligation to the insolvent firm under a netting agreement. Negotiating the assignment of a portfolio can therefore take a long time.

In the future it might be possible to couch the documentation of over-the-counter transactions so as to facilitate subsequent assignment. However, if assignment is made automatic, a third-party firm almost inevitably loses control over its credit risk -- whether or not the instrument in question is in or out of the money at the time. So it is unlikely that all over-the-counter activities could take place on fully-assignable terms. On the other hand, it is difficult to imagine how a third-party firm could be given an effective assurance against loss of value or increase in credit risk. That might require some new institutional arrangement for compensation and risk adjustment, which in turn might need some form of official sponsorship. The range of theoretical alternatives is wide, and choosing between them will be difficult. A forum, bringing together regulators, practitioners and lawyers, would be a good way to start tackling the issues.

Footnote:

6. In the non-financial sector, the International Bar Association has promulgated a useful "Cross-Border Insolvency Concordat" for just such a purpose.