3
I. Introduction
In the recent past, several large, internationally active firms have entered formal insolvency
proceedings, among them Drexel, TWA, Barings, Maxwell, Olympia & York, Enron and Global
Crossing. Others have come close to doing so, for example Long-Term Capital Management. The
asset size of these actual or potential failures is increasing, and with it the risk of a disorderly and
contentious outcome.
Since Herstatt, no major financial insolvency has been disorderly, although there have been some
close calls. This is due to several contributing factors: a combination of improved risk
management, greater transparency, a vastly superior legal framework, international cooperation
among authorities, and good luck. This positive history, however, offers few reasons for
complacency. The failures to date have been of firms of moderate and intermediate complexity.
Firms and markets become ever more complex, more interrelated, and ever larger. An
increasingly deregulated market environment is a less cosseted one: greater competition entailing
greater risk of failure. Sovereignty remains a fact of life, but business operations are increasingly
cross-border, placing greater strain on the legal system.
Therefore, the underpinnings of insolvency law are especially salient for large, complex and
internationally active financial institutions, and the Financial Stability Forum has been
considering issues in anticipating and managing such problems. Both general and financial firmspecific
insolvency regimes are relevant. While the principal operations of such institutions are
primarily housed in supervised entities, such as banks and securities firms, significant activities
may take place in unsupervised holding companies or affiliates. Furthermore, there are now nonfinancial
companies that undertake financial activities on a scale that approaches those of major
financial institutions. These include General Electric and the worlďs major automobile
companies, as well as the recently insolvent Enron. Furthermore, financial firms and markets
often have large exposures to non-financial firms, as both creditors and counterparties. Finally,
many countries do not have special purpose financial insolvency law, but rather treat bank
insolvencies with minor modifications, if any, to their general insolvency system.
This report discusses the insolvency process - particularly that of banks - in countries with well
developed economic, financial and legal infrastructures.
A. Role of the Contact Group
To explore these issues, the BIS formed the Contact Group on the legal and institutional
underpinnings of the international financial system (the Contact Group). The Contact Group was
formed in response to a suggestion at a G10 Deputies meeting by the Bank of Italy. Participation
was voluntary. The members are representatives of the European Central Bank, the Bank of Italy,
the Bank of Japan, the Netherlands Bank, the Bank of England, the Federal Reserve Bank of New
York, the Bank for International Settlements, the International Monetary Fund, the Organisation
for Economic Cooperation and Development and the World Bank.
4
The Contact Group has considered the legal and institutional arrangements for resolving the
insolvency of financial institutions and of non-financial institutions that have substantial financial
activities. The Contact Group has examined both general insolvency regimes and special regimes
applicable to financial insolvency. Some of these financial insolvency regimes are applicable to
particular categories of institutions (such as banks and non-financial corporations with significant
financial activity). Others apply to a particular type of transaction or operation (such as the taking
of collateral, the finality of settlement of transfers of funds or securities and netting transactions).
A particular focus of this examination has been the resolution of institutions that conduct a wide
range of activities on a large scale in many countries.
The Contact Group sought to explore two areas of the legal system of direct relevance for
financial transactions: the nature of insolvency proceedings and rules governing contract
enforcement in insolvency (including netting and collateral), first at the national level and then at
the international level. The effort has involved two stages. Since law and legal rules are based on
nationally defined jurisdiction, the Contact Group has conducted two comparative surveys of the
legal situation in the jurisdictions represented within the group, namely the European Union,
Italy, Japan, the Netherlands, the United Kingdom and the United States. Then, on the basis of the
responses to the questionnaires, a number of policy issues have been identified and discussed.
The first survey compared legal and institutional arrangements for insolvency in different
jurisdictions (the "insolvency survey"). The second survey compared the effect of insolvency
arrangements on the performance of financial contracts (the "contract enforceability survey") in
different jurisdictions. Appendix B contains analyses of these surveys.
B. Summary of findings
This report offers reflections on national insolvency regimes and the coordination of an
international insolvency based on the results of the questionnaires, a review of academic literature
and further analysis. It describes current insolvency law and its context from an economic
perspective, and evaluates legal rules governing insolvency by three criteria: efficiency of the
insolvency process, equity of treatment and reduction of uncertainty. Using this analysis, the
Contact Group has identified two salient issues, discussed in turn in Sections II and III of this
Report. It also discusses law reform efforts addressed to these issues, in Section IV.
1. Business environment and insolvency law
The first issue identified is the gap developing between the rapidly changing environment in
which insolvencies occur and the slower evolution of national insolvency regimes. These
regimes, largely developed in the 19th or early 20th century, have evolved over time. However,
they have not kept pace with the dramatic changes in capital markets, globalisation, corporate
governance and markets for corporate control, and techniques of financial management. On
balance, the gap increases the demand for legal certainty and efficiency on the part of market
participants.
5
Perhaps the most important source of uncertainty and inefficiency lies in the slowness of
traditional insolvency processes. Insolvency processes tend to be initiated later than they should
be, and to be very slow after their initiation. Both earlier initiation and faster resolution tend to be
more efficient; faster resolution decreases legal uncertainty. Late initiation is a problem of
incentives. Creditor incentives to initiate insolvency are strong, and creditor initiation is therefore
made difficult in insolvency regimes. Debtor initiation is easy, but debtors seldom have incentives
to declare early insolvency. Regulatory incentives are blunted, unless the regulator has a direct
financial stake, eg as an insurer. Slow resolution is usually inherent in the ordinary legal rules
governing insolvency, although some specialised bodies of insolvency law do emphasise speed,
especially Japanese and US bank insolvency law.
The demands for legal certainty and efficiency are especially great for risk transfer mechanisms,
such as derivatives contracts, which require active management. The development of a body of
law on financial contract enforceability - still in the process of being extended and refined - is
meant to address these issues. However, because such new law addresses only some of the issues
and only some of the assets and liabilities of financial firms, it creates some tension with existing
insolvency processes, because in effect it alters the priorities in insolvency. Enhancing the legal
certainty and efficiency of the insolvency process beyond the current rules for financial contracts
might relieve the tensions, without necessarily changing the equity concepts in national
insolvency law.
2. Globalisation
The second issue stems from the increasing globalisation of financial activities and the global
scope of financial institutions in a legal environment still defined by national jurisdictions.
Sovereign jurisdiction over global activities raises the possibility of forum shopping.*
Forum
shopping sometimes creates a healthy competition among jurisdictions. But it can also increase ex
ante uncertainty and creates a tendency for the relevant legal rules to be the ones most favourable
to the party who can select the insolvency forum. Efficiency can decrease because the
administration process can be complex and contentious and it may adversely affect incentives for
creditors, debtors and regulators ex ante and ex post. As opposed to the national setting, where
improvements in the general insolvency process can at least in concept be equity 00000neutral, in
the international setting solutions to multi-jurisdiction issues need analysis of equity questions.
Even without forum shopping, globalisation creates many problems. Some of the problems are
strictly legal, such as the uncertain choice-of-law rules governing collateral or netting. Other
problems are inherent in the coordination of disparate insolvency proceedings.
*
Various technical economic and legal terms, when first introduced, are boldfaced. These terms are defined in a
glossary, in Appendix D.
6
3. Law Reform
Many worthwhile insolvency-related reform efforts are already in progress, at both the national
and international levels. For example, many jurisdictions are reducing settlement risk by
improving their treatment of netting and closeout of financial contracts. Difficult choice-of-law
issues are being addressed by international law reform efforts, emanating from organisations such
as the European Union, the Hague Conference on Private International Law ("Hague
Conference") or the United Nations Conference on International Trade Law (UNCITRAL). As
part of this process, two EU directives and an insolvency regulation recently set out an approach
for coordination of international insolvency processes that identifies a main insolvency
proceeding for financial and non-financial firms within the European Union. Many jurisdictions
are streamlining their insolvency procedures, with an eye towards adopting "best practice" from
other jurisdictions.
II. Analysis of national insolvency processes
This section discusses insolvency processes of advanced jurisdictions. It is couched at a very
general level, and is concerned with the structure and organisation of the processes rather than the
specific legal rules associated with them.
It begins with an analytic framework by which to evaluate insolvency processes. The second
subsection then describes insolvency processes, both mainstream and innovative. The mainstream
processes (piecemeal liquidation and most forms of reorganisation) make some implicit
assumptions about capital and related markets, globalisation, dynamism and the distinctiveness of
financial services. As discussed in the third subsection, these tacit assumptions, although once
probably true, seem outdated today. Some of the innovative processes may better accommodate
the new rules of the game. The fourth subsection discusses the special insolvency rules pertinent
to financial contracts. The section concludes with a recapitulation of some of the issues, phrased
in terms of the analytic framework developed at the beginning of the section.
A. A framework for analysis
The process of resolving a troubled company, no matter whether or not it takes place under a
formal insolvency regime, generally involves retrenchment and focusing on viable operations,
and closing down or selling off operations that are either unprofitable or ancillary. It commonly
also involves a restructuring of liabilities. The basic issues of insolvency can be most easily
described in a timeline of financial distress for a firm, as shown in the following diagram1
:
1
The timeline of financial distress is adapted and extended from Matthias Kahl, "Economic distress, financial
distress, and dynamic liquidation", Journal of Finance 62 (February 2002) pp.135-168.
7
time
A financial institution or financially active firm that experiences emerging problems - worsening
loan quality, losses due to internal control lapses - has the option to develop business plans to
address the underlying problems. Many problems are in fact resolved this way. If uncorrected,
however, or if some other devastating blow occurs, the firm's difficulties may be severe enough
to require a private workout.
A private workout usually involves some restructuring of debt, the possible issuance of additional
equity, and sales of assets or businesses. Traditionally, a firm's "main bank" could pressure a
liquidity-constrained firm to enter a workout, and in many cases where creditors rely primarily on
a `main bank', the ability of banks to encourage workouts persists. In the case of large,
internationally active financial institutions and financially active firms, no one creditor has
sufficient influence. But increasingly, far-seeing managements recognise the need for
restructuring and initiate management-led private workouts in order to realise the maximum longrun
value of the firm for shareholders. A notable example was the successful extensive
restructuring of IBM in the early 1990s; currently many financially strong telecommunications
companies are restructuring in the light of the industry's present overcapacity.
As the firm moves closer towards or into insolvency, it may seek to reorganise through an
insolvency law procedure. Most frequently, management initiates such an insolvency
reorganisation. A reorganisation gives management time and powers to restructure an insolvent
firm as a going concern, effectively offering it a second chance to make the firm successful. If a
firm is viable and can be turned around, a reorganisation benefits many creditors if the overall
losses are lower than in a liquidation. If creditors are uncertain about the viability of the firm, they
may still look favourably on reorganisation as a means of developing further information about
the firm's viability.
Finally, usually either the creditors or the debtors can initiate a liquidation process if the legal
condition(s) for insolvency is (are) satisfied. An important exception is banks and some other
financial firms, where, in some jurisdictions, only the regulator or insurer can initiate insolvency.
Acute and
worsening
problems:
private
workout
Emerging
problems:
managementled
correction
Insolvency: but
possible
viability:
reorganisation
Insolvency and
unlikely
viability:
liquidation
8
The entire timeline can be viewed as steps in a process of "dynamic liquidation"2
- firms may
recover and exit the process, they may work their way through the entire process through a series
of business disappointments, or they may leap to the final stage of liquidation. At each stage the
firm traverses, the existence of the statutory possibility of liquidation provides an incentive for
debtors and creditors to cooperate in private workouts and reorganisations, which thus take place
"in the shadow of the law". The extent to which formal insolvency procedures are debtor or
creditor friendly affects the amount and nature of the risk that borrowers are prepared to take and
the terms and conditions on which lenders are prepared to provide money.
The statutory frameworks dealing with insolvency can be explained by three underlying
objectives of insolvency regimes: the reduction of legal and financial uncertainty, economic
efficiency and the fostering of equitable treatment. Legal certainty and efficiency, and to some
extent equity, contribute to lowering liquidity and systemic risks in that they reduce the potential
for market disruption and large deadweight losses. The three dimensions are not strictly speaking
mutually exclusive, but are relatively separate and intuitively appealing. Despite their overlap,
some tension between them is inevitable. Each national legal framework tries to strike a balance
between them on the basis of local preferences and tradition.
Legal uncertainty affects the ability of market participants to form a probability distribution
around the outcomes of financial transactions. If legal uncertainty is too great, market participants
cannot assess the risks in entering into contracts. In such circumstances, it may be very difficult to
talk meaningfully about the efficiency or the equity of an insolvency regime. If legal uncertainty
is sufficiently low, but not necessarily very low, market participants can estimate the probability
of possible outcomes and make choices based in part on their willingness to bear risk. Reductions
in legal uncertainty, therefore, generally represent an improvement for debtors, creditors and
other stakeholders. Nonetheless, the benefits from increased legal certainty need to be weighed
against any negative impact on efficiency or equity associated with the rules used to enhance
legal certainty.
An improvement in economic efficiency can be achieved chiefly in three ways. One is aligning
incentives for all parties to a transaction to maximise the long-run value of their respective firms
and reducing incentives to waste resources in order to avoid the consequences of failure on the
firm's management or owners. In general, incentives that align actions with market outcomes are
seen as improving efficiency (except in markets where market failure is inherent). A second is
reducing the transaction costs involved in the insolvency process, especially deadweight cost.
Finally, efficiency is improved by any other means that increases the total amount of wealth of all
the parties in the insolvency, eg., rules or practices that preserve the value of firm-specific assets
or increase information available to the parties.
2
This extends the concept set out by Kahl in the article cited in footnote 1 to describe the relationship between
reorganisation and liquidation.
9
Issues of equity by their nature involve value judgments. In the case of insolvency law, that
judgment has already been made in national laws and accepted practices which define the rights
of debtors and the rights and relative priority of creditors and other stakeholders. Some
jurisdictions seem more "debtor-friendly," in that they give management more opportunity and
ability to protect the value of the firm's assets; others seem more "creditor-friendly" in that
creditors (especially secured creditors) have greater ability to initiate and control an insolvency
process or realise their collateral in order to maximise recoveries from the debtor's estate.
In the insolvency context, a fairly conservative standard of determining improving outcomes
seems to work well. That standard is that one outcome is better than another if anyone is better off
and no one is worse off in the first outcome than they are in the second, which economists call
Pareto superiority. This standard works well since insolvency laws assign the order in which
claims are paid and generally all claims of senior creditors are satisfied before the claims of more
subordinate creditors are paid. Thus, if the insolvency estate is larger in one situation than
another,3
and the pre-existing debtor-friendliness and priority of claims is maintained, some
creditors will receive more and all creditors will receive at least as much in the first situation as in
the second.
B. The processes
1. The basic processes: workouts, liquidations, reorganisations and others
Insolvency law and practice have evolved through accretion of experience. Much of this
experience has been national, adapted to a particular jurisdiction's unique traditions and social
needs. But despite the diversity of insolvency law and practice, some common themes apply to all
jurisdictions. All business insolvency laws must face the same problem: resolving the competing
claims of stakeholders when the debtor firm may not have sufficient resources to satisfy all claims
on time and in full.
Insolvency law has two traditional solutions to this problem: the "liquidation" and the
"composition" (or "reorganisation".) There are many traditional solutions outside insolvency
law, of which the "workout" is most significant. A liquidation converts the insolvent entity's
assets into cash, and distributes the proceeds to claimants. A composition or reorganisation seeks
to preserve the entity and readjust the claims of stakeholders so that the entity can successfully
meet the readjusted claims. A workout is similar to a composition or reorganisation, but proceeds
outside insolvency law. A liquidation often serves as the backstop to an unsuccessful
3
Technically, it is not the absolute size of the insolvency estate that is relevant, but instead the size of the estate
minus the size of the claims. This distinction is particularly significant with asset securitisations and netting
arrangements. Both devices decrease both the size of the estate and the number of claims against the estate, and
may well be Pareto-superior. For simplicity, this report refers to "size of the estate."
10
reorganisation or composition. A reorganisation, in turn, often serves as the backstop to an
unsuccessful workout.
To introduce various processes for dealing with financial distress, we shall begin with the classic
distinction between liquidation and reorganisation. For our purposes, a liquidation acts on the
assets side of the balance sheet. A liquidation has two steps: reducing the insolvent firm's assets
(or business lines) to money, and then distributing the proceeds to creditors, in accordance with
priority rules. As a result, the firm ceases to exist as an entity, although its business may be
carried on by those who acquire its assets. A reorganisation, in contrast, preserves the firm as an
entity. A reorganisation may affect assets, but operates on the liabilities side of the balance sheet.
A reorganisation temporarily postpones liabilities, and uses the time gained to readjust the claims
on a firm so that they correspond better to the firm's value and expected cash flows. A
reorganisation does not necessarily affect the assets side, although asset and business line sales
are common in reorganisations.
Liquidations often disrupt the going-concern value of an insolvent firm, but need not do so. In
principle, the whole firm can be liquidated as a unit, preserving its economic value, albeit perhaps
at distressed prices. But as is often the case, the troubled firm may not have much of a goingconcern
value or a bulk liquidation may be otherwise infeasible. In such cases, a piecemeal
liquidation may be preferable, yielding a higher return to creditors, or at least a quicker return.4
Reorganisations are more complex, and may be divided into a number of categories. The most
traditional category of reorganisation - the "composition" - is typically consensual, although some
compositions may permit majorities to bind minorities.5
In this process, the creditors negotiate
with each other (and generally the debtor) to readjust the relative stakes in the organisation. The
composition therefore resembles a debt workout, although it is generally aided by a courtimposed
moratorium on debt collection procedures. Compositions suffer from most of the same
weaknesses as sovereign debt workouts, and succeed - if they do - for the same reason: solidarity
among creditors. This solidarity is achieved by use of creditors' committees composed of major
creditors, and the common practice of paying many minor creditors (eg trade creditors) in full.
Nevertheless, compositions are at risk of holdout. Another class of reorganisation lowers this risk
considerably. Such reorganisations generally resemble negotiated compositions. But if the
negotiations fail, these reorganisations allow for substantive judicial intervention: binding
holdouts who resist an appropriate majority of their class of creditors, or even imposing a new
4
One negative externality of piecemeal liquidations of banks is worth mentioning in margin. A piecemeal
liquidation may reduce the availability of credit to smaller borrowers. These borrowers are forced to repay the
insolvenťs estate, but often cannot find replacement credit, because their credit information is lost in the liquidation
proceeding. Although competition assures that this problem is usually temporary, it may be lethal to some
otherwise viable small businesses.
5
The old German Composition Code, for example, permitted a 75-80% majority to bind a minority. This was
repealed on 01 January 1999.
11
capital structure on the firm if negotiations fail ("cram-down"). We shall call these
reorganisations "Chapter 11-style reorganisations", because they were popularised in Chapter
11 of the United States Bankruptcy Code. For most purposes, this report will not distinguish
between compositions and Chapter 11-style reorganisations, because the similarities between
these two techniques far outweigh the differences. Indeed, many jurisdictions use various hybrids
of the techniques discussed here.6
Indeed, the debt workout, the composition and Chapter 11-style reorganisation can all be viewed
as part of a continuum. They all rely on negotiation between the debtor and creditors. All of them
risk failure from holdout problems. In a bilateral workout (which is common for small firms with
a single bank creditor), the holdout risk is between debtor and creditor. The more complex
reorganisations share this risk, but also risk holdout by a minority of creditors. To cope with this
problem, these different reorganisation processes employ different mechanisms of creditor
solidarity. The workout uses a purely private mechanism: limiting the creditor class by excluding
numerous (but low-value) creditors such as employees or trade creditors. These creditors are
usually paid in full notwithstanding any formal parity between their claims and the claims of the
financial creditors. (In some jurisdictions, their claims would have priority in any case.)
Nevertheless, despite assuming an inferior position, financial creditors negotiate because
negotiation is better than the alternatives: liquidation or a rush for assets. The other proceedings
tend to mirror this approach (weak creditors do not negotiate but are often paid in full), but add
additional techniques. The composition adds a judicial moratorium on debt collection. This
increases creditor solidarity by precluding a rush for assets, leaving liquidation or negotiation as
the sole alternatives. The Chapter 11-style reorganisation retains this moratorium, and contains a
cram-down. Some variations of this procedure allow for an insolvency official to create its own
reorganisation, if the parties cannot negotiate their way to one.
The liability transfer is very different from reorganisations, but leads to the same result: a
transformed balance sheet. In this technique, an insolvency official (or court) assigns the old
liabilities, and often, some portion of the old assets, to a solvent organisation. Liability transfers
are particularly useful when the liabilities are contingent (eg. insurance or letter-of-credit
liabilities), when the liabilities themselves are vested with going-concern value (eg a consumer
bank deposit with appreciable switching costs), or when the liability structure is well-suited to the
assets it supports (inexpensive funding or effective maturity matching). A liability transfer results
in a restructured balance sheet, without the need for negotiations.
There are several kinds of liability transfers. We begin with the simple merger. A distressed firm
may be purchased by another firm with a stronger balance sheet, transferring the assets and
6
For example, in the United Kingdom the reorganisation resembles a composition in some respects: lack of cramdown
and moratorium on debt collection or liquidation of collateral. However, it does not contain a continuing role
for existing management, and is less negotiation-intensive than either a classical composition or a Chapter 11-style
reorganisation.
12
liabilities of the old firm into the new one. (Mergers are very common in declining industries that
need to restructure and in dynamic industries in which comparative advantage shifts rapidly.) The
merger is a private transaction that only works when a distressed firm has some net going-concern
value, or has powerful synergies with its acquirer. However, other kinds of liability transfers,
which involve public law, may be used with insolvent firms in general. The "bridge bank"
technique used in US bank insolvency law is akin to a merger, but only a subset of the distressed
firm's assets and liabilities are transferred into the bridge bank. The rest of the balance sheet
remains behind, and is conventionally liquidated. A bank regulator or liquidator decides which
assets are transferred, and which remain behind. Or in some cases (the "liability-based
restructurings" of insurance law), individual assets and liabilities are transferred to a buyer,
without a bridge institution.
As a final alternative, an insolvent firm may be reorganised with public funds. The public funds
may either facilitate an otherwise uneconomic merger, or may recapitalise the firm so that it can
exist as a stand alone entity. This method can always be exercised by a specific legislative
appropriation of funds, but is subject to formal legal regulation only in Japan and the United
States (which impose highly restrictive conditions). Because of its intended relative rarity, we do
not analyse this alternative in this report.
2. Distinctions among the traditional processes
As discussed above, many of the differences between liquidations and reorganisations are selfevident.
Liquidations conform to the statutory distribution scheme; reorganisations typically treat
different classes of creditors in very different ways, depending on their negotiating strength and
holdout power. The distribution of rights and powers affects the efficiency of the process, in
terms of costs, incentives and the amount of post-insolvency wealth.
One important distinction is the respective role of the creditors and the person charged with the
liquidation or reorganisation (eg an insolvency "Official" - sometimes called an administrator,
liquidator, trustee or receiver - or a court). In a liquidation, the creditors do not act collectively,
except perhaps to appoint, monitor, or receive information from the Official. Disputes are
bilateral ones, between the Official and individual creditors on the one hand, and the Official and
the debtors of the insolvent, on the other. In many reorganisations, the creditors are active, often
acting collectively through creditors' committees. In these reorganisations, the courťs (or perhaps
Officiaľs) role is less central, and is not always necessary (as is the case for some compositions).
This leads to a useful distinction: that between a stakeholder-centred and an Official-centred
insolvency process.
As a general rule, the Official-centred model tends to be dominant in liquidations and liability
transfers, including many bank insolvency proceedings. The Official-centred model is also a
norm in many jurisdictions' reorganisation procedures. Traditional compositions and Chapter 11style
reorganisations tend to be stakeholder-centred. In such proceedings, incumbent management
tends to administer the insolvent firm, creditors' committees are institutionalised, and creditor
13
consent (at least a supermajority in each class) is usually expected, if not required as a matter of
law.
This discussion is summarised in the Table 1:
Table 1
Liquidations Reorganisations
Official-centred ˇ Conventional and bank
liquidations
* A speciality of bank insolvency
law
* Sometimes seen in Chapter 11style
reorganisations
* Characteristic of traditional
European reorganisation
proceedings
Stakeholder-
centred
* In conventional liquidations,
only seen in governance, ie the
reporting relation of the Official.
* Workouts
* Conventional compositions
* Explains most aspects of Chapter
11-style reorganisations
The structure of insolvencies affects the degree of legal certainty. The negotiated stakeholdercentred
processes are inherently unpredictable both as to timing and as to outcome. Insolvency
negotiations are slow. The outcome of negotiations is ex ante indeterminate, depending in part on
the relative skills of the negotiators, in part on the holdout power of the negotiators, and in part on
uncertainty in asset and liability valuation. Official-centred processes vary. Some of them - eg
liquidations - are relatively determinate, with claimants' entitlements depending on the asset
value of the firm and relatively predictable legal rules regarding the priority of claims. Some
Official-centred reorganisations can be slower and more indeterminate than stakeholder-centred
processes. Others eg US Official-centred bank insolvency proceedings) can be extremely fast,
with relatively determinate distributions.
With respect to the nature of the process and its transaction costs, traditional liquidations and
reorganisations have several common elements.
Both processes are highly lawyered, even in jurisdictions that use accountants as the primary
insolvency professionals. Liquidations might involve a substantial amount of litigation, because
debtors of insolvent firms often refuse to pay their debts voluntarily. They see little reputational
cost to doing so, expect no ongoing relationship with the insolvent entity, and are likely to settle
only after litigation. Also, the Official may well press claims - such as avoidances of preferential
14
or fraudulent transactions - that did not exist prior to insolvency. There is yet more litigation on
the liabilities side: the Official is likely to dispute claims, especially contingent claims or
undersecured claims, which may involve valuation of the security interest.
Reorganisations may require less litigation (at least of the claims collection variety), but require a
great deal of negotiation. (Chapter 11-style reorganisation negotiations typically occur in the
shadow of litigation, because an insolvency court must ultimately decide, if the stakeholders
cannot agree among themselves.) And reorganisations are prone to their own litigation, especially
that of undersecured creditors (who will argue valuation) and others who try to escape the
proceeding entirely.7
As a result, these processes are usually slow. Some insolvency processes are inherently very slow
because of the nature of the insolvent firm's balance sheet. An insurance liquidation, for example,
must take a long time because of the large number of long-tail contingent liabilities on an
insurer's balance sheet.8
Similarly, real estate often takes a long time to liquidate, prolonging the
liquidation of any firm with appreciable real estate on its books. Asset collections are often
slowed by litigation, and the liquidator can be swamped by details so that insolvency processes
can be slow even for firms with few contingent liabilities and liquid or short-term assets.
Stakeholder-centred reorganisations are inherently slow because the necessary negotiations are
complex, and any adjudication can be even more complex. Also, some parties to the negotiation especially
junior claims such as subordinated debt and equity - tend to benefit from holding out.
Junior creditors have an effective option on the firm during the pendency of the negotiation
process. They will benefit if the firm increases in value during this process, and are unlikely to
lose much if the firm decreases in value. The value of this option increases with time: an
incentive to delay. As discussed above, Official-centred reorganisations can vary tremendously in
speed, from extremely rapid to extremely slow.
Finally, both Official and stakeholder-centred processes typically suffer from serious governance
problems, albeit of different kinds, which reduce efficiency. The governance problems of
Official-centred processes are inherent in the protracted complex multilateral negotiations that are
central to the processes. Official-centred processes do not have notably high negotiation costs, but
suffer from agency costs associated with administration. Private Officials are fee-charging
professionals, hired by a creditors' committee or a court; public Officials may have their own
agency costs. Adequate incentives may not be place to ensure sufficient monitoring by creditors
7
The scope of an insolvency proceeding can be hotly contested, largely because of past transactions. An insolvency
Official will seek to characterise these transactions as still open under insolvency principles, such as those based on
preferences or fraudulent transactions; the counterparties will seek to characterise them as final and not subject to
insolvency.
8
Thanks to closeout netting, banks' balance sheets do not suffer from this problem, notwithstanding the long-term
contingent nature of many derivatives contracts. The major concern for banks is with standby letters of credit and
other independent bank guarantees, which generally expire within a few years of issuance.
15
or courts.9
Similar agency costs attend stakeholder-centred processes, with their creditors'
committees represented by professionals.
However, the most significant governance problem of any of these processes probably arises from
their initiation, rather than their administration. By general consensus, insolvency procedures tend
to be initiated too late. Too often, reorganisations do not result in profitable restructured firms.10
In principle, if a liquidation is initiated at the moment of balance sheet insolvency, creditors
should receive 100%, net of their administrative costs. In practice, the yield on liquidations is
substantially less for banks, and far less for non-financial firms.
Insolvencies are often initiated by the debtor. The debtor owes its loyalty to the equity holders,
who - as residual stakeholders - have only to gain by delaying insolvency in hope of resurrection.
(For much the same reason, the management of a weak firm would prefer high-variance
strategies, even if they have a negative net present value.)11
Different jurisdictions have different
strategies for reducing this problem. Some insolvency systems punish the management of
insolvent firms, eg through management liability for insolvency, or for trading after insolvency.
This may work in some cases, but gives management a powerful incentive to conceal insolvency
in others. Chapter 11-style reorganisations favour the carrot, rather than the stick, and give
incumbent management a major role in the reorganisation. However, as discussed in margin
above, Chapter 11 reorganisations are probably not timely initiated, either.
Creditors do not have the same incentives for late initiation. Indeed, their incentives would be for
initiating insolvency proceedings at the earliest possible time. However, creditors face two
handicaps in initiating insolvency. First, they usually have less information on the firm's
condition than the debtor. Second, they cannot immediately initiate insolvency when they suspect
the firm is troubled: they must generally wait for a violation of loan covenants or a failure to
make timely payment. Debtors will often not accept loan covenants with hair triggers, and even if
they do, violations may not be readily apparent. Also, some legal systems are solicitous of
possible strategic behaviour by a creditor. A creditor - especially one without an ongoing
relationship with the firm - might wish to initiate insolvency for reasons completely unrelated to
the prospects of the debtor firm, eg to escape an unfavourable change in interest rates. Therefore,
9
In one recent international bank insolvency proceeding, the receiver had collected around $10 million, with all but
$1.75 million disbursed to it in administrative expenses (In re Treco, 240 F.3d 148, 159 2d Cir. 2001). Creditors
had not yet been paid anything. This case is certainly not typical, but is illustrative. For a perhaps more typical
example, see "Creditors fear fees tied to bankruptcy of global", The New York Times, 13 April 2002, p B1.
10
Depending on the metric, somewhere between a fifth and a third of US Chapter 11 plans fail. Put another way, the
average reorganisation produced firms that showed no profits. See Lynn M LoPucki and Joseph W Dougherty,
"The failure of public company bankruptcies in Delaware and New York revisited", Vanderbilt Law Review
(forthcoming), at http://www.law.ucla.edu/erg/pubs/Lopucki-DelawareRefilings-20020308.pdf
11
Michael Jensen and William Meckling, "Theory of the firm: managerial behavior, agency costs and capital
structure", 3 Journal of Financial Economics 305 (1976).
16
the insolvency laws of some jurisdictions make creditor initiation ("involuntary insolvency")
more difficult and linked to more objective evidence.12
Bank insolvency law contains a third alternative: initiation by a neutral governmental official. In
principle, governmental officials - especially supervisors - could have the correct mixture of
information and incentives for timely insolvency initiation. But also in principle, timely initiation
implies that creditors get paid in full. It is hard to escape the conclusion that even bank
supervisors are in some instances too slow in initiating insolvency. Even bank supervisors are
hindered, by limited information, bureaucratic inertia, political pressure, risk aversion and
perhaps a reluctance to declare their pre-insolvency rehabilitation efforts a failure. However, they
may be less imperfect than other parties, especially if the bank supervisor is also a bank insurer,
and has a direct financial incentive to initiate timely insolvency.
3. National approaches to insolvency processes
A few of the approaches to insolvency discussed above are universally available in the surveyed
jurisdictions. All jurisdictions surveyed permit mergers and workouts, as matters of private law
outside formal insolvency proceedings. Similarly universal is the piecemeal liquidation.
Reorganisations and compositions of some kind are in fact universal.
Table 2 summarises a survey by the Contact Group.13
In the table "UK" refers to the law of
England and Wales, and "US" law is an amalgam of federal law, New York State law and the
Official Text of the Uniform Commercial Code.
12
The title of one article describing US law is instructive: Brad R Godshall and Peter M Gilhuly, "The involuntary
bankruptcy petition: the worlďs worst debt collection device?", 53 The Business Lawyer 1315 (1998). Other
jurisdictions are more creditor-friendly, eg that of England and Wales.
13
The survey results are in Appendix B.
17
Table 2
IT JP NL UK US
General liquidation Y 2 Y Y Y
General composition Y Y Y 2 N
General Ch 11 reorg 2 2 N N 2
General liability transfer N N N N N
Bank liquidation Y N* N* N* 3
Bank composition N N N N* N
Bank reorganisation Y N* Y N* 2
Bank liability transfer Y Y Y N Y
* No special bank proceeding. Jurisdiction uses general insolvency law for banks.
** Not controlled by legislation, but available through the political process.
"2" and "3" refer to the number of processes under this heading.
As shown in Table 2 some insolvency processes are less universal than others. Bank liability
transfers, for example, have no equivalent in general insolvency law.
Interestingly, many jurisdictions have several classes of specialised proceedings. Bank and
insurance proceedings are frequently specialised, but there are also specialised distinctions within
general insolvency law, based on criteria such as size or significance. These distinctions are
understandable: the liquidation of a large public corporation will be very different from the
liquidation of a small shop. Some of these specialised proceedings are obsolete or used
infrequently; some are fairly common.
4. Bank insolvency law
The insolvency survey shows that countries have chosen different approaches to the treatment of
insolvent banks.14
Some jurisdictions (eg the United Kingdom and, to some extent, the
Netherlands) treat bank insolvency similarly or identically to ordinary insolvency. Other
jurisdictions appear to have very distinct legal regimes governing bank insolvency, particularly
the United States, Italy, and Japan. In some of these jurisdictions, the insolvency of credit
institutions is regulated in the banking laws as a specific subset of insolvency rules, which may be
14
See Appendix B.
18
supplemented by general provisions applicable to all insolvency cases in the jurisdiction in
question.
There is one universal distinction between bank and ordinary insolvency. In the insolvency law of
any of these jurisdictions, a bank regulator can initiate a bank insolvency proceeding, whether it is
conducted under general or specialised insolvency law. However, if we look to trends, rather than
universals, another distinction appears.
The slower kinds of proceedings - especially the stakeholder-centred ones - are disfavoured in
bank insolvency law and the existing faster options - liquidations, mergers and liability transfers are
the norm.15
Chapter 11-style reorganisations appear disfavoured in bank insolvency law, and
compositions seem extremely rare or nonexistent. One reason may be that de facto bank
reorganisations often take the form of bank supervisor-overseen (or -encouraged) workouts,
outside any formal insolvency process, or with the formal approval of an insolvency court as the
last step. In concept, however, there is no reason why procedures for an expedited stakeholder
centred approach could not be developed. Forced mergers are common, even in jurisdictions
whose bank insolvency procedures differ little from conventional insolvency procedures. Some
jurisdictions (the United States and Japan) rely on the most activist procedure of all: the liability
transfer. These fast proceedings often have supervisory involvement.
There may be several reasons for these distinctions in insolvency process between banks and nonfinancial
firms.
Liquidity (and systemic risk) is almost certainly the most important of these reasons. Prolonged
bank insolvency procedures destroy liquidity: either the liquidity of counterparties or that of
markets. This raises the spectre of systemic risk through the very pendency of the insolvency
process. Systemic risk creates a powerful need for rapid insolvency proceedings, deposit
insurance, or both. Rapid insolvency proceedings - be they liquidations, recapitalisations or
liability transfers - are best conducted by an activist Official.
It should be noted that bank reorganisation techniques that call for general moratoriums are
particularly inappropriate to modern banking's need for liquidity. Moratoriums inherently block
the payment stream, and for derivatives and foreign exchange contracts may be difficult to
implement given that contracts can fluctuate from in-the-money to out-of-the-money and back
again. Such moratoriums are also hard to reconcile with the needs of households and small
business depositors and borrowers. Moratoriums have their necessary place in bank liquidations they
stop the run on the bank - but are singularly ill-suited to bank reorganisations.
Pervasive supervision may be significant. As discussed above, supervisory intervention,
conducted well before the appearance of obvious financial distress, can be a good substitute for
15
This is in contrast to insurance insolvency law, which often proceeds extraordinarily slowly, in large part because
of long-tail contingent insurance liabilities.
19
rehabilitation. Regulation also plays a significant role in legitimising activist insolvency
procedures, such as a liability transfer or overt government funding. Concerns about
administrative intervention in the economy tend to be muted in the financial services sector.
Firm-specific assets, other than human capital, are less common in modern banks, as the loan
book of a bank becomes increasingly transparent, thanks to the syndication and bond markets,
and a widespread market for credit information. (This trend is still limited by close bank-firm
relationships in some countries.) This greater transparency aids rapid sale of particular
components of the balance sheet or particular operations. It also makes the liquidation of a bank
as a whole easier, thus strengthening the case for liquidation, especially if a supervisor-overseen
or encouraged workout has been unable to conserve going-concern value. (However, piecemeal
liquidation will dissipate firm-specific human capital.) Finally, it argues that insolvency
moratoriums should not apply to security interests in bank collateral. A moratorium is justifiable
if the security interest is in a firm-specific-asset, such as machinery. Rapid foreclosure of such a
security interest would destroy value. However, a moratorium is much more difficult to justify if
the security interest is in a non-firm specific asset, such as most receivables and almost all
securities. (The traditional justification for extending the moratorium to non firm-specific assets is
that it serves as a temporary loan to the insolvent reorganisation. This justification implicitly
assumes that no outside capital will be available to the insolvent firm: an assumption belied by
modern capital markets and an active market in post-insolvency lending, such as debtor-inpossession
financing.)
Bank creditors, at least traditionally, have been dispersed and weak. The governance problems
of creditors' committees - always an issue in insolvency law - would be particularly pronounced
in a bank insolvency. This problem is dispelled by supervisor-led proceedings, some of which
place the supervisor in the role of an Official.
C. What has changed?
The basic framework of mainstream insolvency law had become established by the 19th century.
The most significant advances in the 20th century were probably the Chapter 11-style
reorganisation (inspired by procedures developed on pre-war US railroad insolvencies) and the
"ancillary proceeding" of international insolvency law (developed in the wake of the Herstatt
insolvency).16
The elements of this framework were probably well-suited to the conditions in
which they were established.17
But do they equally fit the conditions of today?
16
Kurt H. Nadelmann, "Rehabilitating international bankruptcy law: lessons taught by Herstatt & Company", 52 New
York University Law Review 1 (1977).
17
The discussion in the main text assumes that the conditions that formed modern insolvency law were exclusively
economic. This is certainly too narrow a view: political considerations played a role. 19th-century insolvency law
was shaped by 19th-century classical liberal political theory. This law stressed the quick and orderly resolution of
the insolvent and the pro rata satisfaction of unsecured creditors from the assets of the estate. The main alternative
20
Many observers have highlighted the weaknesses of insolvency law, both for banks and for
general firms. Much of the criticism centres around reorganisations or compositions, slow, costly
and frequently ineffective.18
Banks do not usually share these problems, because their
reorganisations follow a different pattern: led or encouraged by supervisors and occurring before
financial distress becomes public and forces the bank into insolvency. But bank liquidations often
share the same problem as general firm liquidations: they are slow, costly and destructive of
going-concern value.
This subsection argues that discontent with insolvency law is related to its age. The global
financial system has evolved substantially in the last 20 years and the insolvency process has not
kept pace. Apart from cross-border issues, the mainstream insolvency process has not responded
to this evolution. International law reform efforts are plentiful, but innovation is largely in
settlement risk reduction and the resolution of cross-border issues. Procedural reform efforts have
been largely confined to adoption of Chapter 11-style reorganisations in jurisdictions without
them. This adoption process will substantially improve the insolvency procedures of these
jurisdictions. Nonetheless, the Chapter 11-style reorganisations have themselves been subject to
criticism, as insufficiently responsive to market developments of the last few decades.
Interestingly, most of these criticisms have come from the United States, the birthplace of the
Chapter 11 process.
This subsection examines developments in rough order of their significance to insolvency:
transformed financial markets, globalisation, a greater need for speed, and an increasing overlap
of the banking and general sectors of the economy.
1. Financial, capital and managerial markets
The case for reorganisation proceedings is based on market incompleteness and, to some extent,
market failure. If capital and managerial markets were complete and perfect, outside capital could
simply purchase and operate all of the insolvent firm's assets, with the purchase proceeds
distributed among the liability holders, after satisfying the claims of the secured parties. This
would preserve the entire going-concern value of the insolvent firm, with minimal transaction
costs and delay.
Of course, capital and managerial markets are not complete or perfect. Economic theory suggests
that firms exist to coordinate the production of goods and services where markets do not provide
that coordination. This suggests that a firm provides an externality in providing its output, which
to the classical liquidation was the composition, which - consistently with classical liberalism - placed party
autonomy at its centre. In the 20th century - particularly the latter part of the century - debtor protection and various
social considerations became increasingly important: hence the trends toward weakening the security interest,
creating wage priorities, and preservation of the debtor's enterprise. This Report does not speculate on the current
political underpinnings of insolvency law reform, but merely notes that they are likely to be significant.
18
See footnote 10 above.
21
is lost in the event of its failure - this is a motivation for giving priority to the rehabilitation of the
firm in reorganisation.
However, the scope of markets has expanded dramatically, the information set available to
participants has increased in both quantity and timeliness, and the functioning of markets has
improved tremendously over the last 20 years or so. Inevitably, better functioning markets make
assets less firm-specific, even if the assets themselves might be more difficult to evaluate.
First, the depth and liquidity of financial markets in general have increased, for several reasons.
Transaction costs and risks of exchanging financial assets have decreased: a result of
deregulation, improving technology and legal reform. The scale of financial firms and the growth
of the institutional investor sector have brought more risk-bearing capacity to markets. Several
jurisdictions have removed product segmentation among different classes of financial firms further
augmenting potential market-making capacity. Improvements in market liquidity and
infrastructure have been particularly notable in the equity markets, and the ability of firms to raise
equity capital at all stages of their life cycle improved notably. Activity in financial markets has
expanded as transactions have shifted from banking to securities markets and as new products
have developed to trade risk and assets. For example, low-transaction cost risk transfer products
have developed in markets as diverse as pure credit risk and reinsurance.
Second, the depth and liquidity of markets for non-financial assets, both tangible and intangible,
have increased: notably the market for firms (or business lines), and the related market for
corporate control. The market infrastructure for takeovers is sufficiently strong that the market is
limited only by the anti-takeover provisions of some jurisdictions' corporate law. Firms and
business lines are frequently auctioned. The availability of merchant banking and institutional
investment capital has made even the largest firms potential candidates for acquisition or
restructuring, and the market seems to provide adequate information on targets. Furthermore, the
market for managerial skills is becoming increasingly sophisticated. Lateral hiring of top
executives is a routine matter these days, facilitated by consulting firms (a good source of nonfirm-specific
top managerial talent) and headhunter firms, which act as brokerages.
Third, in a development particularly significant for insolvency processes, the market for
distressed financial and non-financial assets has developed and expanded dramatically over the
last two decades, at least in some countries. Specialist investors - such as vulture funds, and firms
that specialise in restructuring and remarketing the assets of troubled firms - have created a
market in which distressed assets can be priced and transferred at reasonable transaction costs and
reasonable speed. The markets for distressed assets rely on the wider array of markets in financial
and non-financial assets for the eventual resale of recovering or restructured assets.
22
Insights from the corporate finance literature into the boundaries of firms (what activities occur
inside the firm rather than in markets) have further weakened the case for reorganisation.19
This
literature often focuses on the problem of aligning incentives to produce goods and services, and
the role of firm-internal decision-making, coordination and control mechanisms in that process.
However, evolving corporate practices have highlighted the important role of sophisticated
contracts in solving such incentive problems, aided by modern information technology. The
widespread use of outsourcing, as well as the increased use of joint ventures and business
alliances, suggest that the scope for rapid restructuring and disposition of assets has increased
substantially.
In summary, most arguments for traditional insolvency reorganisation rely on the firm-specificity
of management, assets and capital. If management, assets and capital are sufficiently firmspecific,
any restructuring of the firm must rely on the firm's internal resources eg through
reorganisation proceedings. If outside help is not possible, the firm must rely on its incumbent
management, and liabilities must be restructured to fit the firms' assets and cash flows. But if
outside markets can provide adequate capital and management to distressed firms or absorb these
firms whole or piecemeal, the case for reorganisation or a highly administered liquidation process
weakens. Liabilities need not be restructured when an outsider can buy the firm and pay off the
liabilities in priority rank.
Although the case for traditional reorganisation has been weakened with time, it may still be
valid, at least in some cases, and perhaps many cases. Small non-financial firms cannot yet take
full advantage of modern capital and management markets, and only certain sectors of the labour
market are becoming less firm-specific. Of course, the smallest firms are usually resolved
informally in bank workouts, and reorganisations are reserved for firms with disparate
stakeholders. But there are many firms in the middle, and even the largest firms have many
specific assets, apart from team human capital. Moreover, as discussed in the framework section,
many creditors may actually prefer reorganisation if the viability of the firm is uncertain and more
information is needed to decide on bankruptcy.20
We cannot conclude that reorganisation is
obsolete; merely that it is no longer reflective of modern trends.
2. Globalisation
Law is promulgated by sovereign states, and affects those within the jurisdiction of those states.
At one time, most firms could be attributed to a single jurisdiction, as could their acts. However,
this is increasingly less true for modern firms, which have become international in both their
physical presence and the geographical scope of their activities. Furthermore, the markets in
19
See for example Oliver Williamson, "The economic institutions of capitalism" (1985); Oliver Hart, Philippe
Aghion and John Moore, "The economics of bankruptcy reform", Journal of Law, Economics and Organisation
8(3) (l992).
20
See text accompanying footnotes 1 and 2 above.
23
which firms operate are becoming increasingly international. Finally, more economically
significant acts are intangible, and thus difficult to accommodate to territorially-based
jurisdictions. General business law is strongly affected by globalisation, as witnessed by the
accelerating output of UNCITRAL Conventions and Model Laws. As with general business law,
so with insolvency law.
As larger financial firms have become global in the last few decades, their business and
management structure has evolved substantially. Extensive work by the Joint Forum, the
organisation of banking, securities and insurance supervisors, documents several relevant features
of many global financial conglomerates.21
First, such conglomerates organise and conduct their
business lines to operate across legal jurisdiction and geographical lines, with many jurisdictions
then having some claim to being the insolvency jurisdiction. Moreover, key decision-making and
financial and risk control activities are likely to be centralised in one or a few financial centres,
the result of both economies of scale and the sound practice of consolidated risk management.
Both practices are facilitated by modern communications and air travel.
Thus, even if the firm's business appears separable by jurisdictions,22
as a practical matter, the
business within a single jurisdiction could not be reorganised on a standalone basis. (This does
not preclude sales of business lines, but limits sales of business offices.) The firm's executives
across locations must cooperate for business strategy, management and risk control reasons. The
accounting consolidation of subsidiaries further reduces any jurisdictional compartmentalisation
of an organisation by subsidiaries. In any case, some organisations - notably banks - are unwilling
to employ the subsidiary device for their principal operations, which tends to raise the cost of
operations and degrade organisational creditworthiness. As a result of all these factors, the
business activities of financial firms frequently overlap jurisdictional lines.
While global non-financial firms may have more physical assets that can be easily attributed to
individual jurisdictions, their business line and management structure may have characteristics
similar to those of global financial firms. Moreover, their financial management may have many
similarities to global financial firm management. For example, some global non-financial firms
centralise their treasury operations in one or a few centres.
In addition, many financial markets are now global. The same financial products are traded
worldwide, among counterparties who may be anywhere, while settlement often remains located
in a single centre, depending on the product. This gives rise to complex cross-border relations,
which place stress on cross-border insolvency law. Collateral, for example, is often "located" in
different jurisdictions than the "location" of the debt it secures. Location is a difficult concept for
debt and many of the intangibles that comprise collateral. An intangible is located wherever a
21
Joint Forum on Financial Conglomerates, Supervision of financial conglomerates, BIS (February 1999).
22
The most favourable case may be fairly typical: the subsidiary device used as a form of jurisdictional
compartmentalisation, with relatively simple arm's-length cross-affiliate financial structures.
24
court says it is, raising potential problems of multiple liability and unpredictable perfection of
security interests. Many international treaties have sought - often successfully - to mitigate this
potential confusion. Some of them are discussed below.
3. The need for speed
Today's rapid marketplace conflicts with the slow, deliberate process inherent in most traditional
insolvency processes. Insolvency has two sources of delay: it is often initiated late, when the
value of the firm has greatly eroded, and it usually proceeds slowly once it is initiated. Late
initiation was discussed above, in connection with insolvency governance.23
This subsection
focuses on the speed of insolvency, once initiated.
Reorganisations are inherently slow, because complex negotiation is part of the process.
Liquidations can be fast in principle (especially if the firm is sold as a unit), but are often very
slow in practice. Even if the assets are quickly sold, contingent liabilities can slow the distribution
of assets, especially for firms with a large number of such liabilities on their balance sheets, such
as insurers. The traditional processes, therefore, seem to have a basic time unit of months: months
to initiate, months to organise the parties, and months for subsequent steps. Some processes may
take only a month or even less, eg organising committees or an accelerated liquidation. Others
may take tens of months: negotiations. And the insolvency may be initiated years later than it
should.
The slowness of these processes contrasts with a financial market environment that can change in
hours, a market for corporate information that acts in days or weeks, a product market
environment that be transformed in a few months, and labour markets that often provide rapid
opportunities for employment elsewhere. This gap generates another set of tensions in the
insolvency process.
First, the increased volume and timeliness of information raises the potential for "fire sale" or
adverse pricing, even at an early stage of restructuring or liquidation. When information was
poorer and disseminated more slowly, troubled firms had time to work off large positions. In
recent years, troubled firms, especially large firms or firms with large positions in financial or
non-financial markets, have found that news of their problems travels fast. Markets rapidly
incorporate information about possible asset sales or insolvency into prices, and thus the firm
faces much more unfavourable conditions for restructuring or liquidation when that finally
occurs. This development argues that to maximise the value to shareholders of a troubled firm,
serious consideration of any restructuring needs to take place early, perhaps before the odds of
insolvency become overwhelming.
Second, over the last 20 years, a far more active financial management approach has become
necessary. This has been very noticeable in financial firms, which employ daily mark to market
23
See Section II-B-2.
25
accounting and active risk management for their financial contracting. This practice is in tension
with slow insolvency law. Daily market, credit and liquidity risk management loses much of its
utility if the surviving firm cannot quickly liquidate, adjust or fund its positions. In addition,
skilled personnel need to be retained to manage such portfolios.
The need for active management in financial businesses applies more broadly to all businesses,
because of more intense competition and rapid technological innovation. Firms in protracted
reorganisation tend to lose value, and this tendency is more pronounced in a dynamic economy. A
reorganising firm is probably less able to keep up with the changing markets and shortened
investment cycles in which it must compete. Furthermore, human capital is seldom subject to
reorganisation moratoriums,24
and represents an increasing share of firm value. Much of this
human capital is not so firm-specific that it cannot leave during the period of prolonged
uncertainty of a reorganisation. This is especially true in advanced economy workforces that are
increasingly professionalised and decreasingly bound to their employing organisations. In many
industries, one can question whether the going-concern value preserved by the moratorium is
exceeded by the value destroyed in a prolonged reorganisation process. Furthermore, intangible
assets such as franchise value (not on the balance sheet) tend to decay in value over time if
customers are not carefully monitored and the future of the franchise is not clear. This trend is
especially pronounced in a service economy, in which customers expect an ongoing relationship
with their suppliers, and thus are more sensitive to their suppliers' insolvency risk.
4. The overlap of banking, other financial, and non-financial firms
The convergence of the timing needs of banks and other firms is just part of an increasing overlap
among different types of firms. Banks have traditionally been viewed as special, subject to special
regulatory and (in some jurisdictions) insolvency law procedures. This may be changing to some
extent. Non-bank firms are taking on more bank-like characteristics, such as liquidity- and
market-sensitive trading activities, credit extension, and active risk management.
One principal reason is the well-known blurring of traditional business lines among financial
firms with different licences or charters. Another reason is the growth of sophisticated financial
activities at large non-financial firms, either in the management of their capital, funding and risks,
or in providing financial services for customers, or both. The financial activities of some large
non-financial firms are comparable to those of financial institutions.
More fundamentally, features of financial markets are increasingly being incorporated into nonfinancial
markets, especially with the growth of the internet and widespread availability of
sophisticated data processing. These include more price transparency, the greater speed of
24
There are some isolated exceptions, related to intellectual property and competition issues (eg trade secrets and
covenants not to compete), and trade unions (which in some jurisdictions, can be ordered by courts not to strike).
However, as a good first approximation, these exceptions are just that - exceptional. Most economies permit their
workers to leave their jobs at will, regardless of their employers' financial condition.
26
transactions, easier price and product comparison, better secondary markets for existing goods
and, especially, changes in the mechanisms for payment.
This increasing overlap is important to the extent that a jurisdiction has a special bank insolvency
regime. Most of the distinctive aspects of specialised bank insolvency law are designed with
banks' unique roles in mind. These unique roles are increasingly imitated in the non-financial
sector, especially in the emerging B2B (business-to-business) marketplaces, which are often
modelled on exchanges or auction mechanisms. Thus, the expedited procedures being developed
for banks may be relevant. Banks accommodate liquidity, operate or participate in central clearing
and settlement mechanisms, rapidly intermediate transactions between remote parties, and lack
powerful creditors. Liquidity is impaired - perhaps systemically - if the failure of a major
counterparty ties up the transaction flow. Bank insolvency and payment law therefore seeks to
retain systemic liquidity, through precise finality rules, a preference for very rapid reorganisation
over liquidation, fast liquidation procedures, and the like. Intermediation connotes the
involvement of many entities in a single transaction - which demands sharply defined rights of all
the entities, structured as clean bilateral relations. (In other words, adverse claims on funds are
very difficult to establish.) Central clearing and settlement connotes an extensive reliance on
collateralisation and netting, and thus the rapid enforceability of security interests and netting
arrangements, even in insolvency.
D. Financial contracts at the national level
The differential rates of change between financial markets and insolvency practice have also
created concerns about settlement risk of payments and financial instruments used for liquidity
and risk management, for financial and non-financial counterparties alike. (Settlement risk
extends to payment systems, securities transfers and financial contracts - including derivatives,
securitisations and other contracts which may be collateralised.) Settlement risk in insolvency is
incompatible with the operation of the payment and securities transfer systems, and the uses of
financial instruments to manage risk.
Over the past 15 years or so, these risks have been mitigated through special legal exceptions to
general principles of insolvency law, often called "carve-outs."25
These laws include those
enforcing closeout netting of derivative and foreign exchange contracts, multilateral netting of
payment systems, collateral arrangements securing these contracts, and netting and
collateralisation of financial securities contracts: securities settlement, repos and securities
lending. For example, collateral used for financial contracts or the financial contracts themselves
are insulated from the effects of insolvency arrangements (whether liquidation or reorganisation
measures) either through laws permitting rapid liquidation, closeout and netting of obligations
25
Some jurisdictions have promulgated fewer carve-outs than others, because their law was already relatively friendly
to closeout netting and rapid liquidation of collateral. However, even most of these jurisdictions need special legal
protection for financial specialties, such as clearing and settlement operations.
27
and collateral or title transfer mechanisms. Or, to take another example, securities transfers may
be exempted from ordinary insolvency rules concerning preferential transfer, reducing the risk of
reversing the transfer in insolvency. These special rules resolve either partly or fully large gross
claims and gross liabilities between counterparties immediately upon insolvency, in effect placing
or settling these claims ahead of other secured and unsecured creditors and other stakeholders.
These rules have certainly been successful in their own terms. They have undoubtedly reduced
settlement risk in particular, and systemic risk in general. By way of comparison, the 1974
Herstatt insolvency created worldwide distress, whereas the 1995 Barings insolvency barely
caused a ripple.26
These insolvencies were certainly different in many respects, but decreased
legal settlement risk was certainly a significant one. However, the scope of the carve-out rules is
elastic. There is no widespread consensus on which contracts should be subject to these carveouts.
As markets rapidly evolve, the scope of these provisions is effectively widening. No static
list of carve-outs can accommodate market evolution, and there are no common principles to fill
the gap. Therefore, although most jurisdictions have carve-outs, each jurisdiction has a different
list of carve-outs, and updates its current list at different rates than other jurisdictions. This creates
frictions among jurisdictions, discussed in the section on cross-border insolvency.
It is difficult to evaluate the countervailing forces involved in the rules governing financial
contracts. The argument for an insolvency regime supportive of legally certain netting and rapid
enforcement of collateral is strong. If such regimes did not exist, creditors would avoid heavily
netted contracts or contracts such as repos whose small margins are only justifiable if collateral
enforcement is swift and certain. The result: a thinner market that cannot take advantage of
modern risk transfer technology. The argument is even stronger for payment system netting,
which could not support large payment volumes on little capital without strong netting and
collateral enforcement rules.
However, if resolution regimes are too supportive of netting and collateralisation, the unsecured
creditors of the firm will insist on compensation, in the form of higher interest rates, greater
overdraft privileges or the like. Unsecured creditors who enjoy deposit insurance might not
mitigate their risks, but instead pass it on to the insurer, who would therefore ultimately
underwrite the cost of netting and collateralisation. Excessive netting and collateralisation might
also render the insolvency process largely irrelevant. In the event of actual or impending default,
creditors may merely seek to enforce their security or exercise rights of closeout. In such a
regime, if firm-specific transactions are secured or netted, the institution will cease to exist, to the
detriment of unsecured creditors, and with the loss of any firm-specific value, assuming some
value exists.
26
The Barings insolvency was smooth for at least two reasons. First, most of the Barings derivatives liabilities were
subject to adequate risk management, both that of Barings' counterparties and that of the Singapore exchange,
which served as a central counterparty for many of Barings' exchange-traded liabilities. Second, Barings' bank was
quickly sold as a unit, although the holding company was subjected to a prolonged liquidation.
28
E. Impact on legal certainty, efficiency and equity
The changes described above appear to have changed the relative importance among the three
goals of legal certainty, efficiency and equity.
The gap between the market environment and insolvency processes suggests that the demand for
legal certainty and for efficiency in the insolvency process by creditors and debtors has increased
as market participants manage their risks more closely, as competition and globalisation have
increased, and as assets have taken on a more intangible character that argues for quicker
disposal. At the same time, the increasing liquidity of financial and non-financial assets, the
growth of markets for corporate control, and the ability to fashion contracts to solve incentive
problems in reorganisation or liquidation have increased the possibility of reducing the costs of
legal uncertainty and inefficiency without changing the underlying concepts of equity. Slow and
indeterminate negotiations can now be replaced with faster and surer market-oriented practices. In
effect, the demand for more certain and efficient insolvency resolution has risen, while the supply
(ie legal and transactional technology) for more certain and efficient insolvency has also
expanded.
If uncertainty and inefficiency can be reduced, the overall value of the insolvent estate can
increase. At a minimum in such circumstances, at least some creditors and other stakeholders can
be made better off without making any other stakeholder worse off, and, if desired, basic equity
relationships could be maintained.
The development of financial contracts represents a partial closing of the gap between the market
environment and insolvency processes, for those instruments where the tensions are most acute.
The special exceptions to insolvency law increase legal certainty and efficiency, but at the
expense of reordering priorities within the insolvency regime.
The continued development and refinement of approaches to contract enforceability for financial
contracts suggests that national jurisdictions have found the approaches, on balance, beneficial. A
reasonable inference is that the private benefits in terms of legal certainty and efficiency and the
market-wide benefits of limiting market disruption and sustaining liquidity outweigh any loss to
other creditors or that the losses of other creditors are small. This inference is plausible: financial
contracts do not involve those firm-specific assets whose value would be lost in case of contract
enforcement.
In particular, the gains in legal certainty for financial institutions have broad spillover effects or
external economies. Financial institutions are particularly susceptible to a loss of confidence.
Anticipation of the closure of a financial institution can generate a run on the institution
concerned and become a self-fulfilling prophecy. Regimes which seek to improve the chances of
the institution remaining a going concern can reduce the risk of systemic shocks being transmitted
through the financial sector. But such approaches are likely to be effective only if there is a
reasonable degree of certainty that they would be adopted in a particular case. There may be
29
uncertainty about such adoption even in a single jurisdiction, and the interaction of different
jurisdictions can make less clear what approach is likely to be taken in a particular case.
However, a gap remains between market possibilities and insolvency practice for other assets and
liabilities of the insolvent firm. This gap suggests that additional opportunities exist to increase
the legal certainty and efficiency of insolvency processes. This may involve greater use of market
mechanisms such as auctions of business units in insolvency proceedings. Both the FDIC in the
early 1990s and the bankruptcy trustee in the current Enron case used such methods. These
present the possibility of essentially equity-neutral gains in efficiency and legal uncertainty.
More refined questions about the efficiency consequences of insolvency regimes lie somewhat
outside the scope of this paper, but deserve at least some discussion. Insolvency regimes have
incentive effects on the behaviour of firms and their counterparties. Some incentives enhance
efficiency; others do not. This has been discussed above, in connection with the initiation of
insolvency and subsequent insolvency processes27
but also applies to other insolvency rules. For
example, priority rules that shift insolvency risk from unsophisticated parties (eg employees or
depositors) to financial creditors may enhance the efficacy of monitoring by the more
sophisticated creditors.28
(Such rules would also increase the cost of credit.) Or, to pick another
example, legally certain and rapidly enforceable security interests decrease a secured creditor's
interest in reorganising a firm.
III. Coordination of national insolvency processes:
cross-border aspects of insolvency
Cross-border insolvency has received a tremendous amount of attention during the last decade,
with genuine progress towards international law reform. This progress has been on two fronts:
coordination of national insolvency proceedings and the problems unique to the insolvency of
international financial firms: liquidity management, settlement risk and ensuing systemic risk.
This section begins with an analysis of the problems of coordinating cross-border insolvency.
Problems in coordination apply both to insolvency processes and to substantive insolvency rules,
notably the carve-outs discussed above. It then applies this analysis to the schemes actually used
to coordinate cross-border insolvency, so-called "full universality," "modified universality," and
"territoriality." It describes how cross-border coordination issues are somewhat different for
banks than for general business firms. The section concludes with a brief review of current law
reform movements. Appendix A chronicles this progress in detail, especially with regard to bank
insolvency.
27
See Section II-B-2 above.
28
This argument is a hardy perennial in discussions of deposit insurance. See for example Helen Garten, "Banking on
the market: relying on depositors to control bank risks", 4 Yale Journal of Regulation 129 (1986).
30
The conclusions of this section are complex, and deserve a brief summary here. First, it is
difficult to coordinate international insolvencies, and the appropriate coordination schemes for
banks might not be those of general business firms. Second, the predominant (if not necessarily
optimal) coordination model today is based on comity: voluntary cooperation among insolvency
courts and Officials. Outside the European Union, there is little in the way of treaties that
coordinate cross-border insolvency.29
Third, although the coordination of cross-border insolvency
procedures may be weak, there has been considerable progress in harmonising cross-border
insolvency rules, notably those related to the conflict of laws, financial contract netting and rapid
liquidation of financial collateral.
A. The problems of insolvency coordination in an international context
Legal certainty is as valuable in a cross-border world as it is within a single jurisdiction.
However, legal certainty is more difficult to attain across borders. Which law will govern the
insolvency proceeding or will multiple laws apply? Which law will determine whether a security
interest is perfected or a netting contract is valid? These questions are particularly important for
financial institutions. Within a particular jurisdiction, most financial contracts enjoy considerable
legal certainty. Will this certainty be retained in an international context, especially in
insolvency?
Uncertainty over financial contracting has two salient dimensions in an international context. One
of them concerns the choice of an insolvency regime and the potential for forum shopping. The
other concerns the frictions between different regimes, even after the relevant forum (or forums)
have been selected.
1. Choice of an insolvency jurisdiction
Different jurisdictions have different distributional consequences in insolvency. Therefore, those
who have the ability to invoke an insolvency regime have an incentive to invoke the jurisdiction
most favourable to them: the "forum shopping" problem. If only one party were free to select a
jurisdiction, forum shopping would not necessarily create uncertainty: it would merely mean that
insolvency regime would be the most favourable one, from the party's perspective. Since the
relevant party is usually the debtor, we would expect to see pro-debtor regimes flourish, to the
extent the debtor has a choice of jurisdiction.
However, the ability to initiate insolvency could rest with different parties, such as creditors,
directors or shareholders in the institution concerned, or a relevant supervisory or regulatory
authority or other administrative body. This creates genuine legal uncertainty, as well as the
distortion mentioned above. Who will be the first to file? Another layer of uncertainty may exist
where a court or other judicial body has discretion to decide whether the application proceeds or
29
The Bustamente treaty, in force in much of South America, is perhaps the best-known example.
31
is successful, or perhaps to discontinue one type of procedure and substitute another on the
application of other interested parties.30
Finally, there is the prospect of two competing regimes,
initiated by two parties with separate interests in the insolvency. As a result it may be difficult to
predict whether and which regime will be triggered in any particular circumstances.
One can view forum shopping from two perspectives: ex ante and ex post.
Ex ante forum shopping is simple in concept. Business tends to seek the lowest-cost environment,
for labour, capital and other inputs. This applies as much to the legal environment, as any other. If
a firm could credibly associate itself with a jurisdiction having capital-friendly insolvency law, its
cost of capital would decrease. This could, in principle, initiate a wholesome competition among
legal systems for more capital-friendly insolvency law. In other words, there is not always a race
to the bottom. However, this kind of forum shopping would undercut insolvency policies that
give greater weight to labour and communities.
Although the prospect of ex ante insolvency forum shopping is intriguing, it is also at present not
a major factor in choosing business location.31
It is commonly believed that business, tax and
corporate control reasons dominate the choice of principal jurisdiction, for firms that have such a
choice. Most legal systems do not allow a free choice of principal jurisdiction. Differences in tax
regimes in fact are known to drive choice of corporate domicile.32
Under some circumstances,
discussed below, differences in insolvency regimes might have a similar effect.
Ex post forum shopping, where it is possible, is more likely and more problematic. When a firm
becomes distressed, management is likely to seek a jurisdiction that provides the best prospects
for itself: a chance to participate in reorganisation and a minimum possibility of penalty for errors
in management judgment or misconduct. Although ex ante forum shopping may be desirable, it is
more difficult to justify ex post forum shopping.
Ex post forum shopping is not a major problem in current international insolvency law, because
most legal systems discourage the practice. However, it is not at all hypothetical. It is a well-
30
Official discretion is a general problem, not limited to the cross-border context. It is common in ordinary
insolvency law. For example, a liquidator might be able to decide which contracts to perform (in which case a
counterparty will - in the absence of an enforceable contractual provision dealing with termination - still be under
an obligation to perform the contract) or to repudiate (in which case the counterparty will only have a right in
damages). Alternatively, a counterparty may find that its ability to realise security is exercisable only with the
consent of the insolvency Official and/or the consent of the courts.
31
A firm can engage in credible ex ante forum shopping only if the rules are clear and the choice is irrevocable (or at
least difficult to reverse). Forum shopping can lead to a competition among legal systems only if there are few
constraining relationships between the jurisdiction selection rule and the business activities, assets and locations of
the firm. As a general matter, clear jurisdiction selection rules (eg situs of incorporation) are usually revocable, and
often constrained.
32
For example, many firms headquartered and operating in the United States have obtained Bermuda charters for tax
reasons alone, eg Global Crossing, Tyco and Ingersoll-Rand.
32
known phenomenon in US general insolvency law,33
and explains the popularity of pro-creditor
bankruptcy courts in Delaware and New York. It is a potential problem whenever a firm can
freely change its place of incorporation, incorporation is a sufficient contact for insolvency
jurisdiction, and other jurisdictions are obligated to respect this contact. These conditions may be
uncommon now, but may become more prevalent with increasing globalisation. We shall discuss
the further problems of ex post forum shopping below.
2. Other cross-border dimensions of insolvency procedure
The choice of an insolvency jurisdiction is not the only important cross-border issue regarding
insolvency processes. One closely related issue is the choice of jurisdiction for a firm's assets and
liabilities, which can often have significant consequences in insolvency. (For example, the law
governing setoffs and preferential transfers might be the law of the asset or liability, rather than
the law of the insolvency forum.) The location of most liabilities is a matter of contract,
especially for global firms. Many financial assets are some other firm's liabilities, and therefore
also have a manipulable situs.
As another example, the rules regarding the treatment of subsidiaries within a group of insolvent
companies have not been harmonised in all countries. Some jurisdictions apply the law of the
main place of establishment to the insolvency of all entities within a group of companies affected
by the insolvency. However, most jurisdictions do not consolidate in this fashion, and view the
appropriate law separately for each entity within the group. The consolidation of subsidiaries is a
particularly significant issue for financial firms, and is discussed in more detail below.
One fundamental question at the outset of major insolvency cases is whether a rescue attempt
should be undertaken through reorganisation. Different countries may have different types of
reorganisation measures, as well as different rules and criteria for the situations where a
reorganisation of the insolvent entity may be attempted. In a cross-border insolvency case, these
can create some obvious problems of inconsistencies. Proceedings taking place concurrently in
different jurisdictions with different insolvency laws, where one party's proceedings are not
recognised by the other, could lead to contrary conclusions on whether reorganisation or
liquidation should be conducted.
3. Enforceability of contracts: netting and realisation of collateral
As discussed above, the enforcement of financial contracts in insolvency proceedings tends to be
regulated by "carve-out" statutes. As a general matter, these statutes assure the enforceability of
certain kinds of netting agreements notwithstanding insolvency, and assure that certain collateral
arrangements can be enforced without hindrance from the insolvency. Unsurprisingly, no two
jurisdictions' carve-outs are identical. This can complicate cross-border insolvencies
33
The United States permits a bankruptcy filing to be made in any state in which the insolvent party has an insolvent
affiliate (28 U.S.C. § 1408(2)). For larger firms, this permits the possibility of substantial forum shopping.
33
tremendously. Some examples of different jurisdictional approaches to carve-outs are drawn from
the Contact Group's questionnaire:
* The different treatment of "absolute transfer of title mechanism" under different
laws, for instance under English law and under Italian law. While the majority of
market participants using the ISDA credit support documentation have preferred
title transfer arrangements subject to English law, under Italian law such
arrangements are deemed vulnerable to re-characterisation as disguised security
(these schemes could be recognised as valid and enforceable only if falling in the
category of "repos").
* The different treatment of top-up collateral. While in some jurisdictions top-up
collateral delivered in a specified period prior to the commencement of insolvency
may be declared null and void (Japan) or ineffective (Italy), in other jurisdictions
there is a special protective regime, in the sense that top-up collateral provided for
in the context of an agreement relating to qualified financial contracts is valid and
enforceable (United States).
* Differences in the valuation of obligations and currency conversions. While in some
jurisdictions (Italy) there are mandatory rules for this valuation (with reference to
replacement costs, using market values at the date of the declaration of insolvency),
in other jurisdictions (Japan, Netherlands, United Kingdom, United States) as a
general rule the non-defaulting party has the freedom to determine the valuation
method (although in some jurisdictions, ie United Kingdom, the liquidator may
have the ability to challenge such valuation if it is not based on "reasonable
grounds").
In a cross-border insolvency, it is not enough to know if a particular carve-out rule is enforceable
in one particular jurisdiction. What if the rule is enforceable according to the carve-out rules of
the jurisdiction whose law governs the netting or collateral agreement, but not in the jurisdiction
(however defined) of the insolvent party? (What are the choice-of-law rules for determining
which law governs the agreement?) What if the insolvency proceedings are split among several
jurisdictions, a distinct possibility discussed below? These questions may be arcane, but they are
very significant. As discussed above, the carve-out rules are widely believed to reduce systemic
risk. If they do not work in a cross-border context, they may be less effective than assumed.
4. Perfection of collateral
One important function of the carve-out rules is to permit parties to rapidly realise or enforce
financial collateral in insolvency. The enforceability of collateral depends on whether the security
interest has been "perfected" in the "appropriate" jurisdiction. (Collateral is generally perfected
through publicising the security interest in a matter recognised by law, eg registration,
34
notification, control, physical possession, or the like.) The real cross-border problem is that of
uniquely determining the appropriate jurisdiction in which the perfection is to occur. The problem
is difficult. For the system to work, all jurisdictions must agree on the appropriate conflict-of-law
rules to determine the means of perfection, and the conflict-of-law rules should lead to a unique
jurisdiction's law. This very difficult problem in international cooperation has been universally
solved for only one class of collateral: real estate. Jurisdictions are territorially based, and real
estate is both territorial and immobile. All other classes of collateral are more difficult. Financial
collateral (eg securities, receivables, derivative contracts, bank deposits, and the like) is
intangible, and one might therefore expect that a solution to this problem is impossible. However,
financial collateral has been the subject of considerable recent international law harmonisation
efforts. These efforts promise to bear fruit over the next decade, and allow most jurisdictions to
harmonise their rules on perfection of financial collateral.
The recent European Directive of the European Parliament and of the Council on financial
collateral arrangements34
("EU Collateral Directive") is intended to determine the appropriate
jurisdiction of perfection for cross-border use of financial collateral between EU member states.
The Directive defines financial collateral to include transferable securities and cash deposits. (It
also includes rules encouraging the rapid realisation of collateral and discouraging clumsy
formalities in establishing a security interest.)
UNCITRAL has developed a Convention on the Assignment of Receivables in International
Trade - which includes assignments used in factoring, forfaiting, securitisation, project financing
and refinancing. The primary purpose of this Convention is to determine the appropriate
jurisdiction for perfection of receivables. Although not yet in force, this Convention has been
successful enough for UNCITRAL to be tasked with a broader study of secured credit law, with
the purpose of identifying possible solutions, ranging from a "model law" to a "convention".
The Hague Conference on Private International Law is working on a draft Convention on the law
applicable to dispositions of securities held through intermediaries, with the aim of implementing
the "PRIMA approach". According to the PRIMA principle, the rights of a holder of such
securities provided as collateral will be governed by the law of the country of the relevant
intermediary which maintains the securities account.
The scope of the project of the Hague Conference is not to harmonise the substantive laws
relating to securities (this project is being considered by UNIDROIT). It is limited to determining
the relevant law governing perfection (and most other property rights) for securities held through
intermediaries. The Hague Conference project has been anticipated by other legal systems (eg the
US UCC Article 8 and the EU Settlement Finality Directive), but the Hague Conference project
has the prospect of worldwide reach.
34
Directive 2002/47/EC of 6 June 2002, OJ L 168/43.
35
5. Implications for legal certainty, efficiency and equity
The problems described in this section suggest that in an international context, legal uncertainty
can increase dramatically. The choice of insolvency regime can become uncertain, and sometimes
even manipulable. Conflicting laws and an absence of mutual recognition can make the
enforceability of contracts uncertain. The appropriate jurisdiction for perfecting a security interest
in collateral can become nearly impossible to determine.
This uncertainty contributes to inefficiency. Uncertainty creates incentives for debtors and
creditors to behave in a manner that protects their respective interest, but may undermine the
ability of the insolvency process to maximise the value of the insolvency estate and may create
spillover effects on third parties. In addition, the complexity of administration of multijurisdiction
insolvencies, including the costs of negotiation and litigation, can generate substantial
direct costs and delays. Finally, the ex ante costs of planning transactions to accommodate
insolvency also increase: legal opinion letters become more profuse and less reliable.
As was discussed in the previous section, legal uncertainty and inefficiency can be reduced. But
unlike the closed system of a national insolvency regime, these improvements cannot necessarily
be made assuming that the distribution of the proceeds of the insolvency estate can be held
constant. Since concepts of equity differ across national borders, a result of the democratic
process in each country or region, the choice of regime(s) will affect the distribution of outcomes.
To some extent, providing a clear understanding on how insolvencies across jurisdictions will be
handled can mitigate the problem, in that market participants can gauge the probabilities of
possible outcomes and incorporate them into their ex ante decision-making. It is also possible
that more analysis can identify improvements in the coordination mechanism that are relatively
equity-neutral.
Efficiency has another dimension - spillover effects to otherwise uninvolved third parties,
especially in the form of systemic risk. There is no real disagreement with the objective of
avoiding systemic risk and financial instability, particularly in view of the magnitude of the
problems that could materialise in financial markets. Hence, the need for efficient and effective
insolvency regimes has gained recognition as a means of contributing to the maintenance of
financial stability. For instance, insolvency appears as an important item among the 12 standards
for stable financial systems recently introduced by the Financial Stability Forum.35
35
Final Report of the Follow-up Group on incentives to foster implementation of standards, 11 September 2001, at
p 8 on www.fsforum.org. Other publications include UNCITRAL draft legislative guide on insolvency law, 2001;
Group of Thirty, Reducing the risks of international insolvency, Washington DC, 2000; Orderly and effective
insolvency procedures, key issues, IMF 1999; Principles and guidelines for effective insolvency and creditor rights
systems, World Bank, April 2001.
36
B. Approaches to the coordination of insolvency proceedings
The coordination of insolvency proceedings is the most difficult dimension of general crossborder
insolvency law. It has implications for the problems discussed above: financial
contracting, forum shopping, and the like. Coordination is difficult for several reasons.
Substantive insolvency law varies considerably across jurisdictions, and international
coordination of insolvency law may endanger valued local considerations of public policy. In
contrast, the consensus on cross-border netting and collateralisation is much stronger, and
coordination of choice-of-law rules concerning netting and collateral places far less strain on
national autonomy.
Coordination may be viewed as a kind of continuum: from a completely centralised global
proceeding at one extreme, to a set of completely independent proceedings along territorial lines,
at the other. We here seek to assess the cross-border issue for banks, in the context of general
cross-border insolvency law.
In both general and bank-specific insolvency law, three cross-border models are conceivable. The
first two models go by several names: we shall call them the full and modified universal model
of insolvency. Both full universal and modified universal insolvency assume a cross-border
division of labour, with most "ancillary" jurisdictions acting in aid of a central insolvency
process. In a full universal insolvency model, the "main" jurisdiction's insolvency law will
govern the insolvency of the entire firm. Typical ancillary roles include collection of assets and
imposition of a moratorium, both aiding the central insolvency proceeding, which either
reorganises the liabilities or liquidates and distributes assets. These ancillary roles invoke the
power of the local court, and some of the law of the ancillary jurisdiction. However, all of the
specialised insolvency law - processes, preferences, priorities and the like - is that of the main
jurisdiction.36
As an example of this division of labour, an ancillary court may feel obliged to turn
over all local property to the main jurisdiction, but may use its own law to define whether a
property right exists.
Full universal insolvency implies a surrender of sovereignty: predefined rules which dictate the
respective roles of jurisdictions. Modified universal insolvency retains insolvency sovereignty,
but encourages a cooperative approach. Each jurisdiction decides whether it should take a central
or ancillary role. Ancillary jurisdictions must decide which non-ancillary foreign jurisdiction
should be treated as the "primary" jurisdiction, and which other non-ancillary foreign
jurisdictions have the "secondary" role. In the modified universal insolvency model, jurisdictions
are supposed to show some deference to the choices of other jurisdictions and to the needs of the
system: the so-called "comity of law."
36
Cf Council Regulation (EC) No. 1346/2000 of 29 May 2000 on insolvency proceedings, OJ L. 160, 30.6.2000,
pp 1-18, Article 4(2). The Insolvency Regulation's list includes: the assets associated with the insolvent, debtor and
liquidator powers in the proceeding, setoff, permissible claims, creditors' remedies and priorities.
37
These two models are sharply distinguished from the "territorial" insolvency model, which
expects no international cooperation. In the territorial insolvency model, each jurisdiction acts
independently of the others. The insolvent firm is simply liquidated or reorganised along
jurisdictional lines.
Full universality - which requires an international treaty - is often considered attractive.37
However, the surrender of sovereignty implicit in full universality is difficult to achieve, and
modified universality is often viewed as a more attainable norm. The territorial insolvency model
is commonly viewed as an outmoded system modelled after pre-war notions of sovereignty and
globalisation.38
However, as we shall see, it has its merits.
The following discussion describes the three models in operation. However, its principal goal is
to show that the issues are different for banks. Some problems of each of these models are muted
for banks; and some of the problems are accentuated.
1. Full universality
The notion of a coordinated worldwide insolvency proceeding is quite attractive, especially for
multinational firms. Proponents of universal insolvency admit that full universality requires a
very high degree of international legal cooperation, but otherwise believe that universal
proceedings are a desirable goal, worth the transition costs.
A successful implementation of full universality is the new legal regime on financial insolvency
adopted by the European Union, applicable to all of its member states, except for Denmark.
However, the EU rules clearly represent a special case, not comparable with other attempts at
international rule-making, given that this new European cross-border insolvency regime was
adopted within the existing EU legal and institutional framework. Even with the advantages of the
EU framework and the strong political commitment to general European integration, this
initiative was under consideration for over a decade before its adoption. Outside the European
Union, fully universalistic international insolvency treaties have been very rare.39
It is
extraordinarily difficult for one country to legislatively surrender its sovereignty to another, even
in matters of private law. (Courts have a somewhat easier time surrendering sovereignty on a
case-by-case basis: the so-called "comity of law," discussed below.)
In the end, these efforts led to the adoption of three legal acts with respect to the insolvency of
different categories of legal entities. Non-financial firms are covered by the one of these acts: the
37
Professor LoPucki is a rare dissenting voice. See Lynn M. LoPucki, "The case for cooperative territoriality in
international insolvency", 98 Michigan Law Review 2216 (2000).
38
The universality and territoriality principles are distinct from the dichotomy between the single entity and separate
entity doctrines of bank insolvency law. The single entity doctrine is universalistic. However, the separate entity
doctrine is somewhat different than classic territoriality, in that the local insolvency may assert jurisdiction over
some assets, generally those booked to the local branch.
39
See footnote 29 above and Appendix A, Part 2.3.1.
38
so-called "Insolvency Regulation." Financial firms are subject to the other two acts, which take
the form of EU directives, one for insurance companies and the other for credit institutions.40
These legislative acts will coordinate national insolvency proceedings, but do not otherwise
harmonise them.
The insurance and credit institution directives prescribe a form of full universality. The
administrative or judicial authorities of the "home member state" (the member state where a firm
has its head office) are the only ones empowered to implement reorganisation measures or
winding-up proceedings concerning a firm, including branches established in other member
states. No secondary proceedings may be started by the other member states. In contrast, the more
general Insolvency Regulation permits a secondary proceeding under certain circumstances. The
Insolvency Regulation therefore may be considered a form of modified universality, albeit one of
the more closely coordinated forms.
Although the major objection to full universality is usually impracticability, the argument for
universality has its weaknesses, even if the necessary international cooperation can be obtained.
Some of these weaknesses are heightened in bank insolvency; others are attenuated.
Forum shopping is a potential problem inherent in universal insolvency proceedings. As
discussed above, the problem is largely hypothetical at present (outside the United States), but
might become prominent with increased globalisation and a decreased ability to associate a firm
with a unique jurisdiction. In a globalised world, it is difficult to design rules that locate firms in a
clean and natural fashion. The principal jurisdiction of many major international firms is at least
mutable: consider Royal Dutch Shell, DaimlerChrysler, HSBC, AIG or Vivendi.
Implications for banks. Because of accepted international supervisory norms, banks are even less
susceptible to ex post forum shopping than most non-financial firms. The comprehensive
consolidated supervisor of a bank (or a banking group) is generally quite clear and subject to
licensing, and cannot be changed by a quick filing procedure.
Priority inflation is another possible problem associated with universal insolvency models. It
occurs when local lawmakers seek to safeguard the interests of their local constituencies. This is
possible in a universal regime because the bankruptcy estate is global, but the priority scheme can
40
Council Regulation (EC) No. 1346/2000 of 29 May 2000 on insolvency proceedings (the EU Insolvency
Regulation), OJ L 160, 30.6.2000, pp 1-18 (the EU Insolvency Regulation is effective and directly applicable in all
EU member states (except for Denmark) as of 31 May 2002; Directive (EC) No. 2001/17 of the European
Parliament and Council of 19 March 2001 on the reorganisation and winding-up of insurance undertakings (the EU
Winding-up Directive for insurance undertakings), OJ L 110, 20.4.2001, pp 28-39 (member states are required to
implement the Winding-up Directive for insurance undertakings into their respective national legal regimes by
20 April 2003);. Directive (EC) No. 2001/24 of the European Parliament and of the Council of 4 April 2001 on the
reorganisation and winding-up of credit institutions (the EU Winding-up Directive for credit institutions), OJ L
125/5.5.2001, pp 15-23 (the time frame for member states to implement the Winding-up Directive for credit
institutions is by 5 May 2004).
39
disproportionately benefit local creditors, eg tax authorities and employees. The losers are
unsecured non-priority creditors, who may be less local than the priority creditors.
There is less risk of priority inflation in the territorial insolvency model, because the total asset
and liability pools are smaller than in the universal model, but the local creditor pool is likely to
be about the same size. In other words, a territorial model cannot as easily externalise the benefits
of local priority creditors onto foreign non-priority creditors. Instead, the costs of a priority
regime must be borne more heavily by local non-priority creditors. (As discussed below, the
modified universality model is also resistant to exorbitant priority inflation.)
Implications for banks. The problems of priority inflation seem no different for banks or nonfinancial
institutions.
Corporate groups remain a complex problem, even in fully universal proceedings, including the
EU insolvency directives. The usual rule in general business insolvency is that each corporation is
liquidated separately, often subject to a separate insolvency law. Most attempts at international
insolvency cooperation attempt to coordinate these separate liquidations, not merge them. The
legal constructs of corporate separateness and affiliate independence are strong.
This anti-consolidation rule causes several classes of problem. First, business operations are
frequently entangled among the individual corporate components of groups. This may not matter
for merger or whole-firm liquidation, but can be very problematic with most other insolvency
proceedings. Piecemeal attempts at reorganisation of a corporate group will tend to lose value.
Second, corporate groups generally have inter-affiliate exposures: often a parental guarantee of
subsidiary liabilities, but sometimes far more complex exposures. These can be difficult enough
in an ordinary proceeding, and different jurisdictions' rules on consolidation or inter-affiliate
subordination can have significant effects in other jurisdictions. While regulatory limits to some
extent control inter-affiliate exposures, the risk management and funding needs of financial firms
tend to make inter-affiliate exposures of increasing importance as a risk management issue.
Finally, the boundaries of corporate groups are frequently vague: consider problems posed by
joint ventures, minority shareholdings, cross-ownerships, and decentralised organisations.
Implications for banks: consolidated comprehensive supervision. The issues of corporate groups
are especially important for large, internationally active financial institutions. These firms often
adopt management and business line structures that cut across the legal entities in the corporate
group. This organisation, plus efforts to centralise the risk management of key global activities,
lead to large and diverse inter-affiliate transactions. The indistinct separation of corporate entities
within a group is consistent with the "comprehensive consolidated supervision" ("CCS") concept
of unified bank supervision. The CCS concept demands a central supervisor responsible for the
entire financial group, no matter where the group entities are incorporated. Local supervision and
entity-based supervision are not inconsistent with CCS, but the central supervisor is responsible
40
for the entire organisation. Local supervisors are discouraged from permitting entry of financial
firms that do not enjoy CCS.
The CCS concept is extremely powerful, because it is a jurisdictional and organisational focal
point that could facilitate universal insolvency proceedings applicable to corporate groups. CCS is
not susceptible to forum shopping, because it is indisputable ex ante and (relatively) immutable
ex post. It permits an organic connection between supervision and insolvency on an international
scale, accommodating the business overlap across entity and jurisdictional lines.
CCS is not (yet) an element of financial insolvency law. Financial groups are subject to the same
deconsolidated insolvency procedures as non-financial business groups. Agreement to CCS is not
the kind of surrender of sovereignty necessary to support fully universalistic bank insolvency.
Furthermore, CCS permits considerable diversity of practice, and accommodates a diversity of
objectives between home and host country supervisors, and insolvency law. In particular, in some
jurisdictions, including the United States, the host country supervisor has statutory responsibility
to protect local depositors in the event of a foreign bank failure. Thus, the law provides scope for
an insolvency proceeding in the host country's courts. Furthermore, advanced jurisdictions differ
on basic questions, such as whether a special bank insolvency regime is advisable
Nonetheless, matters may change with time. The CCS concept is only about a decade old
(although it has antecedents in the bank insolvencies of the mid-1970s and the 1970s principle on
consolidation set out by the Basel Committee on Banking Supervision and the Basel Concordat),
and has become increasingly accepted. The role of CCS in crisis management is increasingly
appreciated, and crisis management is a supervisory kin to insolvency. Full universality might be
an ideal in a world of sovereigns. If this ideal ever becomes a reality, CCS might play a major
role, especially for financial firms.
It is worth noting that banks generally operate through branches, unlike most other international
firms, which operate through local subsidiaries. The problem of corporate groups therefore does
not apply to most international banks, viewed as standalone entities. (The separate entity doctrine,
however, may require separate proceedings for individual branches.) However, as discussed
above, larger banks commonly belong to a group and engage in substantial transactions with their
non-bank affiliates, so the problems of corporate groups are applicable to banking organisations,
even disregarding the separate entity doctrine.
2. Modified universality
In some cases, a modified universal insolvency is indistinguishable from a fully universal
insolvency. Sometimes, one jurisdiction is clearly the appropriate one for the main insolvency
proceeding. If the other jurisdictions recognise this, and their courts are sufficiently empowered,
they can agree to serve as ancillary jurisdictions to the main proceeding. Thus, decentralised
cooperative consent amongst courts in different jurisdictions ("comity") can play the same role as
binding international law.
41
Modified universality is becoming the norm, even though full universality is still the ideal. The
norm is embodied in the UNCITRAL Model Law on Cross-Border Insolvency, promulgated in
1997. This model law has already been adopted by a few jurisdictions including Mexico, Eritrea,
Montenegro and South Africa, and is being considered in others, such as the United Kingdom and
the United States. Because it does not sacrifice the sovereignty of insolvency courts, it is probably
far easier to implement than full universality.41
(To the extent that a court decides on an
independent proceeding, the model law may be considered a co-operative form of territoriality.)
But although modified universality is similar to full universality, it contains some different
strengths and weaknesses. Modified universality contains some safety valves that weaken the
promises - and mitigate the weaknesses - of full universality.
* Priority inflation is less of a problem in the modified universal model. A jurisdiction
is not likely to assume the ancillary role if the main jurisdiction's priorities are
exorbitant.42
The UNCITRAL Model Law does not compel a jurisdiction to
recognise another.
* Forum shopping is more likely to be a different kind of problem in modified
universality. The debtor - and creditors - have the option of shopping for multiple
forums. However, it is less likely that the debtor can game the rules to select a
totally inappropriate forum, because such a forum is unlikely to be recognised by
other jurisdictions as the forum of the main proceeding.
* The same issues of corporate groups apply to modified universality.
* Reorganisation is more difficult in modified universality, because of the prospect of
defecting jurisdictions. (This is less relevant to financial institutions, whose
supervisor-led or -encouraged reorganisations usually occur without formal
insolvency.)
* A comity-based system such as modified territoriality may become less stable as the
largest firms regulated by the system grow in size and complexity, especially
because a single uncooperative legal system can impose heavy costs on all others, if
it contains enough assets. Fortunately, however, the legal systems of most major
national economies appreciate the need for international cooperation.
41
It should be noted that the EU Insolvency Regulation also allows for secondary proceedings.
42
See In re Treco, 240 F.3d 148 (2d Cir. 2001). In this case, the US court refused to assume an ancillary role with
respect to a security interest, because the overseas proceeding accorded the administrator's expenses priority over
security interests. Since the administrator had collected around $10 million, with all but $1.75 million disbursed to
it in expenses, it appeared as if the US security interest would be completely worthless, if turned over to the
administrator.
42
In summary, modified universality is no stronger than comity, which has worked well to date.
Comity requires that a court favour systemic concerns over local ones. This implies both support
of global concerns and some degree of insensitivity to local concerns. If local concerns are deeply
felt - and bank regulation often raises strong local emotions - a court may be less likely to grant
comity to another court.
3. Territoriality
Territoriality places fewer strains on international cooperation than full or even modified
universality. It does not preclude cooperation among jurisdictions, but tends to encourage at best
a modest degree of cooperation: information-sharing, ad hoc cooperation, and the like. It does not
permit the more pernicious forms of forum shopping, but - as discussed above - may encourage
some useful forms of jurisdictional competition for financial assets.
Inefficiency is probably characteristic of territoriality, as compared to universality. The territorial
insolvency model involves multiple proceedings and races for assets by both Officials and
creditors. (Intangible assets present the spectre of multiple liability, although this may soon be
mitigated by the Hague Conference and UNCITRAL projects discussed above.) Some assets may
be subject to no insolvency proceeding at all, eg if they are located in a territorial jurisdiction
whose requirements for insolvency are not met. Inter-affiliate obligations - difficult enough in the
universal model - become even more difficult to sort in the territorial model. Such an inefficient
system is probably unfair, because stronger creditors tend to do better in a legal free-for-all.
Reorganisations of unsupervised firms are very difficult in the territorial model, unless the
relevant jurisdictions are extremely cooperative. Reorganisations imply that the insolvent will be
run as a going concern - something very difficult for a multinational business partitioned by an
insolvency proceeding into territorial units. This is mitigated, to some extent, by non-financial
firms' common practice of compartmentalising a multinational firm by subsidiaries that mostly
follow jurisdictional lines. The reorganisation problem is potentially problematic for regulated
financial services, since they are more likely to use branches than affiliated organisations, and are
more likely to conduct the same business line using several affiliates. However, in banking and
insurance, most reorganisations are conducted or encouraged informally by home-country
supervisors, without use of the insolvency apparatus. Such reorganisations are not hindered by the
territorial model, but require supervisory cooperation.
Forum shopping, at least in the forms seen in universal models, is not characteristic of
territoriality. Territorial insolvencies occur where the assets are, not where the parties choose to
file. But another kind of forum shopping is possible in territorial models. Most financial assets and
all liabilities - are intangibles: loans, securities, deposits and derivatives. At least with respect
to securities and deposits (and possibly derivatives), a financial firm can often choose the location
of its assets with a stroke of the pen, without substantially affecting its underlying business. A
financial firm can also choose the location of its liabilities, at least for wholesale placements.
43
Therefore, a kind of ex ante forum shopping is possible in the territorial model. As discussed
above, ex ante forum shopping may be desirable, under some circumstances.
Supervisory cooperation and competition are strongly affected by territoriality. In the territorial
model, supervisors will tend to look out for their local balance sheets and thus have ex ante
incentives to monitor the financial institution closely. In other words, territoriality reduces the
externalities inherent in home-host divisions of supervisory responsibilities. However, vigorous
host-country supervision of financial firms may face the same problems as in the universal model:
business line organisations that cut across jurisdictional lines, centralised decision-making and
control functions, and "global" books that include transactions originated in other jurisdictions. In
other words, territoriality may provide good supervisory incentives for the host country, but the
incentives are more compatible with the supervision of less complex firms.
The flip side of good supervisory incentives for the host country is a reduced chance of
supervisory cooperation on the eve of insolvency, and during the insolvency process. In the
territorial model, home and host supervisors will become unwilling to share financial information
on weak firms once they suspect insolvency is imminent. This may make reorganisation more
difficult, and any insolvency less efficient.
Cross-border carve-out rules can be conceptually difficult to reconcile with a pure territorial
model. Why should an Official recognise the netting of a foreign debt against a domestic debt,
when the Official is committed to ignoring foreign assets and liabilities? Or why should an
Official recognise a domestic asset as securing a foreign debt that is outside the scope of the
administration in any case? Few territorial models are pure, and some territorial models explicitly
recognise cross-border netting.43
Of course, there is nothing inherent to a universal or modified
universal model that ensures that it will enforce cross-border netting and collateralisation
agreements. However, it is safe to say that territorial models are less likely to be receptive to
cross-border netting and collateralisation than universalistic models.
Financial and non-financial institutions are both affected by these problems. Banks rely on netting
and collateralisation in their risk management, which could frequently be cross-border. Nonfinancial
organisations frequently employ cross-affiliate guarantees, which have a cross-border
component.
A coordinated international workout or insolvency of financial institutions is especially important
to manage systemic risk. This places some real strain on territoriality. Without an effective
coordinated response, a run on the institution is conceivable.44
Even secured creditors and market
43
NY Banking Laws § 618-a.
44
The presence of deposit insurance will limit such behaviour but it may not prevent runs both because depositors
will want to avoid the inconvenience of their funds being temporarily frozen until the insurance is paid and because
there may be limits on the extent of cover (eg there may be an upper limit on the amount of insured deposits, or
44
counterparties may join the run if they doubt their position under at least some jurisdictions.45
However, territoriality - although it impairs supervisory cooperation - tends to be less complex,
and thus less legally uncertain, than modified universality or a poorly coordinated full
universality.
C. Current initiatives in cross-border insolvency
This subsection summarises the recent history of cross-border insolvency reform efforts. It does
not discuss the substance of these efforts - that is the role of Appendix A. Instead, it enumerates
the efforts, places them in a chronology, and discusses the various players behind them.
Only some of these initiatives are statutory, ie model laws or conventions. Statutes are not the
only mechanism of legal reform, and are generally promulgated only after a consensus on law
reform has been developed by organisations without legislative power. These organisations may
include public sector entities without legislative power (eg national central banks or various treaty
organisations), or various kinds of private sector organisations. Although private sector
organisations inherently have no legislative power, the trade codes and model agreements that
they promulgate are often drivers for law reform and themselves quasi-legislative in effect, eg the
ISDA model derivatives agreements or the ICC's efforts in letters of credit. (Cases of judicial
decision-making can also be a significant driver of law reform, although we do not discuss them
here.46
)
1. Legislative initiatives
Most of the international legislative initiatives have already been discussed, and merely need to
be summarised here. They can most compactly be presented in tabular form. Particularly
noteworthy are the recent dates of all the initiatives. Table 3 clearly shows that cross-border
insolvency reform is new. The legislation is the fruit of efforts that began in the early 1990s, often
originating from the private or financial policy sectors. These efforts had been preceded, in many
jurisdictions, by national law reform efforts with the same scope: a trend to modified universality,
and various netting and collateral rules that minimised settlement risk.
foreign currency or overseas branch deposits may not be covered). Furthermore, depositors may be uncertain
whether and how the deposit protection scheme applies.
45
For market counterparties relying on rights of setoff or closeout netting, these may well be protected or enforceable
notwithstanding liquidation proceedings or reorganisation measures, but there may in practice be confusion as to
whether the relevant contractual right to trigger such setoff or closeout netting has indeed arisen (which may
depend on whether any formal steps for the commencement of the resolution mechanism concerned have indeed
been started). Such uncertainty will be more likely if the insolvency proceedings are in a "foreign" jurisdiction, and
therefore of unfamiliar nature or extent.
46
For example, the cooperation between English and US courts in the insolvency of the Maxwell companies was a
significant part of the backdrop for the UNCITRAL Model Law on Cross-Border Insolvency.
45
Table 3
Leg. body Name Date Subject
UNCITRAL
Model Law on Cross-Border
Insolvency
1997 Modified universality
EU
Settlement Directive 1998 Netting payments/securities transfers and
financial collateral enforceability
UNCITRAL Receivables Convention 1999 Conflict of law - collateral perfection
EU Insolvency Regulation 2000 Modified universality
EU Winding Up - Banks 2000 Full universality
EU Winding Up - Insurance 2000 Full universality
Hague
Conference
Draft Convention on law
applicable to securities held
through indirect holding systems
-- Conflict of law - collateral perfection
EU
Collateral Directive 2002 Financial collateral enforceability and
conflict of law
UNCITRAL
Draft Legislative Guide on
Insolvency Law
-- Best practices - municipal insolvency law
2. Private sector initiatives
Private sector trade and professional associations have been very active in the international
insolvency law revision process. Some of their products are quasi-legislative, such as the
International Bar Association's Cross-Border Insolvency Concordat. These organisations also
hold symposia, some of which are published, sometimes with seminal results.47
Finally, these
organisations are very active as NGOs in the international legislative process.
These organisations include (in alphabetical order): EMTA (Trade Association for the Emerging
Markets), IBA (International Bar Association), INSOL International (International Federation of
Insolvency Professionals), and ISDA (International Swaps and Derivatives Association).
Most of these groups represent the viewpoint of practitioners: chiefly insolvency lawyers,
accountants and judges. This has many benefits: practitioners are more knowledgeable about the
47
For example, Randall Guynn's 1996 article for the International Bar Association, Modernizing securities
ownership, transfer and pledging laws, started the process that led to the Hague Conference project on securities
held through intermediaries.
46
operational details of insolvency than any other group interested in insolvency. Furthermore, the
practitioners represent all the players in an insolvency: officials, creditors and others. However,
practitioners tend to view the world from an ex post prism, whilst financial institutions and their
attorneys plan transactions ex ante.
3. International financial policymakers
As a core business practice, insolvency engages widespread attention. It is important to many
beyond the group of lawyers, accountants and other practitioners who draft the rules and
administer the processes. Insolvency is particularly important to international financial
policymakers. This report, therefore, is part of a tradition of insolvency law reform that dates
back at least to the BCCI insolvency of 1991. (The Herstatt insolvency of 1974 was an important
antecedent that helped lead to the netting carve-out legislation of the late 1980s.) Subsequent
milestones have included:48
* The G30's 1998 study group report International insolvencies in the financial
sector.
* The BIS 1999 symposium on international bank insolvencies.
* The current IMF project on bank insolvency.
* The World Bank's Bank Insolvency Initiative.
These efforts consider insolvency primarily from a legal perspective and have had substantial
input from insolvency practitioners.
IV. Conclusions and new directions
The analysis in this report suggests that the legal uncertainty, inefficiency and potential inequity
resulting from the existing legal and institutional underpinnings of insolvency may be
incompatible with important objectives of public policy related to financial stability. Moreover,
the risks involved may be growing as the pace of change in the financial system continues to
outstrip that of the insolvency framework.
This section describes some possibilities for what can be done in the medium and longer term.
Many of these possible paths are complementary, and can be pursued in parallel. They are listed
in very approximate order of priority.
48
Most of the information here comes from the G30 study group report Reducing the risk of international insolvency:
a compendium of work in progress (2000).
47
A. Financial contracts and collateral in national law
Financial contracts are not usually a major problem in general insolvency cases, but along with
collateral loom large in the insolvency of financial institutions. General insolvency law places no
particular stress on rapid liquidation of collateral; financial institution insolvency often does. Both
general and financial insolvency law requires international consensus on the perfection rules for
cross-border collateral.
Many of the problems are cross-border, and will be discussed immediately below. But at least two
questions can profitably be viewed as matters of local law. First, when should national law
recognise set-off? Second, when should it recognise rapid liquidation of collateral? The report
hints at some criteria that are useful for answering both questions.
Some set-offs enable valuable practices or markets; others are merely wealth distributions from
general to favoured creditors. The former set-offs should definitely be protected: automatically
permitted without being subject to any moratoriums. These set-offs include payment system
netting and derivatives closeouts, both of which substantially reduce settlement-related risk. Other
set-offs do not enable financial market practices, and have a predominantly distributive effect
among general creditors and setoff claimants. These set-offs may be safely left to individual
jurisdictions' sense of equity.
Even clearer criteria exist for rapid liquidation of collateral, free from any insolvency
moratorium. The conceptual key is firm-specificity of collateral. The economic value of firmspecific
collateral would be dissipated if it were liquidated separately from the firm. This insight
can justify the inclusion of such collateral (eg complex machinery) in a reorganisation
proceeding, denying the secured creditor's enforcement rights. (Firm-specificity appears to be the
main argument for reorganisation over liquidation.) Whatever the merits of this argument, it does
not apply to non-firm-specific collateral. No value is dissipated if this collateral is liquidated by
the secured creditor outside the insolvency proceeding: it is worth as much to the secured party as
it is to the insolvent. Such collateral includes securities and receivables. This argument is
especially salient for the collateral securing financial contracts, but applies to all non-firmspecific
collateral.
There is a residual argument for including non-firm-specific collateral in the insolvency estate. A
freeze on creditors' remedies temporarily provides valuable assets for recapitalising the firm.
However, this argument is not consistent with the existence and availability of outside capital
markets, including debtor-in-possession financing. It is worth noting that asset securitisations which
remove those securitised assets from the insolvency proceeding - almost always involve
non-firm-specific assets, such as receivables.
B. Cross-border issues
There are many opportunities for improvement of cross-border insolvency law. There is much to
be gained by broader adoption of existing international standards for general insolvency and
48
collateral law: the UNCITRAL Model Law on Cross-Border Insolvency, the UNCITRAL
Convention on the Assignment of Receivables in International Trade, and - once promulgated the
Hague Conference's Draft Convention on law applicable to securities held through indirect
holding systems. The EU Insolvency Regulation might also be desirable on a wider scale, despite
the practical difficulties of implementing such ambitious legislation. The universality and comity
approaches implicit in the EU insolvency legislation and the UNCITRAL Model Law are useful
and effective developments in dealing with cross-border insolvency.
Several open questions remain. First, should the UNCITRAL Model Law on Cross-Border
Insolvency be extended to banks or other financial institutions, which the Model Law leaves as a
local option? There may be a conflict between currently different national approaches to bank
insolvency law in a cross-border context when some jurisdictions adopt a single entity approach
and others a separate entity approach. One example where the outcome was affected by these
different approaches was the liquidation of BCCI. In the single entity approach assets are pooled
and distributed to creditors according to priorities set for the whole entity (this requires comity of
law to establish an accepted basis for determining the relevant jurisdiction - such as place of
incorporation). In the separate entity approach certain assets (e.g. in a bank insolvency, those
booked to local branches) may be subject to local jurisdiction and used first to meet local claims.
Where jurisdictions adopt different approaches the outcome may not seem equitable to all
creditors.
Second, will widespread adoption of the Hague Convention be sufficient for securities, or is
harmonisation of substantive law also necessary? Securities, of course, are by far the most
significant source of financial collateral in the market, and cross-border collateralisation is now
common. Well-harmonised conflict of law rules are necessary, but may not be sufficient, for a
harmonised cross-border legal regime.49
Third, is there a need for an international netting treaty for settlement risk, and - if so - should
such a treaty be aimed generally or only at the advanced jurisdictions that pose the greatest
settlement risk?
C. Contract- and market-based insolvencies
Since the mid-1970s or early 1980s, a large academic literature on alternatives to insolvency has
developed on both sides of the Atlantic, associated with names such as Oliver Hart, Philippe
Aghion, Mark Roe and Alan Schwartz.50
These names include both economists and lawyers
49
Luca G Radicati di Brozolo, "Conflicts of laws issues of international payments", in International Monetary Law
(M Giovanoli, ed, Oxford 2000).
50
See for example Philippe Aghion, Oliver Hart and John Moore, "The economics of bankruptcy reform", 8 Journal
of Law, Economics & Organisation 523 (1992); Lucien Arye Bebchuk, "A new approach to corporate
reorganisations", 101 Harvard Law Review 775 (1988); Michael Bradley and Michael Rosenzweig, "The untenable
case for Chapter 11", 101 Yale Law Journal, 1043 (1992); Alan Schwartz, "Contracting About Bankruptcy", 13
49
working in the law-and-economics tradition. This literature generally shows a great distrust for
insolvency reorganisations, for many of the reasons discussed above in this report: particularly
the structure of the governance of reorganisation, including attendant decision processes. The
literature distrusts both Official-centred procedures and the negotiation process inherent in most
stakeholder-centred procedures.
The policy prescriptions in this literature vary, from insolvency proceedings variable by ex ante
contract to various auction or option proceedings.
Auction proceedings assume an outside source of capital: employing either all-cash or more
complex bids to establish the value of the firm, and having the bids approved by the creditors
through some kind of voting or tender procedure, which accommodates holdout problems.
Insolvency auctions are not novel: many restructurings and even bulk liquidations employ
auctions. However, in these contexts, auctions are not the centrepiece of the insolvency
procedure, as has been recommended by some of the academic literature. Conventional Chapter
11-style reorganisations still assume that outside capital is the exception, rather than the rule.
Option procedures employ internal capital, eg by issuing options with appropriate exercise
prices or redemption rights granted depending on the holder's position in the capital structure.
Appropriate transferable options will create incentives for the option holders to, in effect, value
the firm. Like the auction proceeding, an option procedure uses a market mechanism to replace
the negotiation mechanism inherent in modern reorganisations. However, the option procedure
does not require an outside capital market.
Insolvency-by-contract is a truly radical idea: that all creditors can contract in advance to a
particular insolvency process or rule. This idea is probably infeasible if taken to its limit, because
involuntary (ie tort) creditors cannot contract, and many voluntary creditors cannot meaningfully
contract (eg trade creditors, employees). However, it might work well to adjust the claims of
financial creditors, especially if they are viewed as coming serially to the firm and the monopoly
advantages to the first creditor are not excessive. In this respect, it resembles a prepackaged
reorganisation negotiated ex ante, before distress. (The usual scope of prepackaged
reorganisations is limited to financial creditors; trade creditors are typically paid in full.)
Credit derivatives do not formally affect insolvencies, but may transform the incentives of the
players, and thus indirectly insolvency proceedings. Credit derivatives permit a creditor to
contract with a third party to hedge its claim on a debtor and to receive a payment immediately
upon the occurrence of a default or other "credit event", with such a payment usually based on the
actual or proxied market valuation of the claim. The third party usually becomes the creditor,
generally without any pre-existing relationship with the debtor. Post-insolvency trading of claims
("vulture capitalism") is similar in this respect. Experience with insolvencies involving credit
Journal of Law, Economics & Organisation, 127 (1997); Mark Roe, Corporate reorganisation and bankruptcy
(Foundation 2000).
50
derivatives to date is probably too small (and the recovery process too understudied) to draw
immediate conclusions. However, credit derivatives might plausibly reduce the pre-insolvency
monitoring incentives of creditors (including their incentives to effect a workout), and create a
class of specialised creditors-in-insolvency. The legal underpinnings of the law of credit
derivatives are not those of insolvency law, but rather concern their enforceability: both as against
the third party and against the insolvent.
D. Prepackaged reorganisations
So-called "prepackaged" reorganisations, although still fairly new, are popular among insolvency
practitioners, and appear superior in some respects to conventional reorganisations. A
prepackaged reorganisation relies on the same kind of negotiations as a conventional
reorganisation, but mostly occurs before the insolvency proceeding is initiated. The initiation of
insolvency is a fait accompli, intended only to insure legal protection to out-of-court agreements.
A prepackaged insolvency, therefore, disposes with the moratorium, or at least shortens the
moratorium period considerably, and probably lowers the costs significantly.
Prepackaged reorganisations are most plausible when a firm has no difficulty meeting its present
obligations out of ordinary business revenues (and thus does not need a moratorium), but
management nevertheless feels the need to restructure. Recent legal and market evolution has
provided such opportunities. For example, some junk bond financings are not payable in the early
years (or are "payable in kind"), but impose a severe cash drain in later years. During these early
years, management might decide that it will not be able to meet the new cash obligations when
they become due. (By no coincidence, prepackaged reorganisations first appeared around 1986,
and became very popular by the early 1990s.) Prepackaged reorganisations are also potentially
useful in settling pending mass tort cases. Such cases cannot always be settled with a cash
payment, because their settlement value exceeds the value of the firm. A prepackaged
reorganisation involving the tort creditors becomes a sensible way to settle such a case, resolving
tort liability and financial structure simultaneously.
However, if management does not enter the process until the firm has cash flow problems, the
prepackaged reorganisation is unlikely to succeed, unless the negotiations can somehow be
accelerated. And, as discussed above, management will often have no incentive to do so.
E. Insolvency arbitration
Arbitration is an "alternative dispute resolution" procedure. An arbitrator's powers are akin to
those of a judge, for the parties are bound by the arbitration agreement. An arbitrator, however,
has no power over those who have not consented to arbitrate.
Insolvency arbitration is often proposed by insolvency practitioners. Proponents of insolvency
arbitration believe that specialist insolvency practitioners are better than judges, especially
outside advanced industrial countries. They therefore recommend insolvency arbitrations that
51
bind the debtor and its financial creditors (who can all consent), and pay other creditors in full.51
Proponents of insolvency arbitration also believe that arbitration clauses should be enforced for
bilateral insolvency disputes, such as claims resolution and asset collection.
F. Reducing the insolvency estate
As discussed above, modern firms have fewer firm-specific assets than they once had. This is
responsible, in part, for a significant trend in insolvency law, which has not received the
recognition it has deserved. This approach, in effect, segregates firm-specific assets from nonspecific
assets, and excludes many of the non-specific assets from the insolvency proceeding.
This new approach has at least three sources: the increasing use of off-balance sheet financing
(particularly asset securitisations), the increasing popularity of outsourcing, and the new carveouts
to ordinary insolvency laws that protect financial contracting and liquidation of financial
collateral. Outsourcing is usually considered a kind of exogenous factor, unrelated to insolvency.
However, a heavily outsourced firm has a smaller balance sheet, and it will therefore have fewer
assets tied up in a reorganisation or other insolvency proceeding. The more controversial
measures involve off-balance sheet financing and financial collateral liquidation.
Asset securitisation finances various promises to pay, such as mortgages and trade receivables.
These assets are not very firm-specific: they are worth almost as much to an assignee as they are
to the original creditor. They are therefore easy to finance, using a special purpose vehicle
structure, which has the legal form of an asset sale free from the insolvency proceeding, but the
substance of a secured transaction.
Asset securitisation effectively shrinks the size of the insolvency estate, by carving some nonspecific
assets from the estate. (Securitisation reduces the estate's obligations pari passu.)
Securitisation is attractive. The financier obviously benefits, because it can liquidate its claim
outside of the deadweight losses imposed by insolvency proceedings. The debtor benefits through
cheaper financing, extended because the financier bears less risk. The other creditors also benefit
from the cheaper financing, which decreases the burden on the estate. However, this ex ante
benefit to other creditors disappears ex post, when the parties have entered the insolvency
process. Unsecured creditors get no benefit in insolvency from asset securitisation, because
traditional reorganisations typically redistribute wealth from secured to unsecured creditors (at
least with respect to many liquidation priority schemes.) The financiers may still rely on the
insolvent bank for servicing and other ancillary activities.
The same result is achieved directly with financial contracts and collateral, in many jurisdictions.
As with asset securitisations, this collateral does not have any appreciable firm-specific value in
the hands of the insolvent. Furthermore, the ability to liquidate this collateral is essential in
51
See UNCITRAL, "Alternative approaches to out-of-court insolvency processes,"  24, Working Group
A/CN.9/WG.V/WP.55.
52
preserving systemic liquidity: a special concern for financial institutions. As discussed above, this
result is obtained with carve-outs from ordinary insolvency law.
Developments like these, driven by the markets, are likely to increase efficiency and legal
certainty by some measures, but may not be equitable according to current national insolvency
standards or consistent with financial stability. In unusual instances, such developments can on
occasion impair financial stability. The markets, for example, favoured the so-called "walkaway
clause" for derivatives contracts in the early 1990s, in which creditors who owed money to a
defaulting counterparty after all contracts were netted had the right not to pay. In other words, an
insolvent counterparty who would otherwise be "in-the-money" after closeout received nothing.
This right, although jointly agreed by the counterparties, was inconsistent with the common goal
of insolvency processes to maximise the value of the bankruptcy estate, and conceivably could
have perverse incentives for debtors and creditors alike as a troubled firm approached insolvency.
The walkaway clause only disappeared when bank regulators refused to recognise netting in
computing capital requirements if the netting agreement contained such a clause.
However, such market-driven change is not necessarily inferior or inequitable relative to other
approaches. After all, the key legal foundations of modern payment system risk reduction and
derivatives netting have received powerful support - and frequently leadership - from market
participants. Modern financial markets could not function safely without such foundations. With
adequate disclosure, unsecured and secured creditors alike might experience much greater legal
certainty relative to the status quo, permitting them to contract with much clearer expectations
about the mean and variance of their losses net of recoveries, in the event of default.
G. The Official-centred restructuring
The Official-centred restructuring, when viewed as an emerging legal underpinning of insolvency
law, is extremely controversial. In some jurisdictions, it has been long-tried and discredited. In
other jurisdictions, it has been extremely successful over the past few decades.
The theoretical promise of Official-centred restructurings is easy to understand. In principle, it
can offer the speed and efficiency of market-based restructurings, with far greater flexibility.
Unitary decision-makers are far faster than a negotiation process. Liability transfers are possible:
another degree of flexibility unavailable in a reorganisation. Finally, the agency costs of a creditor
representative are almost certainly greater than the corresponding costs for bureaucrats. Or
viewed the other way, an Official-centred restructuring can have most of the flexibility of a
Chapter 11-style reorganisation, without the need for protracted negotiations. Many of these
theoretical promises have proven themselves in practice, eg in the US Federal Deposit Insurance
Corporation receivership, the staple of US bank insolvency law.
However, the practical failure of the Official-centred restructuring in many jurisdictions is also
easy to understand. An Official generally has poor incentives to successfully reorganise a firm,
compared to financial creditors. An Official generally gets little reward from a successful
53
restructuring, and may be subject to political pressures that are inconsistent with success. It is
worth noting that the Federal Deposit Insurance Commission - which restructures US banks - is
also the bank insurer, and has a direct financial stake as residual claimant. Official-centred
restructurings may also suffer from some informational disadvantages compared to stakeholdercentred
reorganisations.52
These informational disadvantages may be mitigated, however, for
supervised institutions.
Official-centred restructurings therefore can be promising if the incentives and institutions are
right. But given current incentives and institutions, they appear at best as a speciality, limited to
insolvencies of insured and supervised financial institutions. The range of historical experience
across countries may provide some valuable lessons about incentives in the insolvency process as
many firms become more bank-like in their activities and risk exposures.
H. Forum shopping and legal arbitrage
Two related issues also need consideration: forum shopping" and legal arbitrage. As discussed
above, forum shopping exploits the power of an interested party to choose an insolvency forum
(and thus insolvency regime) without regarding the interests of other parties. Forum shopping is
currently difficult, but may become easier as firms become more global. Legal arbitrage involves
choosing different parts of different national legal systems. In doing so, market participants may
create a legal framework that does not exist in any national jurisdiction. Often, legal arbitrage is
feasible even in the field of insolvency law, where the possibility to exploit this " la carte"
approach is generally limited due to the mandatory nature of insolvency rules.53
The current level
of legal arbitrage may be tolerable, but the trend to increasing arbitrage opportunities is
worrisome. Rapid legal evolution promotes arbitrage, especially because the evolution is
generally in the direction of greater party autonomy and fewer mandatory rules. Globalisation
further increases the salience of this trend.
A degree of forum shopping and legal arbitrage provides incentives for individual countries to
adapt their laws to changing conditions and market practices. Taken to an extreme, however,
competition among jurisdictions can lead to wasted resources in the form of legal arbitrage and
the possibility of efforts to change the insolvency venue in the late stages of distress that could
alter the positions of creditors and other stakeholders.
52
Specifically, incumbent stakeholders may possess private information inaccessible to even a competent and wellmotivated
Official. Some of this information is related to the value of the firm; other information relates to risk
preference. This latter information can be particularly significant in restructurings that transform claims, leading to
different results for different general creditors, eg financial debt to equity, or short-dated debt to longer-dated debt.
For strategic reasons, these preferences may be best revealed through a negotiation process.
53
For example, some EU initiatives in the field of insolvency law seem to imply some forms of "legal arbitrage",
since the law regulating the effects of insolvency on certain kinds of financial contracts is not the law governing the
insolvency proceeding, but the law applicable to the transaction as chosen by the parties.
54
Hence, issues concerning forum shopping and legal arbitrage need to be examined closely to
identify an optimal international solution, which should assure transparency of legal rules,
certainty of financial transactions, reduction of informational problems and elimination of
unintended effects such as moral hazard and cherry-picking.