The globalization of capital markets has drawn increasing attention
to the prudential regulation of banks and securities firms. In an era
of high capital mobility in the industrialized countries, market
volatility and competitive pressures place great strains on financial
regulators. Shocks such as the less-developed countries (LDC) debt
crisis and the 1987 stock market crash can lead to a crisis of
confidence in a country's regulatory environment. However, if
regulators respond unilaterally with strict and costly regulations,
they may put their financial sectors at a competitive disadvantage with
foreign rivals. Faced with this “regulator's dilemma,”
financial regulators from the industrialized countries have initiated a
range of efforts to harmonize their prudential regulations—that
is, to negotiate internationally accepted minimum requirements for the
stability of financial institutions.1
On
the regulator's dilemma, see Kapstein
1989. Financial regulators use the terms harmonize and
harmonization to refer to the purposive efforts by regulators
to agree on a common set of regulations for a given issue area. I
differentiate harmonization from the process of regulatory convergence
that arises organically through market pressures or emulation. In this
article, international regulatory harmonization implies an
agreement between regulators—that is, the purposive creation of
an international regulatory standard.
The variation in
regulator behavior in these negotiations is striking: certain
regulators are adamant in pushing for an international standard, while
others remain aloof or actively resist the creation of an international
agreement.
The political science literature has offered few studies of the
circumstances under which financial regulators will demand
international standards. Existing studies of the 1988 Basel Capital
Accord—the most prominent example of international regulatory
harmonization—either assume that regulators have the lofty goal
of mitigating global systemic risk, or neglect the importance of
regulators as agents in the negotiation process.2
By treating the
Basel Accord as an isolated case, these studies miss an important
opportunity to offer a more general model of the politics of
international regulatory harmonization. A recent study by Simmons
examines the processes of regulatory harmonization in four financial
issue areas but uses a country's “incentives to
emulate” as an independent variable rather than specifying
systemically what those incentives are and how they vary.
3 As regulators gain prominence as international
actors in the current era of globalization, it is imperative that
scholars of international relations learn more about their patterns of
cooperation. This article contributes to the literature by explaining
the circumstances under which financial regulators will seek to
harmonize with their foreign counterparts—or, in other words, to
explain precisely what the incentives are that lead a regulator to
press for harmonization.
I propose an analytical framework that captures the competing
domestic pressures on regulators, and the role of international
regulatory harmonization in addressing these pressures. The framework
assumes a principal-agent relationship between a legislature and a
regulator.4
The legislature, as the principal,
delegates the responsibility for setting and implementing financial
regulations to a regulatory agency and imposes boundaries on that
agency's policymaking through the threat of legislative
intervention. The legislature maximizes a combination of campaign
contributions and aggregate welfare, whereas the regulator is only
concerned with maintaining its decision-making autonomy. Furthermore,
the regulator is limited to a single policy tool of regulatory
stringency, unlike the legislature, which has a range of policy options
at its disposal. The regulator chooses a degree of regulatory
stringency that falls within its “win-set”—the range
of policy choices that do not result in legislative intervention.
Exogenous shocks to international competitiveness or voter confidence
in financial stability can decrease the size of the win-set and make
intervention more likely. In such circumstances, the regulator has
incentives to seek international regulatory harmonization as a means of
increasing the size of its win-set and safeguarding its autonomy. In
short, the regulator's domestic political environment prompts an
international solution.
I explore the consistency of the framework with two important cases
in the history of financial regulation: the 1988 Basel Accord, which
established a minimum capital standard for internationally active banks
in the Group of Ten (G-10) industrialized countries; and the
negotiations over capital adequacy for securities firms, which took
place under the auspices of the International Organization of
Securities Commissions (IOSCO) from 1988 until 1992. The Basel Accord
is familiar to scholars of international political economy as the one
of the most prominent examples of international regulatory
harmonization. However, the IOSCO effort—which ended without an
agreement—represents a “negative” case and has
therefore been overlooked in academic circles, despite its obvious
parallels to the Basel negotiations.5
While academics may overlook the case, regulators and
industry executives are quick to mention the IOSCO negotiations as a
striking counterexample to the Basel Accord. Author's interviews
with current and former Basel Committee members, 18–20 June 2002,
Basel, Switzerland.
More important for this study is that the
cases reveal significant variation in regulators' desires for an
international standard. In the Basel negotiations, the U.S. Federal
Reserve was one of the primary proponents of an international capital
adequacy standard, whereas the U.S. Securities and Exchange Commission
(SEC) was vocal in its resistance of a similar standard for securities
firms. Regulators in the United Kingdom (U.K.) advocated for
international standards in both cases, whereas Japanese regulators were
resistant in both.
This article proceeds as follows. First, I review two existing
studies of the Basel Accord and use elements of both to develop an
improved framework for analyzing international regulatory
harmonization. I then provide an overview of the importance of capital
adequacy regulations in a globalized world, including a brief overview
of the Basel Accord and a more lengthy account of the IOSCO
negotiations. I follow with a critique of the existing literature in
light of the two cases and demonstrate that the new
“confidence-competitiveness” framework provides a more
compelling explanation of demands for international regulatory
harmonization. I conclude with a reevaluation of the current literature
on regulatory harmonization and a brief typology to guide future
research.
Explaining International Regulatory
Harmonization
In the past twenty years, there have been several attempts by
regulators to harmonize financial regulations, from establishing
capital adequacy standards to payments system requirements to
supervisory standards.6
These negotiations are notable
not for the convergence of views among the participants, but rather for
the wide variation in the preferences of regulators toward the creation
of an international standard. While certain regulators are strong
proponents of an international standard, others are adamantly opposed.
How can one explain the varying preferences of regulators toward
international regulatory harmonization?
The political science literature on financial regulatory
harmonization has focused almost exclusively on the Basel Accord, an
agreement that established an international capital adequacy standard
for commercial banks. Kapstein argues that the accord was created as a
result of regulators' consensual knowledge of the systemic risks
of undercapitalized banks. As banks' capital levels declined
throughout the 1970s and 1980s, they became more vulnerable to losses
from loan defaults and exogenous shocks. Kapstein's argument,
which has become conventional wisdom among economists and political
scientists, is that the adoption of minimum capital standards by the
G-10 countries provided the global public good of financial stability
and thus was in regulators' collective interests.7
When applied beyond the specific case of Basel, this argument implies
that harmonization will occur whenever an international regulatory
standard is needed to address systemic problems, including financial
instability.
8 In a recent paper, Oatley and Nabors challenge this functionalist
logic by arguing that the Basel Accord is an example of
redistributive cooperation: “the creation of an
international institution that intentionally reduces at least one other
government's welfare compared to the status quo.”9
They argue that the U.S. Congress legislated
tough capital adequacy requirements domestically in 1983 and directed
U.S. regulators to impose these regulations onto foreign
competitors—especially the Japanese—through an
international agreement. More generally, Oatley and Nabors argue that
politicians drive the international regulatory process: electoral
incentives lead legislators to shift the costs of their policies onto
foreign countries. International regulatory harmonization, then,
represents the particular interests of one country's politicians,
rather than a jointly provided public good. Oatley and Nabors imply
that regulators themselves are important only in that they carry out
the directives of the legislature.
The framework presented here synthesizes elements of both of these
arguments but offers a more compelling and analytically useful
explanation of regulator preferences. Similarly to Kapstein, I
incorporate regulators as important actors in international regulatory
harmonization. In the past fifteen years, regulators have become
important actors on the world stage, creating new possibilities for
international cooperation. Regulators have considerable discretion in
coordinating with their foreign counterparts. Unlike many other types
of international agreements, regulatory agreements are usually not
ratified by legislatures, nor are they legally binding on signatories.
Nevertheless, these agreements are far from inconsequential: market
forces and pressure from international organizations help to ensure
compliance with global regulatory standards.10
Any model of international
regulatory harmonization must therefore include regulators as key
actors. However, I also incorporate Oatley and Nabors's focus on
legislatures and domestic politics more generally. The behavior of
regulators is constrained by the preferences of elected officials. A
complete analysis of regulatory harmonization therefore requires an
integrative approach that accounts for the incentives of both
regulators and legislatures.
Analytical Framework
Legislators and financial regulators are engaged in a principal-agent
relationship.11
Legislators, as the
principals, delegate the responsibility for setting and implementing
financial regulations to regulatory agencies. Regulators make frequent
modifications to the regulatory environment for firms—such as
altering prudential supervisory protocols, tightening capital adequacy
rules, and changing reporting requirements—but rarely do these
changes arise because of observable pressure from elected officials.
However, if a regulator enacts a policy that runs counter to the
interests of elected leaders, the legislature can intervene and change
the policy.
Political intervention is the bane of a regulator's
existence.12
This corresponds to the
notion in the principal-agent literature that there are costs to the
agent when its policy is overturned. See Ferejohn
and Shipan 1990.
When politicians attempt directly to
influence regulatory policy—for example, by holding hearings and
publicly criticizing the decisions of regulators, or by legislating new
regulations—they threaten the agency's autonomy and
prestige. Intervention may also affect regulators' future job
prospects, especially for an agency head who is forced to resign. The
prospect of intervention by legislators therefore creates
ex
ante constraints on regulators' discretion, which ensures
that the principals can maintain some control over the agent.
Regulators will use all strategies at their disposal to minimize the
possibility of intervention.
I begin by assuming that the legislature maximizes a combination of
aggregate welfare and campaign contributions, in accordance with
standard political economy models of politics.13
This means that the legislature responds to the
demands of financial firms (the primary source of campaign
contributions), but also that it seeks to enact policies that are
beneficial for voters as a whole. In an open economy, firms are
interested in profits, and they seek to maximize their competitiveness
with foreign rivals. Voters, on the other hand, demand financial
stability—or what the financial press often calls
“confidence”—which arises when banks and securities
firms are resistant to collapse and insolvency. The legislature can
influence competitiveness through industrial policy (for example,
tariffs, subsidies, barriers to entry), regulatory policy, and other
legislative options. Likewise, the legislature can influence confidence
by creating new regulatory agencies or oversight boards, altering
existing agencies, or legislating new prudential restrictions for
firms. Consider, for example, the creation of the Public Company
Accounting Oversight Board (PCAOB) in the wake of the Enron and
WorldCom financial scandals.
14 The PCAOB
emerged from the Sarbanes-Oxley Act of 2002, U.S. Public Law 107-204;
30 July 2002.
Congress created the PCAOB to increase
confidence in U.S. companies after the scandals galvanized the
attention of voters toward accounting lapses and corporate
malfeasance.
Note that confidence and competitiveness do not move in lockstep.
Policies that improve one variable often have negative consequences for
the other variable. For example, new legislation that allows financial
institutions to engage in high-risk activities may improve
competitiveness at the expense of confidence. Also, the legislature
must take into account the effects of legislation on other welfare
considerations, such as aggregate economic growth and productivity.
The legislature delegates the responsibility for setting and
implementing prudential regulations to the regulatory agency. I further
assume that the legislature incurs a cost for intervening in regulatory
policymaking. This cost can be seen as the time and resources needed to
monitor and criticize a regulator's policy, the opportunity cost
of not working on other legislative issues, or the resources needed to
draft and implement new legislation.
To minimize the possibility of legislative intervention, the
regulator must take into account the legislature's preferences.
Unlike the legislator, the regulator only has one tool at its disposal:
regulatory stringency. Regulations that are too lax (for example, low
minimum capital levels for financial institutions) will ultimately
contribute to faltering firms and a crisis of confidence among voters,
triggering a swift intervention by elected officials. On the other
hand, in an open economy with competitive financial markets,
regulations that are too strict will put domestic firms at a
competitive disadvantage with foreign firms. Regulators therefore face
a trade-off between confidence and competitiveness. As Walter notes,
In going about their business, regulators continuously face the
possibility that “inadequate” regulation will result in
costly failures, on the one hand, and on the other hand the possibility
that “overregulation” will create opportunity costs in the
form of financial efficiencies not achieved, or in the relocation of
firms and financial transactions to other regulatory regimes.15
The threat of political intervention creates a limited zone of
acceptance—or what could be called a
“win-set”—for regulatory policy.16
On the “zone of acceptance” for bureaucracies,
see Meier 1985; on win-sets, see Putnam 1988. Moe 1987
offers a similar argument about constraints on bureaucracies.
I depict the win-set in
Figure 1. The two
graphs show the effects of regulatory stringency on competitiveness and
confidence, respectively. The horizontal axis on both graphs represents
the regulator's range of policy choices, with lax regulations on
the left side and stringent regulations on the right side. For
simplicity, I depict each relationship as linear: as regulations become
more stringent, competitiveness declines (top graph) and confidence
increases (bottom graph). Each graph contains an intervention threshold
for the legislature, labeled I
cmp for competitiveness (top
graph) and I
cnf for confidence (bottom graph). When levels
of either confidence or competitiveness fall below these thresholds,
the legislature's utility for intervening will outweigh the cost.
To avoid intervention, the regulator must maintain confidence and
competitiveness above these thresholds.
In Figure 1, the legislature's
intervention thresholds correspond to regulatory policy choices on the
horizontal axis. To maintain a level of competitiveness above
Icmp, the regulator must enact a policy that is no more
stringent than point B, as indicated by the vertical dotted line.
Regulations more stringent than B will lead to a loss of market share
to foreign firms and will trigger political intervention on behalf of
the domestic financial sector. Similarly, to maintain confidence above
Icnf, the regulator must enact a policy that is no more lax
than point A. Regulations to the left of this point will lead to a
crisis of confidence in the financial system and will trigger political
intervention on behalf of voters. The space to the right of A and to
the left of B represents the regulator's win-set, because policies
that fall between these thresholds will not trigger political
intervention.
It is critical for the framework that a win-set generally exists
(that is, that A is to the left of B). Of course, if there were no
win-set, then there would be no delegation from the legislature to the
regulator. When a win-set exists, regulators can generally make small
changes in regulation without significant consequence—and indeed,
regulators frequently modify existing regulations without much fanfare.
However, a substantial loosening of regulations will eventually lead to
widespread financial instability, whereas a substantial tightening of
regulations will eventually lead to a sizeable loss of market share for
domestic firms. The range of policies within these extremes constitutes
the win-set.
The Role of International Regulatory Harmonization
The applicability of this framework to international harmonization
can be seen when exogenous shocks are taken into account. Because it is
impossible to ascertain the precise location of the win-set, I focus
instead on changes in its size. When a win-set exists, regulators can
use their discretion to set regulatory policy within the threshold
points. However, exogenous shocks to confidence or competitiveness can
shift the thresholds toward each other, reducing the size of the
win-set or eliminating it altogether. For example, in the event of a
high-profile failure of one or more financial institutions, voters will
have less confidence in the efficacy of existing regulations to ensure
the stability of domestic firms. The U.S. savings and loan (S&L)
crisis in the late 1980s is representative: as S&Ls throughout the
country began to fail, voters became aware that the existing regulatory
environment was insufficient to ensure the stability of their
neighborhood thrifts. Voters were not well informed of the specifics of
thrift supervision, but the series of S&L insolvencies led to
implicit demands for more stringent regulations to shore up stability.
More generally, shocks to confidence occur when there are innovations
and structural changes in financial markets, resulting in increased
risk for financial institutions. In the S&L example, one of the
most salient changes was an increase in interest rates, which created
tremendous pressure on S&Ls to find higher-yielding assets. As
shown in Figure 2, an exogenous shock to
confidence shifts the confidence line downward (to the bold dotted
line), indicating that more stringent regulations are required to
maintain any prior level of stability. As a result, A shifts to the
right (to A′), indicating the legislature's implicit demands
for more stringent regulations to maintain confidence at its existing
level. Note that the legislature's preferences over confidence do
not change, but the exogenous shock alters the regulatory policy needed
to maintain that degree of confidence.
On the other hand, an exogenous shock to competitiveness would occur
if domestic firms incurred a loss of market share to foreign firms with
less stringent regulations. As an example, consider the rise of the
Eurocurrency markets in the 1960s, largely as a result of the
congressionally mandated interest equalization tax and Regulation Q
interest rate ceilings in the United States.17
These policies
pushed dollars out of U.S. banks and into banks in London and elsewhere
in Europe, where regulations were not nearly as strict.
18 Reinicke 1995.
The outflow of dollars was not entirely detrimental to U.S. firm
competitiveness, because some of these funds landed in foreign branches
of U.S. banks.
An exogenous shock to competitiveness, such as
the rise of the Euromarkets, shifts the competitiveness line downward
for the adversely affected domestic financial sector, as shown in
Figure 3. This shift implies that less stringent
regulations are required to maintain the prior level of
competitiveness. After the shock, the regulator must choose a policy to
the left of B′ to avoid legislative intervention.
When the size of the win-set decreases, the probability of political
intervention increases. If an exogenous shock is powerful
enough—or if there are simultaneous shocks to voter confidence
and firm competitiveness—then the thresholds can cross each other
such that no win-set exists for regulators, as shown in Figure 4. Given that there are no policies both to
the right of A′ and to the left of B′, there is no win-set
for regulators. Unless the regulator can shift one or both of the
threshold points, then it is impossible to meet the minimum
requirements of the legislature.
No win-set after exogenous shocks
The absence of a win-set necessarily implies that the legislature
will intervene. But it also raises the question, how and why would the
government intervene if it is impossible to satisfy both groups? The
answer is that the win-set applies to regulators, not legislatures. The
framework posits that regulators can only employ regulatory stringency
to affect voter confidence and firm competitiveness. A legislature,
however, has other options at its disposal. For example, the
legislature can enact trade barriers, subsidies, and tax breaks to
bolster firm competitiveness, or repeal costly legislation, such as the
Regulatory Q interest rate ceilings or restrictions on bank lending.
These options are not available to regulators. Likewise, it can reverse
a downward spiral in voter confidence through a highly publicized
change in the structure of the regulatory agency. For example, the
legislature can create a new regulatory body with oversight and
enforcement powers, such as the PCAOB. In more extreme cases, the
legislature can abolish an agency entirely and create a new one in its
place. Such was the case in the aftermath of the S&L crisis, when
Congress dismantled the existing S&L regulator—the Federal
Home Loan Bank Board (FHLBB)—and created the Office of Thrift
Supervision.19
The congressional
restructuring of S&L regulation is contained in the Financial
Institutions Reform, Recovery, and Enforcement Act of 1989 (FIRREA),
U.S. Public Law 101-73.
It was certainly possible for Congress
to institute substantive regulatory changes within the existing FHLBB,
but the creation of the OTS constituted a signal to the public of a
new, more prudent regulatory environment.
20 Figure 4 indicates that regulators are
powerless to protect their authority unless they can effect a shift in
the threshold points. Such a shift can be achieved through
international regulatory harmonization. Smoothing over differences in
regulation affects the competitiveness of firms, either positively or
negatively depending on the stringency of regulations before
international harmonization. This is important in cases where domestic
firms are losing market share to foreign firms in less stringent
regulatory environments. It is also important when declining voter
confidence requires the enactment of more stringent regulations that
would put domestic firms at a competitive disadvantage with foreign
rivals. If international regulatory harmonization is possible at the
more stringent level, then the regulator can bolster confidence without
harming domestic competitiveness. More generally, I argue that a
regulator is more likely to seek international regulatory harmonization
when its win-set is shrinking because of exogenous shocks. An inability
to fend off domestic political pressures leads the regulator to seek an
international solution.
Consider the threshold points in Figure 4.
Ideally, the regulator wishes to increase the stringency of regulations
to point A′ without harming domestic competitiveness. If the
regulator can successfully push for international regulatory
harmonization at this more stringent level, then the competitive
effects of the more stringent regulations will be minimized. In other
words, the competitiveness line will shift upward—indicating a
change in foreign regulations that has competitive implications for
domestic firms—and B′ will correspondingly shift to the
right and create a viable win-set for regulatory policy.
This analytical framework yields observable implications about
demands for international regulatory harmonization. First, I expect
international regulatory harmonization to be initiated most often by
regulators themselves, not by legislatures. This is an important
implication of the principal-agent framework and stands in contrast to
scholars who argue that elected officials are the only relevant actors
in the creation of international standards. The regulator seeks an
international standard in response to an increased probability of
legislative intervention, which in turn is generated by changes in
voter confidence and financial competitiveness. While the threat of
legislative intervention creates constraints on the regulator, the
legislature itself does not dictate the regulator's strategy. The
legislature may in fact express opposition to an international
agreement, as shown in the Basel case below, or remain largely aloof,
as in the IOSCO case.
Second, a regulator is more likely to seek international regulatory
harmonization when less stringently regulated foreign firms are
capturing market share from domestic firms. The regulator could enact
lax standards in line with the foreign competitor, but such a policy
choice would result in decreased stability, which shrinks the win-set
and increases the probability of legislative intervention.
International regulatory harmonization in this instance addresses the
competitive distortions caused by the difference between foreign and
domestic regulations without contributing to a decline in voter
confidence. Regulators facing a loss of market share are likely to push
for an international standard based on their more stringent domestic
regulations. On the other hand, when domestic firms are dominant and
face minimal competition from foreign firms, then the domestic
regulator will be unlikely to press for harmonization.
Third, regulators are more likely to seek international regulatory
harmonization when confidence in the stability of financial
institutions is declining. It is assumed in the model that the only
tool for regulators to bolster confidence is regulatory stringency.
Because the enactment of stricter regulations affects the competitive
position of domestic firms, regulators can push for international
regulatory harmonization to mitigate the competitive effects of
increased stringency. Simultaneous exogenous shocks to confidence and
competitiveness (as shown in Figure 4) have
the effect of shrinking the regulator's win-set more sharply than
a single shock to just one of these variables. Thus, when dual
exogenous shocks occur, regulators are especially likely to push for
international regulatory harmonization. Confidence and competitiveness
are therefore additive in their influence on regulators'
decisions.
The framework presented above is not designed to capture all of the
nuances of regulatory policymaking in the industrialized world. Rather,
it is designed to be broadly applicable to any political environment in
which the regulator is granted authority by a popularly elected body.
Certain regulators are more insulated from political pressures than
others, and not all regulators have the same scope or mandate. These
sources of variation are deliberately overlooked by the model in
exchange for a more general account of regulatory politics in an
international context.
Finally, the framework is designed to explain why regulators demand
international harmonization, not how they achieve it. Understanding
preferences is the first step in a more theoretically complete analysis
of the circumstances under which regulators will create international
standards. Once one understands who wants what and why, one is in a
much better position to explain harmonization outcomes using variables
such as market power and international institutions.21
Kapstein 1989 and Oatley and Nabors 1998 are in agreement on the
importance of market power in explaining the emergence of a
multilateral agreement for bank capital adequacy. On the role of
international institutions in the process of harmonization, see Simmons 2001.
In the cases that follow,
the final outcomes—harmonization for banking, no harmonization
for securities—are not particularly puzzling once one understands
the preferences of the key participants in the negotiations.
Case Studies: Capital Adequacy for Banks
and Securities Firms
Banks and securities firms are in the business of taking risks. Banks
lend money directly to businesses, governments, and individuals,
thereby incurring credit risk: the risk that borrowers will
fail to repay their loans. Securities firms underwrite new issues of
marketable securities, serve as market-makers in the secondary markets,
and often engage in their own speculative trading. In these activities,
firms assume market risk for as long as it takes them to sell
any securities in their inventories.22
As
an example, consider a securities firm that agrees to underwrite an
offering of stock by fictitious ABC Corporation. In a so-called
firm commitment, the securities firm purchases all of the
stock from ABC for resale to the public. The profit from this
transaction consists of the difference between the price paid to ABC
and the price charged to the public. The securities firm faces market
risk from the time it takes possession of the ABC stock until it is
sold to the public, which could be anywhere from a couple of days to
several weeks. If the market value of ABC's stock were to plummet
during this period, the securities firm could incur a substantial
loss.
The willingness of firms to assume these risks allows
for the transfer of surplus capital from savers to borrowers—the
essence of global capital markets.
Capital itself is critical for both banks and securities firms, as it
provides a cushion against losses that result from borrower default or
changes in asset prices. Capital also provides incentives for bank
managers to lend prudently, as the capital of the firm—and not
just customers' deposits and other government-insured
liabilities—is at risk.23
I thank
the editors of IO for this point.
Both types of firms view
capital reserves as necessary for their prosperity and stability.
However, in the event of severe adverse conditions, regulators view
capital quite differently for banks and securities firms.
24 For
banks, the regulator's goal is to enable the bank to weather
adversity and thereby protect depositors' funds and public
confidence in the banking system. Capital levels should be sufficient
to absorb losses and enable the bank to continue as a going
concern.
25 For securities firms, capital regulations are
designed to allow a firm to wind down its operations in an orderly
manner in the event of severe market losses. This would include selling
off marketable securities, repaying any outstanding obligations, and
shutting down operations with sufficient warning to the market.
26 In the United States, the SEC's net
capital rule (SEC Rule 15c3-1) stipulates that a broker-dealer should
have the capacity to wind down its operations and protect its customers
within one month. See Haberman 1987.
Note that capital requirements are designed to prevent insolvency and
default for both banks and securities firms; however, in the event of
severe trouble, banks should stay standing whereas securities firms
should wind down.
Initially, capital adequacy was a purely domestic issue. Regulators
feared the possibility of contagion—where the collapse of a
financial institution could lead to similar collapses and a widespread
crisis of confidence in the financial system—but this fear did
not extend to other countries until banking and securities markets
began a rapid process of internationalization in the 1970s and 1980s.
In 1974, the failure of Bankhaus Herstatt, a medium-sized German
commercial bank involved in the foreign exchange markets, led to a
temporary halt in the international payments system. Stock markets
began to show similar interdependence beginning in the 1980s,
culminating in the worldwide stock market crash in 1987. This evidence
of “systemic risk” gave regulators incentives to
communicate with their foreign counterparts through international
committees and organizations. Groups such as the Basel Committee and
the IOSCO enabled regulators from different countries to share
“best practices,” discuss different regulatory approaches,
and sometimes create international regulatory standards.
The Basel Case
The events leading up to the Basel Accord have been thoroughly
documented,27
so only a brief summary is provided here.
Soon after the Herstatt fiasco, central bankers from the G-10
industrialized countries formed the Standing Committee on Banking
Regulations and Supervisory Practices, now known as the “Basel
Committee” because of its home at the Bank for International
Settlements in Basel, Switzerland. The purpose of the committee was to
provide a forum for bank regulators from the industrialized world to
share ideas and best practices, and to address the issue of systemic
risk in international banking markets. In the early 1980s, the Basel
Committee began to investigate the wide-scale deterioration of capital
levels in internationally active banks. The LDC debt crisis of 1982
served as a wake-up call to regulators about the dangers of low capital
levels, as banks in New York, London, and Tokyo struggled with
substantial losses on their lending portfolios. The committee began
work on a set of guidelines for capital adequacy, but progress was
slow. Central bankers could not decide on a proper definition of
“capital,” let alone agree on an appropriate minimum level
that banks should hold. Also, Japanese banks were operating with
substantially less capital than Western banks, which allowed them to
offer more favorable pricing than their competitors. French banks also
had relatively low capital levels and resisted any movement toward
stricter regulations. There may have been a superficial consensus that
bank capital levels were a systemic problem, but French and especially
Japanese regulators were resistant to creating an international
standard that would prove enormously costly to their respective banking
markets.
In 1987, in a move that surprised the Basel Committee, the Bank of
England and the U.S. Federal Reserve signed a bilateral agreement on
capital adequacy. This Anglo-American agreement established a
“risk-weighted” standard in which capital requirements
would increase with the degree of risk of a bank's loan portfolio.
It was clear from the beginning that the agreement was not intended to
last in isolation; rather, it was a strategy to force the Basel
Committee into a multilateral agreement favorable to U.S. and U.K.
regulators.28
Author's interviews
with senior Bank of England officials, 24–25 June 2002, London.
See also Kapstein 1989.
Implicit in
this Anglo-American “zone of cooperation” was the threat of
excluding noncompliant countries' banks from British and American
markets.
29 As Kapstein notes, “The tacit threat
of preventing foreign banks from expanding operations or establishing
new ones within that zone was apparently credible enough to move
discussions to the multilateral level.”
30 After several months of negotiations
to smooth over the differences between the U.S./U.K. coalition and
the rest of the G-10, the Basel Committee published the Basel Accord in
late 1987. The accord immediately superceded the Anglo-American
agreement, and established a global standard for minimum capital
levels.
The IOSCO Case
Several studies were published in the 1980s that validated what
regulators and investors already knew: equity prices in the major
markets—including the United States, Europe, and Japan—were
becoming increasingly correlated.31
A sudden plunge in equity prices in one exchange
would often reverberate quickly throughout other exchanges, regardless of
market fundamentals. A prime example was the stock market crash of 1987,
which was centered in New York but spread quickly to other markets throughout
the world. This interdependence highlighted the potentially deleterious
consequences of the collapse of a major securities firm. If a firm in London,
for example, were to go bankrupt, the firm's outstanding transactions
with broker-dealers in other countries would go unrealized. If these
firms were not sufficiently capitalized, they could default or
collapse, creating an ever larger circle of bankruptcies worldwide.
32 An influential OECD
publication in 1991 stated the problem more starkly:
The extreme systemic threat arising from a collapse of
securities prices is that default by one or more large securities
dealers will lead to further defaults and that the failures will extend
into the core of the banking system and cause a breakdown in the flow
of payments in settlement of financial transactions throughout the
world.33
Fortunately the stock market crash of 1987 did not lead to such a
dire outcome, but the event served as a jarring lesson for regulators
about the potential dangers of interdependent markets. The crash was
responsible for spurring a number of research reports on
“systemic risk” in securities markets.34
For a comprehensive list of reports stemming from the 1987
stock market crash, see Tobin 1991,
282–83.
Regulators were also growing more aware of the
risks involved in securities firms' investment strategies, from
currency trading and futures to interest rate swaps and complex
derivatives.
35 The late 1980s marked the emergence of the IOSCO as an important
forum for securities regulators. In 1986, IOSCO created a
“Technical Committee,” consisting of regulators from the
developed countries, to guide the work of its members.36
In 1988 these countries were the United
States, United Kingdom, Japan, West Germany, France, Australia, Canada,
Hong Kong, Italy, the Netherlands, Sweden, and Switzerland.
It
also increased its membership to include most of the world's major
stock exchanges.
37 By the end of 1989,
IOSCO had forty-eight members. Tobin 1991,
315.
In 1987, the Technical Committee created a working group
to study the issue of capital adequacy for securities firms, and
included regulators from the United States, France, Japan, and the
U.K.
38 In 1989, the
group issued its first report, which was then approved by the full
Technical Committee for presentation at IOSCO's annual meeting.
The report concluded that a common framework was needed regarding the
capital requirements of securities firms, and should contain the
following elements:
39 1. Liquidity and solvency should be covered by a standard that
provides for a firm to have sufficient liquid assets to meet its
obligations given the risks a firm faces.
2. Marking of marketable securities and commodities positions to
market is necessary to prevent firms from storing up losses and also to
give a true picture of a firm's position.
3. Risk-based requirements should cover all the risks to a firm and,
in particular, should contain:
i. A base requirement reflecting the scale of a firm's
activities to capture nonmeasurable risks.
ii. Position risk requirements (for both on- and off-balance
sheet items) reflecting the price volatility of individual securities with
provisions for concentrated positions and allowances for risk reduction
measures such as hedging.
iii. Settlement risk requirements reflecting the risk of
nonperformance in a timely manner.
The bottom line was that firms should hold enough capital to exceed
the sum of these risk-based requirements.
The 1989 IOSCO report was merely a set of guidelines that set the
agenda for further negotiations. The group had yet to decide how to
measure capital, let alone what the specific minimum level of capital
should be for securities firms. Another obstacle was the relationship
between the proposed IOSCO standard and the Basel Accord. Because an
increasing number of banks were involved in securities activities, the
competitive implications of any new regulations were of critical
importance. In the three years after the 1989 report was published, the
IOSCO Technical Committee and the Basel Committee exchanged a number of
issue papers in an attempt to arrive at a consensus view of the
appropriate capital requirements for any firm conducting securities
business. At a meeting in late January 1992, negotiators tentatively
agreed that securities firms would be required to hold capital
equivalent to four percent of their gross holdings plus eight percent
of their net holdings, after netting out long against short
positions.40
Economist, 31
October 1992, 76. See also Steil 1994. A
“long position” means that a firm has ownership of a
security, whereas a “short position” means that a firm has
“sold short” a security by delivering borrowed shares to
the purchaser, but has yet to cover its position by buying the shares
in the market. Firms (and individuals) often take short positions to
protect the profits in their long positions.
At the conclusion
of the meeting, the U.K.'s Securities and Investments Board
(SIB)—the most vocal advocate of a capital standard—was
optimistic that a formal IOSCO agreement would be forthcoming at the
October annual meeting.
41 Author's
interview with former SIB official, 27 June 2001, London.
The Technical Committee had scheduled one final meeting in July 1992
before the annual meeting, with the intent of drafting a preliminary
agreement. However, as it became clear that an agreement was actually
feasible, U.S. regulators surprised the other participants by opposing
the creation of a capital standard.42
International Securities Regulation Report, 28
July 1992, 1.
A particularly contentious issue was the extent
to which a securities firm should be able to reduce its capital
requirement through hedging.
43 The SIB backed a plan that would allow a firm to
carry 2 percent of the sum of its long and short positions, assuming
the firm had a perfectly hedged book as defined in the proposal.
44 Financial Times, 28 October 1992,
30.
This plan was substantially lower than the SEC's
existing standard of 15 percent. Richard Breeden, Chairman of the SEC,
said that under the SIB's proposed rules, a major U.S. securities
firm would have failed after the October 1987 crash, and refused to
endorse a standard that he called “dangerously low.”
45 Economist, 31 October 1992, 76. In
early 1992, the SEC requested that the Federal Reserve Bank of New York
run a simulation of the 1987 stock market crash, assuming that U.S.
firms were capitalized at the SIB's proposed level. The study
found that the crash would have led to the collapse of one of the
largest securities firms in the country. Author's interview with
former senior SEC official, 7 May 2002, New York.
He also
argued that IOSCO should be a “clearing house of ideas” and
not a rule maker.
46 Financial
Times, 28 October 1992, 30.
The IOSCO annual meeting
therefore produced no agreement, and the Technical Committee went back
to the drawing board to see if a consensus could still be created.
However, despite the SIB's protests, it was clear that no further
progress would be made on the issue, and IOSCO officially abandoned the
effort to harmonize capital adequacy regulations.
47 Financial Times, 11 February 1993,
29.
Analysis
Under what conditions will regulators seek to create global financial
standards? The Basel and IOSCO cases described above present a puzzle.
In both cases, bank and securities regulators came together to discuss
capital adequacy regulations for their respective industries.
Negotiations occurred at roughly the same time under similar market
conditions. But ultimately the negotiations unfolded in different ways,
with varying coalitions advocating or resisting harmonization. How can
one explain this variation, and what do these cases say about the
larger phenomenon of international regulatory harmonization? I begin by
critiquing the current literature on the Basel Accord, which for
convenience I divide into “systemic risk” and
“redistributive” arguments. After showing that these
explanations cannot satisfactorily explain the variation in the two
cases, I advance a new argument based on the confidence-competitiveness
framework that is more successful in explaining the variation.
Systemic Risk
Kapstein argues that the Basel Accord was created as a result of
international “consensual knowledge” of the systemic risks
of bank lending, combined with the leadership of the United States and
the U.K.48
In establishing the importance of consensual
knowledge, he points to the failure of the Franklin National Bank and
Bankhaus Herstatt as the impetus for the creation of the Basel
Committee. He argues that these prominent firm failures, along with the
LDC debt crisis in 1982, led to a consensus among regulators of the
risks of global capital markets. This consensus was echoed in a number
of publications by international financial organizations, including the
OECD's influential publication
The Internationalisation of
Banking.
49 Kapstein acknowledges that consensual knowledge of systemic risk was
necessary but not sufficient to bring about an international agreement.
When it became clear that the Basel Committee was reaching a stalemate,
the United States and the U.K. announced a bilateral agreement on
capital adequacy in 1987 to jumpstart the Basel negotiations. He
implies that all of the G-10 countries were eager to create a global
standard, but it took a show of market power to move the negotiations
along. This view is in fact common among scholars who have studied the
Basel Accord.50
The main challenge to Kapstein's argument is that it cannot
explain the dynamics of the Basel negotiations. It is crucial to
remember that capital adequacy regulations are costly, as they affect
banks' profit margins. Assuming that regulators are rational,
there are tremendous incentives for countries to free ride and let
other countries assume the costs of global financial stability. As the
primary explanatory variable, systemic risk cannot explain why U.S. and
U.K. regulators exerted significant energy to bring about an agreement,
while Japanese regulators (with seven of the ten largest banks at the
time51
) remained on the sidelines and resisted an
increase in capital standards.
Systemic risk is also an unhelpful variable in the IOSCO case. Note
that the IOSCO capital adequacy story unfolds almost in parallel to the
Basel story. The debt crisis galvanized the attention of bank
regulators just as the 1987 stock market crash galvanized the attention
of securities regulators, and academic publications noted the
international interdependence of markets in each case. Securities
regulators all agreed about the systemic risks of international
securities markets, but the IOSCO negotiations revealed stark
differences in regulator preferences toward harmonization.52
Some scholars have argued that the risk of
contagious collapse is unique to banks, and that the rationale for
regulating the capital levels of securities firms is not clear. Dale,
for example, argues that a securities firm in trouble can simply
liquidate its holdings at market value without disrupting other firms.
Dale 1996; see also Herring and Litan 1995. While there is debate about
the systemic risks of securities markets within academia, securities
regulators—the key actors in this case—agreed that capital
adequacy for securities firms was a global concern. See IOSCO 1989.
During the negotiations, U.S.
regulators realized that their capital standard for broker-dealers was more
stringent than virtually anywhere else. Yet rather than use their market
power to push through an IOSCO standard at existing U.S. stringency levels,
U.S. regulators ultimately pulled out of the negotiations and ensured the
defeat of a global standard.
A functionalist theory also cannot explain the behavior of regulators
during the IOSCO negotiations. The leader in the push for harmonized
capital standards was the U.K.'s SIB. This is especially curious
because broker-dealers in London faced less stringent capital
requirements than in the United States. Rather than boost capital
requirements to U.S. levels, the SIB wanted to harmonize at or near its
existing capital levels. U.S. regulators deemed these levels to be
insufficient to prevent against insolvency, and they were concerned
that the SIB's proposed standard—albeit a minimum
requirement—would put downward pressure on U.S. and European
capital levels.53
Author's interview
with former senior SEC official, 7 May 2002, New York.
If
systemic risk fuels regulator interest in global standards, it is
difficult to understand the SIB's demands to validate a capital
standard deemed “dangerously low” by the world's
largest securities market.
54Redistributive Cooperation
In a recent paper, Oatley and Nabors challenge Kapstein's
functionalist logic by arguing that the Basel Accord was an example of
“redistributive cooperation.”55
In 1983, the developed world looked to the International Monetary Fund
(IMF) to bail out large Western banks that were failing because of the
LDC debt crisis—an effort that would require a substantial
increase in IMF resources. The United States agreed to increase its IMF
quota by $8.4 billion, subject to the approval of a wary Congress. With
a recession in full swing, voters became indignant at the prospect of
using taxpayer dollars to coddle a handful of multinational banks.
56 It should be noted that an IMF quota
increase does not require additional government spending or taxes. This
fact is overlooked in the Oatley and Nabors analysis.
At the
same time, the banks were struggling with wide-scale default from Latin
America and a rising competitive threat from less-regulated Japanese
banks. Oatley and Nabors argue that from the perspective of Congress, a
reasonable compromise was to move ahead with the IMF quota increase,
but to force banks to take responsibility for their imprudent lending
behavior by enacting stricter regulations. Such a move, however, would
entail imposing a competitive disadvantage on the domestic banking
sector, potentially hastening the rise of Japanese banks.
To address the competing pressures from voters and banks, Congress
embedded the IMF quota increase in the International Lending
Supervision Act (ILSA) of 1983. The ILSA dictated that U.S. regulators
should increase domestic capital adequacy standards and seek to
coordinate prudential standards on an international level. Oatley and
Nabors argue that the ILSA satisfied voters by forcing banks to raise
new capital and assume at least part of the responsibility for their
aggressive lending behavior. At the same time, they argue that the
mandate to seek an international agreement assuaged banks'
concerns about the loss of market share that could result if they faced
stricter regulations than their foreign counterparts, especially the
Japanese. Oatley and Nabors contend that ultimately it was the exercise
of U.S. market power—as evidenced by the initial Anglo-American
agreement—that led inexorably to the Basel Accord.
While the importance of market power in the creation of the Basel
Accord is beyond question, a sole emphasis on congressional directives
yields an incomplete story. It is important to note that there is a
five-year gap between the passage of the ILSA in 1983 and the creation
of the Basel Accord in 1988, and that U.S. regulators unilaterally
increased domestic capital standards a second time during this
period.57
Congressional sentiment in
the years leading up to the accord was inconsistent. Throughout the
negotiations over ILSA, Congress was adamantly in favor of new capital
adequacy regulations for U.S. banks, reflecting voters' desire to
make the banks take responsibility for their lending behavior during the
debt crisis. However, as the Basel Accord was being finalized in late 1987,
Congress began to vocalize its concerns about the competitive ramifications
to U.S. banks, and to speak out against the Basel Accord. For example,
U.S. Representatives Charles Schumer (D-N.Y.) and Norman Shumway
(R-Calif.) circulated a memo to the House Banking Committee that
questioned the competitive implications of the Basel agreement.
58 Schumer stated, “I am
concerned that unanticipated and unnecessary effects of the regulations
may seriously jeopardize the international competitiveness of American
banks.”
59 The irony
is unmistakable: in 1983 the House Banking Committee mandated higher
domestic capital adequacy standards and an international regulatory
agreement through the ILSA, but five years later that same committee
began holding hearings on the competitive implications of the Basel
Accord.
60 Legislative pressures were clearly salient in
influencing regulator behavior, but a principal-agent framework is
required for a full explanation of the preferences of U.S. regulators
during the Basel negotiations.
A sole focus on elected officials is similarly unhelpful in the IOSCO
case. In the securities markets there was no sudden regulatory
change—akin to the putative voter-mandated increase in capital
adequacy standards in the United States in 1983—that would prompt
a regulator to shift the costs of stringent regulations onto foreign
competitors through international regulatory harmonization. In fact,
during the 1980s and early 1990s, securities capital adequacy
regulations were quite stable throughout Europe and the United States.
Moreover, politicians in Europe were focused on creating a Capital
Adequacy Directive (CAD) that would promulgate a single capital
adequacy standard for all financial institutions in the European Union,
including banks and securities firms. During the early 1990s, most
observers expected that the formula in the Basel Accord would
ultimately find its way into the CAD, leaving no room for a separate
standard for securities activity.61
Author's interviews with former senior official of
the SIB and financial industry executives, 25–27 June 2002,
London.
European politicians, then, were not involved in
initiating the IOSCO negotiations, nor did they play a significant role
as the negotiations progressed.
Nevertheless, the Oatley and Nabors framework, with its emphasis on
redistributive politics, draws attention to the competitive element
behind the IOSCO negotiations. In Europe, regulators calculated capital
requirements on a consolidated basis—that is, capital
had to cover the risks associated with all divisions and subsidiaries
of the firm. In the United States, capital requirements only applied to
registered “broker-dealers”; parent or holding companies
were not regulated. This lack of regulatory oversight gave U.S. firms a
competitive advantage over European firms in the growing derivatives
market.62
Author's interviews with
former senior official of the SIB and senior financial industry
executive, 25–27 June 2002, London.
I address the
importance of these competitive pressures in the following section.
The Confidence-Competitiveness Framework
The confidence-competitiveness framework assumes that regulators
choose policies that safeguard their decision making from direct
political intervention. This leads regulators to strike a balance
between the competitiveness of regulated firms and voter confidence in
the stability of financial institutions. Regulators are therefore more
likely to seek international regulatory harmonization when confidence
is declining, or when less-regulated foreign firms are infringing on
the market share of domestic firms. As I discuss below, this framework
provides a compelling explanation of regulator demands in the Basel and
IOSCO cases.
Basel: Dual Shocks for the United States and the
United Kingdom
Both the United States and the U.K. faced a rising competitive threat
from Japanese banks during the 1980s. Table 1
shows the total assets of the ten largest banks in the world for 1974,
1984, and 1994, indicating the tremendous growth of Japanese banks at
the expense of U.S. and British firms. By 1986, Citicorp fell from its
symbolically important perch as the largest bank in the world to the
sixth largest, behind five Japanese banks.63
More
importantly, U.S. and U.K. markets were home to a growing proportion of
Japanese bank assets. The lending activity of Japan's overseas
branches exploded throughout the 1980s. As shown in
Table 2, Japanese branches in the United States
experienced a 315 percent increase in total assets between 1981 and
1988, while the figure for Japanese branches in the U.K. was 232
percent.
64 By 1988, more than 38 percent
of the assets of Japanese banks were held in overseas branches, mostly
in the United States and the U.K.
65 In 1985, Japanese international lending
outpaced U.S. lending for the first time ever.
66 Financial Times, 31 January 1986, 24.
The United States and the U.K. therefore each experienced an exogenous
shock to competitiveness in the mid-1980s, as shown in
Figure 3.
The ten largest banks in the world (in millions of U.S.
dollars)
Location of assets of Japanese banks, 1981–88 (in billions of
U.S. dollars)
It is beyond the scope of this article to address the specific
reasons for the growing strength of the Japanese banking sector.
However, it is clear that if Japanese banks were to hold the same level
of capital as their competitors in the United States and U.K., their
competitive advantage would be severely cut.67
Kapur 2000. It was estimated
that individual Japanese banks would have to raise $20 to $30 billion
by 1992 to meet the new requirements. Washington Post, 13 July
1998, F1.
In 1986, Citicorp and Barclays (U.K.) had
capital-to-asset ratios of 4.73 and 4.71, respectively, while
Japan's Dai-Ichi Kangyo, Sumitomo, and Fuji had ratios of 2.38,
2.89, and 2.95.
68 Gerald Corrigan, head of the New York
Federal Reserve during the Basel negotiations, stated bluntly:
“The single item on which I place the greatest emphasis relates
to … the goal of moving Japanese bank capital standards into
closer alignment with emerging international standards.”
69 Exogenous shocks to confidence were also salient in the Basel case.
In 1982, the LDC debt crisis exposed the imprudent lending behavior of
a number of U.S. and British multinational banks. By the end of 1981,
the exposure of U.S. banks to Latin America amounted to nearly 100
percent of capital.70
When Mexico and other
countries in Latin America defaulted on their loans, U.S. and British
banks were threatened with insolvency. An increase in IMF resources
helped to stem the crisis in the short term, but it was clear that
market confidence was badly shaken. The U.S. Federal Reserve tightened
capital requirements in 1983 as part of the International Lending
Supervision Act, but a number of developments continued to challenge
the confidence in banking markets in the United States and the U.K. In
1984, the tenth largest bank in the United States, Continental
Illinois, began to falter after it acquired a large number of risky
oil-related credits from Penn Square Bank, which itself failed in
1982.
71 To avert a wide-scale financial crisis, the
Federal Reserve stabilized the foreign exchange market and provided a
bailout package to the bank. That same year, the U.K. faced a similar
crisis when Johnson Matthey Bankers became insolvent after it
concentrated its lending to a small number of high-risk firms. The Bank
of England had no choice but to provide a rescue package and appoint a
new management team for the bank.
72 Both of these bank failures were
embarrassing to regulators and called into question the stability of
each country's banks. Moreover, U.S. bank failures continued into
the decade: 468 banks failed between 1985 and 1987, resulting in
disbursements from the U.S. Federal Deposit Insurance Corporation of
nearly $13 billion.
73 Voter confidence in financial stability was badly
shaken in both countries, resulting in the confidence shock depicted in
Figure 2. Political intervention seemed close at
hand; indeed, regulators in the U.K. were ignominiously summoned to
Parliament to discuss their role in enabling the Johnson Matthey
insolvency.
74 Author's interview
with senior Bank of England official, 26 June 2002, London.
While bank failures were drawing attention to regulators, banks in
the United States and the U.K. were devising new financial products
that could be kept off their balance sheets, and therefore immune from
capital requirements. These “contingent liabilities” were
just as risky as conventional loans, but current capital regulations
did not apply to them. It was clear to regulators that the decline in
market confidence, coupled with banks' clever avoidance of capital
requirements, warranted stricter regulations.75
The environment in Japan in the mid-1980s was much different.
Japanese banks were not nearly as exposed to LDCs during the debt
crisis as were banks in the United States and U.K., and there were no
high-profile bank insolvencies throughout the 1980s. The relationship
between business and government in Japan was much tighter than in the
West, and banks had an implicit guarantee of government support in
difficult times, well before insolvency became a possibility; indeed,
until recently, no bank in Japan had failed.76
The Japanese government's support of banks implied that exogenous
shocks to confidence were rare. This contrasts with the United States
and the U.K., in which government financial support of banks was
reactive—in the form of public bailouts in the event of
confidence-jarring bank failures—rather than preemptive.
With the United States and the U.K. experiencing simultaneous shocks
to confidence and competitiveness, regulators in each country shared a
strong desire for international regulatory harmonization. In the
absence of a change to Japanese regulations, it was becoming
increasingly difficult for these regulators to obtain sufficient levels
of confidence and competitiveness. More stringent regulations were
necessary to bolster stability, but the resulting loss of
competitiveness was too great to bear. In order to create a viable
win-set for regulatory policy, regulators from the United States and
the U.K. were adamant in their support of an international capital
adequacy standard, as evidenced by the Anglo-American agreement in
1987.
IOSCO: The U.K. as the Primary Advocate
Throughout the IOSCO negotiations, the U.K.'s SIB was the
leading advocate of a global standard for securities firm capital
adequacy, while the United States was a vocal opponent. At the heart of
the controversy was the issue of consolidated supervision. As mentioned
earlier, European regulators imposed capital requirements on securities
firms on a consolidated basis, whereas the U.S. SEC imposed
requirements only on registered broker-dealers. While the SIB's
requirements were lower than in the United States, the fact that they
were imposed on a consolidated basis was more important in
considerations of international competitiveness. Parent companies in
the United States could engage in a number of highly lucrative
financial activities without concern for the SEC's capital
regulations. In particular, U.S. firms were able to offer very
competitive pricing on derivatives contracts and other off-balance
sheet items, threatening the market share of firms in London.77
Author's interviews with former senior
SIB official and senior financial industry executive, 25 June 2002,
London.
Capital requirements for broker-dealers, on the other
hand, are less important in terms of international competition, in
large part because the global market for debt and equity remains
relatively fragmented by geographic and regulatory barriers.
78 The SIB's concern over the competitive implications of
consolidated supervision began with the growth of the derivatives
market in the 1980s. The Chicago Board of Trade first developed
financial futures contracts in 1975, and the United States dominated
the market through the mid-1980s.79
Other types of
derivatives, including stock index futures and interest rate swaps,
emerged in 1982.
80 British
firms quickly entered the business of derivatives trading, and a
significant futures and swaps market emerged in London by the late
1980s. By the end of 1992, the volume of outstanding derivatives
worldwide was approximately $21 trillion. Firms in other European
countries, especially Switzerland, were active in this market, but with
the exception of Mitsubishi Bank, the Japanese were not yet major
players.
81 By the early 1990s, firms in London were fighting for their share of
this lucrative market. Reliable country-specific data on derivatives
markets in the early 1990s are notoriously difficult to find, but U.S.
firms such as Goldman Sachs and Morgan Stanley were viewed as superior
derivative providers compared to British firms such as Morgan Grenfell
and Warburg.82
British
industry executives and regulators are quick to point out that U.S.
derivatives providers avoided SEC capital rules by using their holding
companies to effect transactions. As a result, these providers could
offer more competitive pricing on swaps and futures and maintain a
dominant position in the global market.
83 Author's interviews with former SIB official and
executive at Goldman Sachs, 26 June 2002, London.
Regulators
in the U.K. therefore experienced a shock to competitiveness, as shown
in
Figure 3.
On the other hand, maintaining confidence in securities markets (and
securities firms) was a constant constraint on regulators, because
instability increased the probability of legislative intervention. The
existing capital adequacy regulations in the United States were deemed
just right by U.S. regulators, especially after the October 1987 stock
market crash. Despite the volatility in U.S. equity prices, one former
regulator said that, from the public's perspective, the SEC made
it through the crash “with flying colors.”84
Author's interview with former senior
SEC official, 7 May 2002, New York.
There were no major firm
failures,
85 Three small, specialized
investment firms became insolvent after the crash. Wall Street
Journal, 21 October 1987, 26.
and no generalized fear
that the capital held by securities firms was insufficient to weather
the crash. Investors in London, however, were less confident in the
stability of their securities firms. The crash occurred in the middle
of a £7.25 billion equity offering, the world's largest
public offering of stock. On October 15, just four days before
“Black Monday,” the British government released details of
an offer to sell its remaining 31 percent stake in British Petroleum
(BP). Merchant bank N.M. Rothschild was the lead manager of the
offering, along with a syndicate of sixteen underwriters in London and
a small number of firms in North America, Continental Europe, and
Japan. When the stock market plunged, underwriters were faced with
tremendous losses, estimated at nearly £1 billion.
86 After intense pressure from investment banks, the
British government instructed the Bank of England to repurchase a
portion of the shares, but underwriters were nonetheless left with
losses totaling some £700 million.
87 With confidence in British firms in
question, the SIB did not have the option of enacting less stringent
regulations to counteract the loss of market share to U.S. securities
firms.
British regulators therefore found themselves with a shrinking
win-set for regulatory policy and political pressures on all sides.
Financial institutions in London faced tremendous difficulty in
maintaining their share of the derivatives market and were frustrated
by the absence of consolidated supervision for U.S. firms. In addition,
the declining capital levels of British securities firms indicated an
overall drop in international competitiveness. This was borne out by
the acquisition of all but one of the major securities houses in London
by foreign financial institutions soon after the IOSCO negotiations
failed to change the regulatory environment in the United States.88
Acquisitions included Morgan Grenfell by
Deutsche Bank, SG Warburg by Swiss Bank Corporation, Kleinwort Benson
by Dredsner, and Barings by ING. Shroder was the only major securities
firm in London that was not acquired. See Rogers
1999.
Overseeing the downfall of an industry is a sure
way for a regulator to invite legislative intervention.
89 The SIB ceased operations in 1997 with the
creation of the Financial Services Authority.
On the other
hand, the stock market crash reminded the SIB that it did not have the
luxury of enacting less stringent regulations for providers of
derivatives. As the financial press questioned the viability of the SIB
and the job prospects of its embattled chairman,
90 See, for example, Independent, 28 January 1990,
4.
the agency sought an international solution: if the United
States adopted consolidated capital rules, then British firms would
become more price-competitive as U.S. firms were forced to hold more
capital to back up their derivatives trading. The SIB therefore took
the lead in promoting international harmonization, even if that meant
that the SEC would lower its capital requirements—or at least
validate a lower international standard—for broker-dealers to
match the SIB's regulations.
Consolidated supervision was an implicit component of the IOSCO
negotiations. It is difficult to reconcile the SEC's sole focus on
broker-dealers with the spirit of the 1989 IOSCO report, which
emphasized such concepts as “a firm's true position”
and an amalgamation of all risks faced by a firm.91
The SEC
itself was pushing for new legislation to force parent companies of U.S.
broker-dealers to share financial information with the SEC. SEC Chairman
Richard Breeden was a staunch supporter of a bill sponsored by Sen. John
Heinz (R-Pa.) and Sen. Christopher Dodd (D-Ct.) that would amend the 1934
Securities Exchange Act by requiring holding companies and affiliated
entities of registered broker-dealers to submit periodic financial reports
to the SEC.
92 The bill, called the Market Reform Act, received
a push by the failure of securities firm Drexel Burnham Lambert in
early 1990. As the Drexel holding company was on the verge of declaring
bankruptcy, it began to siphon capital away from its solvent registered
broker-dealer, which was regulated by the SEC. The SEC only become
aware of this transfer after more than half of the broker-dealer's
capital—some $400 million—had been depleted, and just
eleven days before the Drexel holding company declared bankruptcy.
93 Washington Post, 3 March 1990,
C1.
As Breeden stated during a Senate Banking Committee
hearing, “If there's a five-alarm fire raging, we think we
ought to know about it.”
94 The bill was eventually
passed as the Market Reform Act of 1990, and constituted a significant
change in the supervision of holding companies.
95 U.S. Public Law 101-432. Note that the legislation applied
only to holding companies of broker-dealers. Bank holding companies
were already subject to consolidated supervision in the United
States.
Given the growing awareness of the vulnerability of a registered
broker-dealer to the unregulated activities of its parent, the SEC had
to be wary of a global capital adequacy standard. Requiring information
from holding companies is a clear step toward regulating their
activities. The SEC had no interest in expanding its capital
requirement to include holding companies; indeed, financial firms made
their resistance to such a development clearly felt in Congress during
the debates over the Market Reform Act.96
Washington Post, 3 March 1990, C1.
But
in light of Breeden's push for more information from unregulated
affiliates, a global capital adequacy standard—enforced by nearly
every other regulator on a consolidated basis—would put the SEC
on a clear path toward regulatory supervision of holding companies.
97 Author's interview with former senior
SIB official, 27 June 2002, London. Such pressure is hinted at in Walker 1992.
As these developments
unfolded, the SEC—eager to defend its own policymaking discretion
and the competitiveness of U.S. securities firms—made its
preferences known by unceremoniously pulling out of the IOSCO
negotiations.
Conclusion
Efforts to harmonize financial regulations across countries have
become a common, although often overlooked, occurrence in the world
economy. However, there have been few attempts in the political science
literature to explain cross-national variation in preferences toward
harmonization. In this article, the principal-agent relationship
between legislatures and regulators is critical to understanding why
countries demand global financial standards. I argue that
regulators' incentives emerge from the possibility of legislative
intervention; the legislature's incentives, in turn, derive from
the need to choose an optimal trade-off between financial stability and
international competitiveness. Regulators must, therefore, use the only
tool at their disposal—regulatory policy—to maintain a
balance between confidence and competitiveness. In the event of an
exogenous shock to confidence or competitiveness, regulatory policy may
be ineffective in maintaining this balance unilaterally, in which case
regulators have incentives to seek an international regulatory
agreement to maintain their autonomy.
The arguments that I present here differ from the existing literature
in two main ways. First, I focus on the varying preferences of national
regulators in the context of legislative constraints, rather than on
systemic concepts such as international market failures and global
public goods. I take issue with studies that neglect the
principal-agent relationship between the legislature and the regulator,
and instead I argue that regulators maintain decision-making discretion
within the bounds set by domestic politics. Second, I focus on the
trade-off between voter confidence and financial sector
competitiveness. Much of the public debate on globalization emphasizes
the possibility of “regulatory arbitrage,” where capital
will flow to the least regulated areas, thereby inducing a regulatory
race to the bottom.98
These analyses do not account for
the critical role of voter confidence in constraining regulatory
policymaking.
My focus throughout this article has been on prudential regulations,
which are designed to foster the stability of financial institutions
and other enterprises. However, the dual variables of confidence and
competitiveness are relevant to a variety of other international
harmonization efforts. One example is money laundering, a growing
concern among governments, regulators, and financial institutions.99
Money laundering is the processing of the
proceeds from criminal activity in an attempt to hide its illegal
origins.
Anti–money laundering regulations require banks
and other financial institutions to file currency transaction reports,
keep adequate records on the identities of account holders, and monitor
suspicious activity. Such requirements can be expensive, especially for
smaller banking sectors that are eager for capital inflow. Regulatory
laxity can be a significant competitive advantage, as shown by the
vibrant banking sector of the tiny island nation of Vanuatu. The
members of the Group of 7 (G-7) industrialized countries, eager to stop
money laundering and the predicate criminal activities that fuel it,
have led the charge in attacking the market confidence of countries
that maintain lax regulations. The G-7's Financial Action Task
Force (FATF) conducts periodic regulatory reviews of countries around
the world, and publishes the names of countries that do not comply with
its “40 recommendations” to fight money laundering. By
drawing attention to these countries, the FATF hopes to discourage
investment in their markets—that is, effect a downward shift in
the competitiveness line in
Figure
3—and eventually bring about appropriate regulatory
changes.
100 As a final note, the framework presented here, with its emphasis on
the industrialized world, highlights an important set of distinctions
that should be noted in the literature on harmonization. There are in
fact three broad types of international regulatory harmonization. The
first, regulatory convergence, is the organic process by which
countries modify their regulations based on the policies of other
countries, especially dominant countries such as the United States, or
simply converge on a common set of rules inadvertently.101
Simmons and Elkins (forthcoming) use the related term “diffusion”
to refer to the international spread of liberal economic policies.
A prime example is bank deposit insurance, which is common in many countries
but was until recently a U.S. phenomenon.
102
The second type of harmonization, which I label
core harmonization,
is the process emphasized in this article in which a small group of advanced
industrialized countries agree, through overt negotiation, to harmonize
their regulations. The result of successful core harmonization is an
international standard, often with a formal name (for example, the
Basel Accord). The creation of an international standard gives rise to
the third type of harmonization,
peripheral harmonization, in
which countries outside the core group of industrialized countries
choose whether to accede to the standard or to maintain divergent
regulations. Third World countries, for example, often tout that they
are “Basel-compliant” to increase investor confidence in
their banking systems; other countries intentionally diverge from the
FATF's anti–money laundering recommendations as a means of
attracting capital.
Scholars should be clear about which type of harmonization is under
investigation, because each requires a separate analytical approach.
Analyses of regulatory convergence might look at epistemic communities,
global economic and geopolitical conditions, and other factors that
lead regulators and policymakers in multiple countries to adopt similar
policies. Economic incentives in emulating a dominant country are also
important, especially for countries with open markets. Core
harmonization requires an analytical framework rooted in domestic
politics—such as the one presented in this article—because
the varying preferences of the small number of core countries are
critical. Finally, peripheral harmonization focuses one's
attention on the coercive power of the core countries through
international institutions and economic sanctions, as well as the
economic incentives of the peripheral countries themselves in acceding
to the international standard.103
The Basel Accord provides a useful example of the importance of these
types. In the mid-1980s, before the existence of an international
standard, there were strong incentives for countries to undercut each
other with lax capital requirements, because doing so provided a
competitive advantage. However, after the G-10 countries established
the Basel Accord in 1988, these incentives largely disappeared, and
“Basel-compliance” became an important signal to investors
of bank stability. For core harmonization of capital requirements,
competitive pressures were a hindrance; but after the G-10 countries
harmonized, competitive pressures actually ensured peripheral
harmonization.104
On worldwide
compliance with the Basel Accord, see Ho
2002.
Regulatory harmonization, in its various forms, will continue to be a
controversial topic for policymakers and scholars in the years to come.
The host of analytical questions that harmonization raises implies that
scholars must be sensitive to their choice of analytical framework. In
the case of core harmonization, domestic politics and the policy
trade-offs of regulators should take center stage in the political
science literature.
