The Financial Crisis and Global Policy Reforms1

Barry Eichengreen

University of California, Berkeley

Introduction

Given the profusion of papers on the financial crisis, the decision to write another

requires some justification. I have accepted the commission for this one because it is

designed to address an aspect of the problem that remains understudied, namely the

design and coordination of global policy reforms. Scads have been written on reforming

national financial markets. But the reach of financial markets and institutions is global.

Although lip service has been paid to the need for global coordination of those national

policies, progress remains halting. Whether in fact there will be a response justifying the

label “global policy reforms” remains an open question.

My analysis is in two parts corresponding to the two main explanations for the

crisis. The first explanation sees the crisis as rooted in inadequate regulation and

distorted incentives in financial markets. The second sees it as the result of a global

savings glut that fueled an unsustainable credit boom in the United States. The debate

between these two viewpoints – not so much their validity as the weight to be attached to

them – will be played out in the history books. For present purposes I do not attempt to

rank their importance.2 I simply assume the validity of both explanations and, by

1 Prepared for the Federal Reserve Bank of San Francisco’s conference on Asia and the Financial Crisis,
Santa Barbara, California, 19-21 October.
2
For my views on this question see Eichengreen (2009a).

1
implication, their compatibility. In each case I draw first the lessons of the crisis and then

the implications for policy.

National Financial Markets and Global Policy Reforms

The first explanation for the crisis focuses on distorted incentives in financial

markets and the failure of regulators to contain their effects. Deregulation, as symbolized

by the Gramm-Leach-Bliley Act in the U.S. and the Single Market Act in Europe,

allowed financial institutions to take on additional risk. The resulting intensification of

competition encouraged them to do so in response to the pressure to survive. Risk taking

was fueled by the knowledge that big banks were important for financial stability – that

they stood to receive official assistance in the event of difficulties. Compensation

schemes for investment managers, whose bonuses depended on current performance

irrespective of the future profitability or even viability of the firm, encouraged strategies

that maximized short-term returns at the expense of long-term stability. The originate-

and-distribute model of securitization allowed firms to neglect the long-term performance

of the securities they issued since the originating entity had no skin in the game. Rating

agencies, lacking reliable models of delinquency probabilities on subprime mortgages in

a down market, failed to warn of the risks. As advisors on structuring securitizations as

well as issuers of ratings, they faced conflicts of interest.

Regulators, meanwhile, lacked the resources to keep up with the regulated. They

bought into the argument that the banks’ internal models provided reliable forecasts of

value at risk and that improved ability to diversify, repackage and otherwise manage risk

justified ever-lower capital ratios as necessary to minimize the costs of financial

2
intermediation. The fragmented structure of regulation within and across countries meant

that no single regulator possessed a comprehensive picture of the regulated. The

fragmentation of regulation also created scope for regulatory arbitrage – for bank and

nonbank financial firms to shift risky activities both to more permissive jurisdictions and

off-balance-sheet, where capital charges could be evaded.3

Policy Implications. From this diagnosis flows a prescription. First, regulators

should require banks to hold more capital. By giving bank shareholders something to

lose, capital is a disincentive for excessive risk taking. It is also a buffer against risks to

the balance sheet. Well-capitalized banks are less likely to be driven into insolvency by

shocks. They are less likely to lose the public confidence that is critical for the

willingness of other market participants to do business with them. They will be better

placed to expand their balance sheets when investment opportunities arise.

Against this backdrop it is striking that Tier 1 capital (banks’ core capital as

defined in the Basel Capital Accord) has been trending steadily downward in the

advanced economies in recent decades, reflecting the belief that lower capital

requirements reduce the cost of financial services and that banks can safely manage their

affairs from a narrow capital base. The result is that Tier 1 capital in the United States

has effectively gone negative in every recent recession. The traditional response of

regulatory forbearance to allow the banks to earn their way out of this hole becomes less

3
Given this litany of incentive problems, it is not hard to see how the world fell prey to a financial crisis.
In retrospect the more appropriate question would seem: how could it have avoided one given these
conditions? All this might seem blindingly obvious in retrospect. What is less obvious is why these
problems were so inadequately appreciated before the fact. The failure of financial market participants to
sound alarm bells is perhaps understandable, given that they were able to profit handsomely from
exploiting incentives for risk taking and, in the words of Chuck Prince, to keep dancing so long as the
music is playing. The failure of the regulators to do more plausibly reflects intellectual regulatory capture
– the tendency for regulators to buy into the worldview of the regulated. For my money, the most troubling
aspect is the failure of independent observers – including academics – to appreciate the prospective risks
(Eichengreen 2009b).

3
feasible in an era of mark-to-market accounting. Much higher Tier 1 capital requirements

are the obvious answer.

At the same time they adjust capital requirements upward, regulators can usefully

revise them along a number of additional dimensions. Capital requirements should apply

to the consolidated position of the institution whether investments are held on or off

balance sheet.4 Required capital ratios should be predicated on realistic holding periods.5

They should reflect not just the volatility of banks’ investments but also the volatility of

their funding. They should reflect the risk posed by failure of the individual institution to

the stability of the financial system; from this flows the conclusion that more capital

should be required of big banks and connected banks. Required capital ratios should

have a countercyclical influence; they should rise when a bank’s balance sheet is

expanding, rather than falling because its internal model signals less value at risk or the

rating agencies upgrade the securities on its portfolio.

Second, there is the need for a resolution regime for systemically significant

nondepository institutions (both banks and nonbanks) to create a third way between

government assistance and uncontrolled bankruptcy. A commitment not to extend a

bailout will be time inconsistent in the absence of a resolution regime. Auto companies

can keep selling cars even while under bankruptcy protection.6 But standard bankruptcy

procedures are not feasible for financial institutions which require people to believe in

their solvency in order to keep funding them. In addition to the confidence problem there

is the fact that putting a financial institution through bankruptcy can place derivatives

4
As Spanish regulators insisted prior to the crisis.
5
Thus, capital requirements for asset backed securities have been predicted on the assumption of a ten-day
trading horizon, which is patently unrealistic in many cases.
6
At least if someone stands behind their warranties.

4
contracts in limbo, resulting in a domino effect destabilizing other institutions as claims

are suspended and collateral is frozen.7

The very fact that the United States has the Federal Deposit Insurance Corporation

as an alternative to bankruptcy for banks is an indication that the country needs an

equivalent resolution mechanism for bank holding companies. But while we have the

FDIC for banks like the old Citibank, we do not have the equivalent for bank holding

companies like Citigroup. Bank holding companies in the United States are regular

corporations and are therefore subject to regular bankruptcy procedures that lack

preemptive provisions.8 The UK similarly has a special resolution regime for deposit

banks as of 2009 but not for merchant banks and other nonbank financial institutions. So

long as this remains the case, moral hazard will be a problem, and potentially a serious

one.9

7
The Federal Deposit Insurance Act allows the FDIC, when resolving a bank, to transfer certain derivatives
and other qualified financial contracts to third parties, eliminating this problem. But not so the U.S.
Bankruptcy Code to which nondepository institutions are putatively subject. For more on this see the
immediately following paragraph.
8
The Fed does have the power to require a bank holding company to divest its banks if it fails to meet
minimum capital requirements, although the holding company normally has 180 days to complete the
divertiture (Elliot 2009).
9
The U.S. Treasury has proposed extending FDI-like resolution authority to bank holding companies and
their subsidiaries but not to hedge funds, private equity firms, and other non-holding company financial
entities. Seizing, restructuring and re-privatizing a large bank holding company either as a unified whole or
in parts is likely to be more complex than doing so for a depository institution, not just since bank holding
companies are more complex but because each one is unique. Banks should therefore be required to
provide a roadmap for how this can be done. Anil Kashyap, the Bank of England, and the Committee on
Capital Markets Regulation all suggest that bank holding companies should be required to plan their own
funeral arrangements in advance; they should be required to draft a set of instructions for how their
institutions could be quickly dismantled should the need arise. See Kashyap (2009); Bank of England
(2009); Committee on Capital Markets Regulation (2009). Providing regulators with a roadmap would
give them an alternative to bailouts. Reducing bailout incidence would in turn mitigate moral hazard. In
addition, having to sign off on their own funeral arrangements would focus the attention of managers and
directors on the mortality of their institution. It would deter them from taking on additional risks that made
orderly unwinding more difficult – especially if banks whose own plans indicated that more days would be
required for orderly resolution were required to hold commensurately more capital.

5
 A resolution regime for these entities would have the following features.10 The

regulator would be empowered to appoint an administrator. The administrator’s powers

would include firing the management and board and suspending the voting and other

decision rights of the shareholders and unsecured creditors. The administrator could

ring-fence specific activities (securities clearing, settlement and custodian activities) and

instruments (derivative securities etc.). He could transfer deposits to another bank, sell

assets, and mandate debt-for-equity swaps, and finally break up the institution or order its

liquidation.

Third, supervisors must be provided with the information they need to make

informed assessments of risks to stability. This means not starving regulatory agencies of

the resources required for information gathering.11 It means imposing reporting

requirements on hedge funds and other nonbank financial institutions of potential

systemic significance, requiring them to provide information relevant to the regulator’s

micro- and macro-prudential tasks.12

Fourth, regulators should address problems in markets for derivative securities.

They should seek to better align the economic interests of originators and investors by

requiring the originator of any securitized asset to retain a meaningful portion of the

equity or first-loss tranche. This will give the originator a stake in the subsequent

performance of the issue. The Treasury blueprint for reforming the U.S. financial system

proposes that the originator should be required to hold five per cent of each issue.13 It

10
More detailed discussion is in Cohen and Goldstein (2009).
11
It means compensating their staffs appropriately. The problem of bloodhounds and greyhounds is a
perennial: the greyhounds (financial market participants) run very fast while the bloodhounds (their
regulators) struggle to stay on the trail. But a starvation diet does not help the bloodhounds keep pace.
12
Confidentiality should be ensured, but this should not be something to lose sleep over insofar as reporting
will be to the regulators, who can aggregate the information before releasing it.
13
See U.S. Treasury (2009a).

6
can be argued that five per cent is not enough – that the originator’s skin should be

thicker. The Committee on Capital Markets Regulation argues the other way: it warns

that such measures may result in greater concentration of risk for financial institutions.14

But , if so, then they should be accompanied by higher capital requirements. This may

make originating such securities more costly, but such as the price of a stable financial

system.

Further, regulators should move over-the-counter derivatives transactions into a

clearinghouse where there exists a central counterparty. The central counterparty is the

single buyer and seller to every other party. It collects margins on every trade and places

them into a reserve fund to protect against losses if other parties default. Traders only

have to worry about the creditworthiness of this one counterparty. If any single trader

goes under, this is unlikely to create the kind of domino effect that resulted from the

failure of Lehman Brothers. Some derivative securities (such as interest rate swaps) are

already central-counterparty cleared by LCH.Clearnet Ltd. Credit default swaps based on

indices have been cleared by Intercontinental Exchange’s CCP service since March. The

authorities in the U.S. and Europe have announced the intention of further expanding

their use of central counterparties for clearing over-the-counter derivative products.

Better still would be to move transactions onto an organized exchange whose members

commit their capital jointly and separately to guarantee transactions.15

14
Committee on Capital Markets Regulation (2009), p.22.
15
Related to the preceding, the authorities should encourage standardized derivative instruments that lend
themselves to centralized clearing and exchange-based trading. Bespoke instruments, being one of a kind,
are necessarily bought and sold over the counter. Banning such instruments, which would deny issuers the
ability to ensure themselves against idiosyncratic risks, might be a bridge too far. But the associated
externality – that securities traded over-the-counter pose greater risks to systemic stability – should be
internalized by holding investors in such instruments to higher capital charges.

7
 Fifth, the compensation practices should be reformed to better align the interests

of key decision makers and shareholders. Officials are reluctant to interfere in these

intimate private sector matters, preferring to leave compensation decisions to the

consenting adults involved. But experience suggests that corporate boards do not always

have the independence and outside directors do not have the information needed to take

decisions appropriately incentivizing those they oversee. Bonuses linked to the

performance of the trading desk, division or firm in the most recent year can encourage

practices that inflate current returns at the expense of the long-term viability of the firm.

A compromise would be to leave decisions regarding overall compensation to the firm

but require that bonuses be linked to medium- rather than short-term performance and

mandate that they be clawed back in the event of subsequent problems.16

Sixth and finally, there should be an agency responsible for macro-prudential

supervision – for the stability of the financial system as opposed to just its constituent

parts.17 The Obama Administration would create a Financial Services Oversight Council

chaired by the Treasury (and with its own permanent professional staff) to identify

potential risks to systemic stability and give the Fed regulatory power over individual

systemically important financial institutions. The Brown Government would vest the

Financial Services Authority, the UK’s unified regulator, with this responsibility. The

16
This problem of compensation practices points to larger problems with the corporate governance of large
financial institutions. Fixing these problems is not straightforward: strengthening the fiduciary
responsibility of directors would more effectively incentivize existing board members but discourage
qualified individuals from serving. One desirable reform would be more independence for the risk
management function. The chief risk officer should be required to report directly to the board of directors
as opposed to the CEO, and his compensation should be tied to the stability of the firm and not simply its
profits. Buiter (2009) recommends subjecting all new board members to a written test, set by the regulator
and marked by independent experts, on the products, services and instruments traded and managed by their
financial institutions, to guard against the danger that directors are inadequately knowledgeable of the
business they oversee.
17
The latter being known, for present purposes, as micro-prudential supervision.

8
EU would create a European Systemic Risk Council made up of European central banks

and regulators and chaired by the president of the ECB.18 Other countries have yet to

fully specify how they would organize macro-prudential supervision.

Challenges at the National Level. Implementing these principles at the national

level is not straightforward. To start, moving to higher capital requirements creates a

transition problem. Assuming that there is no appetite for a taxpayer gift to the banks, it

implies a long period of subdued lending as banks limit the growth of the denominator of

the capital-to-asset ratio. Then there are practical issues of how to reform other

dimensions of capital adequacy requirements. While it may be straightforward to index

capital adequacy to the size of a financial institution, size really is of concern only insofar

as it has implications for systemic stability. Size in and of itself is a poor proxy for the

threat to systemic stability posed by the failure of a particular institution; the variables of

interest are systemic importance or connectedness. And it is not clear that regulators

know how to measure these things.19

Indexing capital ratios to the cycle may similarly be less than straightforward.

Should these be indexed to the growth of GDP or the growth of bank lending? There

may be agreement that internal models and commercial credit ratings are weak bases on

which to assess the riskiness of a bank’s investments for purposes of capital adequacy,

but it is not clear what to replace them with. The current fashion is to supplement

existing measures with a simple leverage ratio (the ratio of capital to unweighted assets).

18
More on this below.
19
Avinash Persaud (2009) has suggested relating capital requirements to cross-institution correlations
(whether a bank holds the same assets as other banks and may be inclined or forced to sell them at the same
time, posing a threat to the stability of the system). Do regulators in fact know how to implement such a
complex capital adequacy regime? The U.S. Treasury evidently proposes to place financial institutions into
a couple of categories by size and connectedness, requiring so-called Tier 1 financial institutions to hold
more capital (U.S. Treasury 2009b); in principle one would want a more nuanced categorization.

9
But the latter ignores entirely the correlations between the returns on different assets. It

thus implies very different degrees of self-protection for different financial institutions.20

A special resolution regime will send a clear signal to banks’ unsecured creditors

that they too are at risk in the event of insolvency – that their claims will be wiped out if

the institution is dismantled and that they will be liquidated or converted from debt to

equity if it is restructured and allowed to continue. But this will make it more difficult

for banks to access debt finance precisely when capital is scarce.21 Then there is the

question of to what entities exactly the special resolution regime should apply: insurance

companies, hedge funds, finance companies and credit unions as well as banks, for

example?22 Finally there is the question of who in a fragmented regulatory system

activates the special resolution regime. The Insurance Commissioner of the State of New

York regulated AIG’s investment activities, the Pennsylvania Insurance Department its

property and casualty businesses, and the Delaware Insurance Department its life

insurance business. Which would be responsible for determining that the company had

failed to meet the threshold conditions for financial solvency?23

To ensure adequate information on hedge-fund operations, U.S. regulators are

moving toward a system where all hedge funds register with the authorities. Under the

Treasury’s July 15, 2009 proposal, hedge funds, private equity funds and venture firms
20
By way of example, these questions are all implicit in U.S. Treasury Secretary Geithner’s statement of
principles for reforming the capital adequacy regime in the United States and globally (U.S. Treasury
2009b), but they remain unanswered.
21
There is also an issue of fairness insofar as bondholders purchased the bonds of bank holding companies
in the expectation that they would be protected by the provisions of the currently applicable bankruptcy
code.
22
My own answer is “all of the above” if they are systemically significant.
23
This problem could in principle be solved by establishing a single consolidated regulator, but in the U.S.
at least this does not appear to be in the cards. The Obama Administration’s White Paper (U.S. Treasury
2009a) would have the Treasury Department invoke this authority after consulting with the President and
the relevant regulators. Cohen and Goldstein recommend that the decision to activate should be vested in
the Treasury on the written recommendation of two thirds of the members of the Federal Reserve and FDIC
boards.

10
with more than $30 million in assets would be required to register with the SEC and to

report assets, leverage, off-balance-sheet holdings, and other investments on a

confidential basis.24 The SEC would possess power of examination and the right to share

this information with other regulators. But the form and content of these “periodic”

reports remains to be specified. And the speed with which hedge funds trade raises

questions about the timeliness of the information. Information on last week’s hedge fund

portfolio is at best only marginally more informative than no information at all.25

Neither is requiring underwriters to retain a portion of any security they issue as

straightforward as it may seem. Financial engineers would be quick to identify other

securities that are correlated with the issue in question and short them as a way of

hedging out the position and its implications for behavior. Long positions in other

investments whose returns are negatively correlated with the issue could be used to the

same end. It is hard to imagine that requiring the originator to hold onto a fraction of the

issue would have much impact on behavior when the entire portfolio was taken into

account.

And even ignoring the scope for hedging, it is not clear that requiring a bank to

hold onto 5 per cent of the issue, as proposed by the Obama Administration, would have

much impact on its underwriting activities. Investment banks in fact had been forced to

warehouse more that 5 per cent of the subprime related securities issued in the late stages

of the housing boom, and this did not deter them from originating them as fast as they

24
Some hedge funds are already required to register with the SEC. The existing loophole is for private
advisors with fewer than 15 investors that do not proffer general investment advice. But these existing
registration requirements do not come with reporting requirements.
25
And for hedge funds that churn their portfolios rapidly, it may be positively misleading. Whether it is
possible for hedge funds to provide and regulators to process in real time information on funds’ portfolios,
as advocated by Blinder (2009) and Calomiris (2009), is an open question.

11
could. Requiring the underwriter to retain a larger share of the issue might have more

substantial effects, but this could result dangerous portfolio concentrations.26 And even

then, to the extent that underwriting and investment decisions take place in different arms

of the financial group, it is not obvious that the exposure of those responsible for the

latter will impact the behavior of the former.

Forcing transactions in derivative securities through a clearinghouse or exchange

would have costs in terms of instrument diversity, since only a limited number of

securities subject to a critical mass of transactions would be feasible for centralized

clearing and exchange-based trading. Offering customized contracts for insuring against

idiosyncratic risks would become more costly. The central counterparty would also, by

definition, be a locus of systemic risk. In the United States clearinghouses are regulated

by the Fed and the Commodity Futures Trading Commission. For stability purposes they

will have to be regulated closely. The question then becomes how to plan for a

government rescue of the clearinghouse without distorting the incentives of its operators.

Finally there are questions about the effectiveness of macro-prudential oversight.

Recall how under the Obama Administration’s proposal macro-prudential supervision

would be the responsibility of a Financial Services Oversight Council chaired by the

Treasury while macro-prudential regulation of systemically important institutions would

be undertaken by the Fed.27 Making macro-prudential supervision the responsibility of a

council of regulators is a recipe for endless bickering over the existence of risks, the

definition of bubbles, and the assignment of responsibilities. When setting guidelines and

26
If so the appropriate response would be still higher capital requirements. This might make securitization
more costly, but so be it.
27
In the UK there is similarly a debate over where to place ultimate responsibility for macro-prudential
supervision, with the Financial Services Authority (the preference of the current Labour Government) or
the Bank of England (as proposed by the shadow finance minister of the Conservative opposition).

12
identifying risks is the responsibility of one entity but enforcing regulations and

otherwise intervening in the operations of systemically important institutions is the

responsibility of another, can we be confident of adequate coordination?

The alternative of placing both functions within a single institution eliminates this

coordination problem but creates other difficulties. Making the central bank the sole

macro-prudential supervisor and regulator may create a conflict with its price-stability

objective and expose it to unwelcome political scrutiny. But assigning those

responsibilities to an independent supervisor may limit information flows to and from the

central bank and complicate the lender-of-last-resort function.28 It is not clear that there is

a single optimal solution to this assignment problem or even a clear second best.29

Challenges at the Global Level. In a world in which human and financial capital

are mobile, few of these reforms are likely to be feasible and effective without

international cooperation.

This has long been recognized in the case of capital adequacy, given the incentive

for individual jurisdictions to impose less demanding requirements in order to attract

business. The Basel process was designed to address this problem, but its track record is

less than reassuring. The negotiation of Basel II, now shown to have been deeply

deficient, occupied the better part of a decade. Given the recent demonstration of the

costs of financial instability, it is clear that we cannot wait another decade for Basel III.

What has been done? The Basel Committee has agreed to increase trading book

capital requirements to reflect liquidity as well as default risk. It has pledged to submit

proposals by the end of 2009 for countercyclical buffers and provisioning over the cycle.

28
As arguably happened in the UK in the case of Northern Rock.
29
The author, for what it is worth, inclines in the direction of making the central bank the consolidated
macro-prudential supervisor, notwithstanding the associated conflicts and unwanted political attention.

13
The Financial Stability Board has produced recommendations designed to reduce reliance

on cyclical Value-at-Risk-based capital estimates and to supplement risk-based capital

requirements with a leverage ratio. The question is whether these limited steps will be

enough.

For capital adequacy there at least exists an established process. Negotiating rules

for executive compensation will be harder insofar as there is no established international

venue. The Financial Stability Board has agreed on principles for the Governance of

Compensation Systems, but the mere existence of principles does not guarantee that

anything will be done.30 Implementation is by national authorities who will be concerned

with brain drain and institutional flight.

One approach would be to proceed under existing corporate governance

conventions. Compensation rules are properly the domain of corporate boards and

compensation committees. The OECD has promulgated principles for sound corporate

governance and consulted with its members and civil society on the role of governance

failures in the financial crisis. Unfortunately, aside from monitoring the compliance or

otherwise of members, it has no way of enforcing its recommendations.

Similarly, in a world where bank holding companies operate globally, an effective

resolution regime cannot be organized at the national level. Host supervisors and

creditors may discover that all of a group’s liquidity is in another jurisdiction, available

first to creditors there. In turn this can alter the incentives for home regulators to be

forthcoming in their interactions with host supervisors.

In the case of nonfinancial firms there has been some progress at harmonizing

these arrangements. The International Association of Restructuring, Insolvency and

14
Bankruptcy Professionals (INSOL International), Committee J of the International Bar

Association, and UNCITRAL (the United National Commission on International Trade

Law) all have commissioned working papers, organized meetings and created working

groups to encourage informal cooperation, foster the recognition of foreign insolvency

proceedings, and promote model legislation. But the harmonization of special resolution

arrangements for financial institutions has lagged.

Only in the EU has there been agreement on the adoption of a uniform insolvency

law for banks. The Commission’s Credit Institutions Reorganization and Winding-Up

Directive introduced a single entity regime in 2001.31 But even there difficulties remain.

Recall Fortis, the Dutch-Belgian-Luxembourg financial services company hit by the

subprime crisis. In the interest of maintaining existing synergies, the Belgian authorities

agreed to put up 50 per cent of the rescue fund, the Dutch authorities 40 per cent, and the

Luxembourg authorities 10 per cent. But these efforts at cross-border fiscal burden

sharing lasted all of a week after which they collapsed over disputes between the

contributors. Rather than savings the baby, the three national authorities chose to

dismember it. The Dutch government took 100 per cent of the Dutch operations, the

Belgian authorities 100 per cent of the Belgian operations, and the Luxembourg

authorities 100 per cent of the Luxembourg operations. So much for synergies. And not

even this was straightforward. In June 2009 Fortis Bank Netherlands (owned by the

Dutch State) and Fortis Holdings (a Belgian-listed company) ended up in court, litigating

which rump unit was responsible for paying preference shareholders.32

31
See Tucker (2009).
32
In a dispute that looks like it will take several years to resolve.

15
 Turning to securities markets, there is the fact that over-the-counter transactions

can migrate. Banning them in one jurisdiction may only cause them to move to another.

Even heavier capital requirements for institutions with positions in credit default and

interest rate swaps may be ineffective when not all jurisdictions cooperate in applying

them. The solution presumably is to incorporate such measures into the revised Basel

capital accord, though this has yet to be done.

Then there is where to locate clearinghouses and how to avoid unnecessary

proliferation.33 In addition to the two clearinghouses set up in the United States in 2009,

five more have been established or proposed in Europe, and it will not be long before

more follow in Asia. Netting positive and negative exposures is difficult if some CDS

positions of an individual derivatives dealer are cleared through one clearinghouse while

others are cleared through different clearinghouse.34 In addition, clearinghouses may be

tempted to relax collateral standards and reduce guarantee fund contributions in order to

attract business, in which case their stability will be at risk.

One solution is to work on netting across clearinghouses. But this will require

strong standardization of contracts and collateral terms – stronger standardization than in

current proposals. Another solution would be to agree on the location of one or a small

handful of clearinghouses. But not only is this is politically difficult, it puts regulation of

that global clearinghouse in the hands of a particular set of national authorities who may

or may not follow policies congenial to the others.35

33
For more on these issues see Duffie (2009).
34
They are inefficient if different derivatives (credit default swaps, interest rate swaps) are cleared through
separate clearinghouses.
35
Still another idea would be for national authorities to agree on a single clearinghouse to be operated and
backstopped by a multilateral organization such as the IMF. But this would be a radical departure for what
is a fundamentally monetary institution. And it would entail ceding significant national prerogatives to an
international organization.

16
 Then there is what to do about the rating agencies. In July the Obama

Administration proposed new legislation under which rating agencies would have to

register with the SEC and document their internal controls, due diligence and rating

methodologies. The proposed legislation would also prohibit rating agencies from

advising clients whose securities they also rate. The EU has agreed to create a college of

supervisors for each rating agency and committed to adopting legislation for a single

European supervisor for rating agencies by July 2010. That supervisor will presumably

demand action limiting conflicts of interest and otherwise ensuring minimally acceptable

practice. But even if the EU moves to a single supervisor, the international community

will inevitably be left to rely on a college of supervisors, given resistance at the global

level to establishing a single supervisor. Whether U.S., European, Asian, Latin American

and other officials can work together to ensure that the rating agencies adopt sound

methodologies and avoid conflicts of interest, while at the same time reducing the

dependence of the regulatory regime on the ratings they issue, remains to be seen.36

Finally there is how to conduct macro-prudential supervision at the international

level. Financial institutions and markets are international; so too therefore must be the

macro-prudential response. Here again Europe epitomizes the challenges, given the deep

integration of its financial markets. The Commission has endorsed the recommendations

of the de Larosiere Group to establish a European Systemic Risk Council (ESRC) for

macro-prudential supervision to identify risks, sound warnings and issue guidelines for

corrective action.37 The ESRC will function under the Governing Council of the

European System of Central Banks, chaired by the president of the ECB, with the

36
While the Financial Stability Forum analyzed the role of the rating agencies in a 2008 report (FSF 2008),
it did not recommend moving toward a new regulatory regime. This is not promising.
37
On the report of the de Larosiere Group, see de Larosiere (2009).

17
representatives of all 27 EU central banks present along with the presidents of three

supervisory bodies: the European Banking Authority, the European Insurance and

Occupational Pensions Authority, and the European Securities Authority.38

But the ESRC will have no responsibility for regulation, financial supervision, or

monetary and fiscal policies, the three key pillars of macro-prudential supervision. It will

have no ability to change actual policies in these areas, only the ability to communicate

its findings to the competent national authorities. As the de Larosiere Report emphasized,

binding mechanisms are needed to ensure that the macro-prudential findings of these

bodies are followed by micro-prudential supervisors. It is proposed that the ESRC or its

constituents could request the Commission to adopt a decision that there has been a

“manifest breach” of Community law and thereby require a national supervisory

authority to take a specific action in order to come into compliance. In this way the

ESRC and the Commission together would have the power to compel corrective action

on the part of national authorities – assuming of course that national authorities assign

them this power.

What is difficult at the European level is even more difficult globally. The

approach that emerged from the G20 process would put the newly expanded and

reconstituted Financial Stability Board in the pilot’s seat, with the IMF as co-pilot. The

FSB will set guidelines for and to support the establishment of supervisory colleges for

systemically important firms, collaborate with the IMF on early warning exercises, and

38
There would seem to be heavy overrepresentation of central bankers and underrepresentation of
supervisors on the risk council. But it is not clear how to fix this given the presence of 50 some supervisors
in the EU. A further problem is that the lines between insurance, pensions and securities are blurring.
Goodhart and Schoenmaker (2009) suggest moving directly to two institutions a one for banking and one
for securities markets. But then there would be even heavier numerical overrepresentation of central
bankers.

18
undertake strategic reviews of the policy work of international standard setting bodies. It

will do so by establishing Standing Committees on Vulnerabilities Assessment, on

Regulatory and Supervisory Cooperation, and on Standards Implementation and by

expanding its secretariat. As I read it, the idea is that the IMF would report its findings

on financial stability risks to the FSB. The FSB would then identify “enhancements” to

regulatory frameworks to mitigate risks and communicate these to the relevant national

authorities. But it would not have enforcement powers.

Then again, creating a global macro-prudential supervisor with enforcement

powers may not be as infeasible as commonly believed. Imagine creating a World

Financial Organization analogous to the already-existing World Trade Organization

(WTO).39 In the same way that the WTO establishes principles for trade policy

(nondiscrimination, reciprocity, transparency, binding and enforceable commitments)

without specifying outcomes, the WFO would establish principles for prudential

supervision (capital and liquidity requirements, limits on portfolio concentrations and

connected lending, adequacy of risk measurement systems and internal controls) without

prescribing the structure of regulation in detail.40 The WFO would define obligations for

its members; the latter would be obliged to meet international standards for supervision

and regulation of their financial markets and institutions. Membership would be

obligatory for all countries seeking freedom of access to foreign markets for

domestically-chartered financial institutions. The WFO would appoint independent

panels of experts to determine whether countries were in compliance with their

obligations. Importantly, it would authorize sanctions against countries that failed to

39
I have proposed this in Eichengreen (2009c), from where the next couple of paragraphs are drawn.
40
The Basel Core Principles for Effective Banking Supervision would be the obvious place to start when
defining these principles.

19
comply. Other members would be within their rights to restrict the ability of banks and

nonbank financial institutions chartered in the offending country to do business in their

markets. This would provide a real incentive to comply.

It will be objected that national governments will never let an international

organization dictate their domestic regulatory policies. The rebuttal is that the WFO

would not dictate. The specifics of implementation would be left to the individual

country. Members would be able to tailor supervision and regulation to the particularities

of their financial markets. But those regulatory specifics would have to comply with the

broad principles set down in the WFO charter and associated obligations.

We already do the equivalent for trade. Dispute settlement panels already

determine whether, inter alia, U.S. tariffs on timber imports from Canada are in

compliance with the United States’ WTO obligations. If not, we have the choice of

whether to change those laws or face sanctions and retaliation. If the U.S. and other

countries accept this in the case of trade, why should they not accept it for finance?

Global Imbalances and Global Policy Reforms

The other view of the crisis focuses on global imbalances. The run-up in asset

prices and associated financial excesses derived in this view from the combination of

accommodative policy in the United States and large capital inflows from emerging Asia

and petroleum-producing countries. China’s current account surplus rose from less than

2 per cent of GDP early in the decade to a whopping 11 per cent in 2007. Under other

circumstances such large surpluses resulting in foreign-asset accumulation would have

led to real exchange rate appreciation through some combination of nominal appreciation

20
and inflation. But China sterilized the asset accumulation, squirreling it in international

reserves held largely in U.S. treasury and agency securities. Moreover, while China’s

reserves grew rapidly it was not alone: India, South Korea and Taiwan, among others,

also saw very sharp increases in reserves. Oil-exporting countries also ran large current

account surpluses and accumulated considerable quantities of reserves after 2000.

Russia’s reserves rose from negligible levels at the beginning of the decade to nearly

$150 billion at the end of 2007. More generally there were large petrodollar flows from

the Persian Gulf states, Russia, Nigeria and Venezuela toward the United States and other

advanced-country markets.

While China, the rest of emerging Asia, and the oil exporters all had rising

national savings after the turn of the century, their national investment rates also rose.

But those increases were inadequate to absorb the growing pool of savings. Investment

rates in China were in fact extraordinarily high by international standards, and it is not

clear that the country could have deepened its capital stock even more rapidly without

significant inefficiencies.41 Other emerging Asian countries had maintained higher gross

investment rates and kept their aggregate current account broadly in balance prior to the

financial crisis of 1997-8, but some of this investment had been inefficient: it reflected

directed-credit policies of governments resisting the inevitable deceleration from the

high-growth period and empire-building by highly-leveraged family-owned

conglomerates. These problems were corrected, at least in part, following the crisis.

Investment rates in emerging Asia ex China recovered from their post-crisis lows after

the turn of the century but never re-scaled the inefficiently high levels of 1993-6. The oil

41
In a sense, the 2008-9 fiscal stimulus, a considerable fraction of which was devoted to additional
investment, will provide a test of the hypothesis.

21
exporters saw modest increases in their investment rates in 2000 and 2006-7, but not on

the scale needed to absorb their rapidly deepening pool of saving.42

All this money had to go somewhere, and much of it flowed to the United States.

One estimate (Warnock and Warnock 2006) suggests that U.S. treasury rates were 100

basis points lower than they would have been in the absence of foreign inflows. The Fed

raised short-term rates from 1 to 5.25 per cent between mid-2004 and mid-2006, but the

yield on ten-year treasuries actually fell over the period.43 This was Chairman

Greenspan’s “bond market conundrum,” which his successor and others ascribed to the

so-called “global savings glut.” Low real interest rates on ten-year treasuries pushed

capital into other assets. The consequent higher home and stock market valuations had

positive wealth effects on spending. Measured household savings in the U.S. fell from

approximately 10 per cent of disposable income in the first half of the 1980s and 7 per

cent in the early 1990s to close to zero in 2005-2007.44

Spending was further encouraged by U.S. monetary and fiscal policies. While the

Fed raised short-term interest rates in 2004-6, it started from exceptionally low levels,

42
In the Bretton Woods II view, China’s investment in U.S. treasury and agency securities reflected the
inefficiency of its financial system: the Chinese authorities invested (on behalf of their residents) in U.S.
financial assets, and U.S. financial institutions used the resulting liquidity and their superior investment
expertise to channel resources to U.S. corporations which invested directly in China (see Dooley and
Garber 2005). The flaw in this view was always that FDI into China plus domestically-financed investment
fell short of Chinese savings. In other words, there did not exist profitable investment opportunities
sufficient to absorb the pool of Chinese savings, regardless of who did the intermediation. The flaw in the
Bretton Woods II story, in other words, is that while it could explain the two way flow of capital it could
not explain the current account imbalance.
43
From 4.7 to 4.5 per cent.
44
The ratio of total household debt to disposable income rose meanwhile from 80 per cent in the 1990s to
nearly 135 per cent in 2007. It was also argued at the time that increased consumer spending reflected the
belief that productivity growth had accelerated permanently—that household debt could rise now because
of expectations of increased disposable income in the future. The problem with this argument is that it
doesn’t explain why U.S. households chose to leverage in response but the U.S. corporate sector did not,
since higher expected future incomes for households should have had as their counterparts higher expected
future revenues for firms, which would have encouraged them to assume higher debt ratios, which they did
not.

22
reflecting the low rates put in place in response to 9/11 and the 2002-3 deflation scare. It

hesitated to normalize too fast for fear of choking off the expansion. The Federal funds

rate was consistently below Taylor Rule levels between 2002 and 2007.45 Fiscal policy

operated in the same direction. The most reliable way of preventing overheating and

discouraging capital inflows is of course by tightening fiscal policy.46 And, after a long

period of deficits, net borrowing by the U.S. public sector had actually gone negative in

1998-2000.47 But public-sector borrowing resumed in late 2001 with the recession and

the Bush tax cuts. Public sector net borrowing in fact exceeded net borrowing by

households in most quarters between the beginning of 2002 and the end of 2007.

The effect of all of this was that the U.S. current account deficit reached 6 per

cent of GDP in 2006.48 Slightly more convoluted is the link to the particular constellation

of financial weaknesses that culminated in the crisis. The story as typically told goes like

this.49 Low yields on ten-year treasuries encouraged money to flow into higher-yielding

assets backed by, inter alia, residential mortgages. Mortgage originations as a share of

total mortgage debt outstanding thus rose from 6 per cent in 1985-2000 to 10 per cent in

2001-2006. To meet the demand for mortgage-back securities, lending standards for

residential mortgages were relaxed. Agency problems between mortgage brokers who

originated the loans, financial institutions who packaged and distributed them, and

investors who purchased them allowed this problem to go uncorrected.50 The income

45
See Taylor (2009).
46
This is the advice that the U.S. Treasury, among others, has regularly given emerging markets over the
years.
47
Recall contemporary fears of a shortage of marketable U.S. treasury securities and questions, which now
seem quaint, about how monetary policy might be conducted in their absence.
48
Note that the U.S. was not alone in seeing its current account deficit widen: similar trends were evident
in, inter alia, Australia, New Zealand, the UK and Spain, among others.
49
My favorite rendition is Goldman Sachs (2009).
50
Or at least to remain uncorrected.

23
streams associated with those mortgages were then sliced, diced, repackaged and insured

to render them compatible with the covenants and capital requirements of institutional

investors.

A variety of feedbacks amplified these dynamics. Lower lending standards and

easy mortgage finance pushed housing prices even higher, which encouraged further

reductions in standards by lenders impressed by the increase in the value of real-estate

collateral. The increase in mortgage activity encouraged entry by brokers and squeezed

margins, further aggravating agency problems at origination. Higher asset prices

encouraged larger flows into U.S. financial markets from domestic and foreign investors

convinced that past performance was a guide to future returns. Higher asset prices also

meant more revenues for state and local governments that depend on capital gains and

property taxes, respectively, for much of their income; these ramped up their spending

accordingly. All these were typical responses to a surge of capital inflows. The only

difference from earlier capital flow bonanzas was that this time the country on the

receiving end was the United States.

Starting in 2007 these same feedbacks shifted into reverse. Private foreign

demand for additional U.S. portfolio investments disappeared in the early part of 2007.

The result was a weaker dollar and tighter U.S. financial conditions. U.S. housing prices,

having reached historically high levels, had already been in decline since the summer of

2006, causing delinquencies, starting in the subprime segment of the market, to rise

sharply.51 This created problems for debt securities backed by claims on pools of

mortgages and, in turn, for institutions like American Home Mortgage Investment

Corporation, BNP Paribas and Countrywide Credit heavily involved with them. They had
51
The peak in housing prices was already in 2005 according to the Case-Shiller 10-city index.

24
to sell other assets in order to square their books, raise liquidity, and meet shareholder

redemptions. The more highly leveraged the institution, the more extensive the resulting

fire sales. The U.S. economy as a whole and its financial sector in particular having

become more leveraged during the expansion, deleveraging now was dramatic. Banks,

under balance-sheet stress and seeing the value of collateral eroding, raised lending

standards, putting further downward pressure on housing prices.52 Private foreign

investors who had previously bought U.S. financial assets with wild abandon now

withdrew from the market en masse. The story of the crisis can be told more colorfully

and in more detail. But from the point of view of our second explanation emphasizing

global imbalances, this was the essence of the matter.

Policy Implications. From this diagnosis again flows a prescription for policy

responses. As in the first half of the paper, I highlight six.

First, monetary policy makers must worry about imbalances even in the absence

of inflation. The first half of the present decade was notable for the absence of overt

inflation in the United States. But even if low interest rates did not spur commodity-price

inflation, they fed asset-price inflation through the various channels enumerated above.

They encouraged the financial excesses that set the stage for the crisis. It follows from

their commitment to the maintenance of macroeconomic stability that monetary policy

makers should lean against unsustainable asset- as well as commodity-price

developments. Regulation alone, no matter how comprehensively reformed, cannot be

relied on to prevent unsustainable asset-market conditions or to fully contain their

consequences. The textbook view of inflation targeting in which asset-market

developments matter only insofar as they convey information about prospective future
52
And on economic activity generally.

25
commodity-price inflation should be abandoned or at least modified to admit an

independent role for asset-market conditions.

Second, policy makers in the U.S. and elsewhere should attend to the procyclical

bias in fiscal policy. The budget deficit and net borrowing by the public sector were

allowed to explode in 2003-6 precisely when the U.S. economy was growing strongly

and the current account deficit was widening. These were years when the economy did

not need a shot in the arm from deficit spending. Fiscal policy makers would have done

better to keep their powder dry. Better still would have been to have taken the same

advice that they regularly doled out to emerging markets – that tightening fiscal policy is

the best way of moderating the impact of large capital inflows and, indeed, of moderating

those inflows themselves.

Third, a large current account deficit cannot be regarded as benign even by a

country like the United States that borrows in its own currency. Foreign finance for the

current account can dry up abruptly. The U.S. has not felt the full effects this time,

foreign central banks having stepped in to replace much of the foreign private investment

that evaporated, but it may not be so lucky next time.53 A current account deficit that

cannot be financed will necessarily be compressed; unlike other deficits, it cannot then be

financed at home. Such compression in turn requires a change in relative prices,

including in the exchange rate, which can catch investors wrong footed. It can also

precipitate a recession if it takes time to shift resources between the production of

nontraded and traded goods. The same arguments leading to the conclusion that

monetary and fiscal policy makers cannot afford to disregard asset-market developments

similarly suggest that they cannot afford to disregard the current account.
53
Nor may other countries.

26
 Fourth, countries equally cannot regard large current account surpluses with

equanimity. If someone else’s current account deficit puts financial stability at risk, so

too by implication does your surplus, since the former is not possible without the latter.

Countries where domestic saving exceeds domestic investment by a large margin have

tools with which to boost spending, from increasing public spending directly to easing

credit terms for household and corporate spending.

Fifth, relative prices must adjust to accommodate these changes in the pattern of

demand. Insofar as the residents of each country exhibit home bias in consumption, the

relative price of home goods will have to rise in the surplus country and fall in the deficit

country. Ruling out deflation – which policy makers in the deficit country will work hard

to avoid – this adjustment will have to occur through some combination of inflation and

currency appreciation in the surplus country. For well known reasons, currency

appreciation is the preferable alternative.

Sixth, reserve accumulation will have to be less insofar as surplus countries

encourage domestic absorption and see their real exchange rates appreciate in response.

Such countries will have to seek other ways of insuring themselves against shocks.

It is worth asking how history would have differed had these recommendations

been adopted at the beginning of the decade. U.S. monetary and fiscal policies would

have been tighter.54 Chinese and more generally Asian fiscal policies would have been

looser. Global imbalances would have been less. Less accommodating monetary policy

54
My friends at the Fed will no doubt object that a tighter monetary policy in 2002-3 would have exposed
the U.S. economy to very serious danger of deflation. Perhaps, but this does nothing to weaken the
argument that monetary policy should have been tightened more aggressively starting in 2004 in response
to the housing bubble and other evidence of financial excesses – more so insofar as fiscal and regulatory
policies were not doing their parts.

27
and less capital inflow would have dampened financial excesses in the United States.55

The combination of less demand stimulus in the United States and more demand stimulus

abroad would have left global demand unchanged to a first approximation. To the extent

that there were worries about overheating in China and elsewhere, foreign stimulus might

not have been increased sufficiently to offset the reduction in demand stimulus in the

United States, and the global economy would have grown more slowly. Recall however

that global growth in 2005-6 was the fastest in more than 30 years. Slightly slower

growth would have been an acceptable price to pay for warding off the most serious

financial crisis in generations.

Challenges at the National Level. Challenges again arise when attempt to move

from principle to practice. It is easy to say that inflation targeting should be modified to

admit a role for asset market conditions but harder to know exactly how to modify it. For

example, it is easy to assert with benefit of hindsight that monetary policy should have

been tightened faster in 2004-6 in response to the rise in housing prices and widening of

the current account deficit, but it is more difficult to say by how much. By exactly how

much do asset prices and the current account have to move before they justify a monetary

policy response over and above that warranted by their implications for expected future

inflation and the output gap? The presumption in the debate over whether central banks

should target asset-market conditions may have tipped away from the Jackson Hole view

in favor of the BIS view, but earlier questions about the BIS view – starting with how

central banks know when asset prices pose a significant threat to stability – have not gone

away. If the effect of asset prices on the economy is complex, nonlinear and contingent,

then the monetary policy response to asset price fluctuations will have to be complex,
55
This, recall, is the premise of the second half of the paper.

28
nonlinear and contingent.56 Attempts to routinize monetary policy in the form of a post-

Taylor Rule are unlikely to succeed.

Similarly, inadequate fiscal restraint in good times is an old problem with no

simple solution. Institutional reforms can help, but effective reform will depend on

circumstances. In general, fiscal rules that limit deficit spending (but also limit fiscal

flexibility) work best where ideological differences between political parties are

relatively pronounced, while fiscal procedures that delegate decision making to, inter alia,

the executive work best where ideological polarization is relatively limited.57 In a

presidential system like that of the United States, it may be necessary to have both

supportive rules and procedures. Given the electoral returns to pork-barrel spending it is

important to have powerful party leaders and a strong committee system to exercise

agenda-setting powers and discipline members of Congress. At the same time, when the

balance of power between the legislative and executive branches is relatively even – also

the U.S. case – ex ante agreements (balanced budget rules, multi-year fiscal targets) can

be critical for fiscal discipline. But reform to give party whips and committee chairs even

more power would be strongly resisted. And the unhappy record of the Gramm-Rudman-

Hollings legislation and its successors does not reassure one about the prospects for

binding fiscal rules.

Countries with large current account surpluses may similarly encounter

difficulties when attempting to narrow them in short order. In China, significant

reductions in household savings rates will require a stronger social safety net, something

that cannot be built overnight. Government saving and investment can be adjusted more

56
A point that is not original to me; see for example Visco (2009).
57
See Hallerberg, Strauch and von Hagen (2009).

29
quickly, but here too there are limits. For example, doubts have been voiced about the

efficiency and productivity of the additional investment spending undertaken by China’s

local governments since the outbreak of the crisis.58 Such questions are likely to deter the

central authorities from relying yet further on expansionary fiscal policy to address the

current account imbalance.

Challenges at the Global Level. What to do about global imbalances may be the

thorniest question of all. In the years leading up to the crisis, both the United States and

China followed the policies they did, despite voices warning of risks, because they

perceived them as in their self interest.59 It could be that they misperceived the balance

of risks and rewards and that more effective advice could have alerted them to their error

and prompted corrective action. It could be that they failed to understand the impact of

their policies on other countries and that more effective consultation would have caused

them to recognize the existence of these spillovers and, good global citizens they are, to

internalize them. It could be that there existed policy adjustments that would have been

mutually beneficial if taken in tandem even though either would have been welfare

reducing for the country concerned if taken in isolation. It could be, in other words, that

what was needed was more effective policy coordination.

The problem is that there already exist mechanisms for correcting these

deficiencies. Warning of the risks posed by large current account deficits is at the heart of

the IMF’s country surveillance. The Fund issued warnings about the danger that chronic

large U.S. current account deficits could result in a disorderly adjustment, but these led to

58
See Shih (2009).
59
Here “China” is shorthand for surplus countries generally. “The United States” is, similarly, a stand-in
for deficit countries generally, although the fact that United States accounted for the vast majority of global
current account deficits in the years leading up to the crisis means that this shorthand does little violence to
the facts.

30
no changes in U.S. policy. It expressed reservations about the constellation of policies

that resulted in large and growing Chinese surpluses but again without noticeable results.

Similarly, the IMF’s multilateral surveillance is designed to alert countries to the

spillovers and external effects of their policies. Instruments here include the IMF’s two

flagship reports, the World Economic Outlook and Global Financial Stability Report, its

Regional Outlooks, its contributions to interregional committees and forums (the G7/8,

G10, G20 etc.), and confidential briefings on internal evaluations like those undertaken

by the Coordinating Group on Exchange Rate Issues. The spring and autumn 2005 WEOs

devoted considerable space to the factors underlying global imbalances, the risks, and the

appropriate policy response, again without discernible results.

Since 2006 the IMF’s arsenal has included a Multilateral Consultations Initiative

bringing together a small number of countries for whom such spillovers are first order.

The first such consultation, announced on June 2006 and concluded with an Executive

Board discussion in July 2007 (note the date), brought together the United States, China,

Japan, the Euro Area and Saudi Arabia to discuss the cross-border impacts of global

imbalances. This consultation started with bilateral staff visits with the five participants

followed by multilateral meetings and a joint report. The report mentioned how the

process had been “useful” and how it had “contributed to an improved understanding of

the issues and each other’s positions.”60 Again, however, it is hard to see that this useful

initiative and improved understanding led to significant adjustments in the policies of the

countries in question.

60
The quotes are from the Public Information Notice summarizing the Executive Board discussion (IMF
2007).

31
 Finally, if the problem is to coordinate policy adjustments that are unappealing in

isolation but mutually beneficial if undertaken jointly, then there already are mechanisms

for achieving this. There is the aforementioned Multilateral Consultations Initiative.

There are country grouping ranging from the G7/8 to the G20. There are bilateral

consultations among governments such as the annual U.S.-China Strategic Dialogue. If

there is a shortage of coordinated action, it is not for a shortage of venues for

coordination.

Why these processes did not lead to different outcomes is familiar enough.

While the IMF can issue warnings, it cannot compel policy adjustments by countries that

do not borrow from it, either because they have no difficulty borrowing on the market

(the U.S. case in the period leading up to the crisis) or because on net they do not borrow

at all (the Chinese case). Louder warnings might be more likely to elicit action, but these

would be problematic on a number of grounds. IMF staff and management operate under

the oversight of the Executive Board, which speaks for the governments about whose

policies they are warning. Large shareholders could push back against warnings that

cause them significant embarrassment, through actions in the Board that make

management’s position untenable. Staff and management know this and choose their

language accordingly.

Likewise, the notion that good global citizens should internalize the cross-border

spillovers of their policies and that difficult policy adjustments may be easier when

coordinated internationally runs up against domestic political constraints. The

Multilateral Consultation on global imbalances resulted in statements by the United

States that it would pursue fiscal consolidation and from China that it would encourage

32
domestic spending. But for the U.S. meaningful fiscal consolidation would have meant

tax increases, which were a nonstarter politically. For China ramping up domestic

spending more rapidly (which, in practice, would have meant ramping up public

spending) would have meant ramping down something else given that the economy was

operating close to capacity. That something else would have been exports, which would

have antagonized politically influential export interests.61 For other Asian countries, it

would have meant forgoing the reserve accumulation seen as the first line of defense

against financial instability.

The familiar responses to these problems go as follows. The IMF needs to devote

more resources to surveillance – and so it has gone on a hiring binge since the crisis. It

needs to develop better early warning indicators – notwithstanding the fundamental

difficulties of crisis prediction and the failure of all concerned to predict the last one.62

Governments should take the results of such surveillance and early-warning exercises

more seriously – despite their manifest reluctance to do so over the years. They should

behave more cooperatively.

In addition to relying on clearer crystal balls and better behavior, it might be

worthwhile to think about more ambitious reforms. None of the initiatives I am about to

describe will happen overnight. The political obstacles are formidable. But if one takes

seriously the risks posed by global imbalances, they are worth contemplating.

61
This adjustment to prevent the economy from overheating would have been achieved by allowing the
currency to appreciate. Of course, there was no such currency adjustment after November 2008 when
Chinese public spending was ramped up, but then there was no longer a danger of the economy overheating,
export demand having collapsed. And given that increased public spending no longer threatened to crowd
out exports, given that growth had slowed relative to capacity, opposition to increased public expenditure
was less.
62
My favorite statement of the limitations of such early warning exercises is Eichengreen and Rose (1999).
A recent analysis attempting to predict the incidence of the 2008-9 crisis and link it to causes – reaching
essentially the same conclusion – is Rose and Spiegel (2009).

33
 One option would be strengthen IMF surveillance by giving greater independence

to those vested with the surveillance function. The IMF department responsible for

surveillance would function independently of management and the Board. Firewalls

would separate surveillance from other IMF functions like emergency lending. The

surveillance unit would have its own budget. It would be overseen by a director

appointed to a single long term in office. It could issue reports without the prior approval

of management or the Board. It would be able to call a spade a spade. The IMF has

adopted this kind of structure for its Independent Evaluation Office, which is independent

of management, operates at arm’s length from the Board, and is overseen by a director

serving a single six-year term. The UK government in 2003 proposed this kind of

structure for the surveillance function.63

But what we have learned about the effectiveness of Chinese Walls in other

financial institutions gives grounds for questioning whether they would be effective in

this context. Staff will be moving back and forth between the surveillance unit and other

departments.64 Can they really be expected to ignore the preferences of management and

the Board? Can they really be expected to disregard the ability of management and the

Board to shape their career prospects in other departments? The crisis has also alerted us

to the kind of problems that can arise when the monitoring function is allocated to one

entity and the lending function to another.65 Effective firewalls and seamless information

sharing do not go hand in hand.

63
See Balls (2003).
64
Certainly this is the case if the experience of the IEO is any guide.
65
Can you say “Northern Rock?”

34
 A stronger alternative to imagining that a unit within the IMF can be made more

independent is to make the entire institution more independent.66 Members of the

management team would service long terms. They would not have to seek the approval of

an Executive Board of political appointees. They could issue strong surveillance

statements. At the same time, surveillance could be adequately coordinated with other

functions. An independent management team could react quickly to the events, in the

manner of national central banks. They could adopt innovative tactics and instruments,

much like central banks in the recent crisis.

For such strong independence to be acceptable, management would have to be

strongly accountable for their actions. They would have to be more transparent about

their decisions and their criteria for taking them. One could imagine publication of

minutes of their deliberations.67 One can imagine the Managing Director holding press

conferences summarizing the management team’s decisions, much like the president of

the ECB.

Management would have to be strongly accountable to the International Monetary

and Financial Committee (IMFC), the oversight committee of 24 officials whose

composition mirrors that of the Executive Board. Much as the president and monetary

policy committee of a central bank are accountable to their national parliament or

congress, IMF management would have to justify their actions to the IMFC. They should

be subject to a formal vote of no confidence. The IMFC, for its part, would be

66
I will be excused, I hope, for repeating this argument, having made it now for fully a decade. See
DeGregorio, Eichengreen, Ito and Wyplosz (1999). The excuse for repeating it is that the case is, if
anything, stronger than ever in the wake of the crisis. I argue this in Eichengreen (2009c), from where this
paper’s material on this subject is drawn.
67
In the manner of the Fed. This would be a small step technically, since minutes of Board meetings are
already kept and a highly sanitized version is published as the conclusions of the chair.

35
accountable to the Board of Governors of the IMF. The IMFC would have to be

reconstituted as the IMF Council, as provided for under the Articles of Agreement, so

that what are now recommendations become binding instructions to management.

It would of course be necessary to abolish the convention that the managing

director should be a European and his first deputy an American. Leadership selection

would have to reward the most qualified candidates.68 It would be necessary to devise a

selection mechanism for the entire management team that both picked out high-quality

candidates and ensured a reasonable degree of geographic and economic diversity.

For some, delegating these sensitive functions to independent technocrats would

be a bridge too far.69 If so, the alternative to delegation is rules. The other way of

insulating surveillance from politics, in other words, is by mechanizing it. If chronic

surpluses and deficits pose a threat to systemic stability, then another way of applying

stronger pressure to correct them would be by automatically levying penalties on the

countries running them.70 A country that had run a current account surplus or deficit in

68
As G20 finance ministers reportedly agreed at their mid-March 2009 meeting in Sussex, England.
69
The objection to schemes of this sort is that the decisions of the IMF are more complex and therefore
entail more discretion than those of a central bank and that they require the Fund to put taxpayer money
more directly at risk. Since a central bank just sets interest rates rather than applying detailed prescriptions
for changes in the fabric of social and economic policy, it is said, independence for its monetary policy
committee is politically tolerable. Since it just sets interest rates, an action which is easily monitored and
assessed, holding its independent management accountable for their actions is relatively straightforward.
And since central banks accept only high-quality collateral in their lending operations, they do not put
serious taxpayer money at risk (typically, in contrast, they are a profit center). In the wake of the crisis it is
clear that none of these objections hold water. We have seen national central banks engage in very detailed
interventions in financial and other markets. They have purchased all manner of collateral as required by
policies of credit easing, exposing themselves to significant balance-sheet losses. The reality is that modern
central banks, not unlike the IMF, are required to do much more than just set interest rates. This has
created some discomfort among observers and demands that central bankers do a better job at justifying
their actions; it has similarly created pressure that mechanisms for holding them accountable, be these
oversight committees of or appointed by the U.S. Congress or the relevant committees of the European
Parliament, be strengthened. Ron Paul notwithstanding, it has not given rise to the view that central bank
independence is intolerable or, for that matter, undesirable.
70
I made this case for chronic surplus countries in Eichengreen (2009c); here I generalize the argument to
deficit countries.

36
excess of 3 per cent of GDP for three years, for example, might be required to transfer

resources to the Fund at the end of every year in which that excess persisted.71 The

transfer might equal one half of the current account balance in excess of 3 per cent of

GDP.72 Nothing would prevent countries from running large and persistent external

surpluses and deficits if they found it difficult and costly to adjust saving and investment,

but their doing so would entail an additional cost, in turn ratcheting up the pressure to

adopt policies of adjustment.73 This tax could be written into the Articles of Agreement

so that collecting it would not require, inter alia, a decision by the Executive Board.74

A problem with a symmetrical scheme of this sort is that deficit countries may

lack resources to transfer to the Fund. They will be losing reserves rather than gaining

them. But they will be subject to market discipline. The same is not true of surplus

countries that feel no direct pressure from the market to adjust. This asymmetry was

what motivated the decision to include a scarce currency clause in the IMF’s original

Articles of Agreement so that other countries could apply pressure for chronic surplus

countries to adjust. If the present measure was applied to surplus countries alone, it could

be thought of as a price-based scarce currency clause.

To the extent that surplus countries are motivated by the desire to accumulate

reserves, a tax requiring them to transfer dollars or the equivalent to the IMF could
71
The particular thresholds mentioned in the text are purely illustrative; readers are free to substitute their
own. Note that nothing requires that the tax revenues be paid in to the Fund. They equally well might go
to the World Bank for development assistance or the United Nations for peacekeeping operations.
72
Or the charge might initially be set at a lower level and raised to, say, 50 per cent over time (as members
who wished to minimize it had more time to adjust). More recently Prasad (2009) has suggested that such
a tax might be applied to countries’ holdings of Special Drawing Rightst at the IMF and would be levied if
a country failed to hit its target for its current account (and fiscal) balance over a three-year horizon.
73
Economists not liking tax schedules with discontinuities, one can imagine a tax on all increases in foreign
reserves that started at infinitesimal levels but rose fairly quickly as the increase in reserves rose as a share
of GDP and as a function of its persistence.
74
It would presumably be easiest to implement in a period when it was not so obvious on which members it
would predominantly fall. It would presumably be easiest to implement in a period when it was not so
obvious on which members it would predominantly fall.

37
conceivably have the perverse effect of encouraging them to undervalue their currencies

still more so that they could replace the reserves that they had been forced by the

provision to transfer to the Fund. In other words, the tax would have a relative price

effect and an income effect working in opposite directions.75 Thus, marrying the current

measure to other sources of emergency liquidity besides own reserves would make it

more effective. More generally, global reforms enhancing those other sources would

mitigate the tendency for countries to run chronic surpluses.

The most obvious source of emergency liquidity is, of course, the IMF itself. The

Fund’s original raison d’etre was to act as a reserve pooling arrangement. It has recently

received a considerable increase in the resources that it can deploy in emergency lending.

It has streamlined its procedures for deploying them.76 It has established a Special

Liquidity Facility for making substantial loans of reserves without conditions to countries

with strong policies. Quota reform has begun to better align voice in the institution with

21st century realities. Yet no Asian country has requested eligibility for the Special

Liquidity Facility. Other conference participants will have to explain what further

reforms would make it politically acceptable for an Asian government to again borrow

from the IMF.

The alternative is regional reserve pooling. Different countries being hit by shocks

at different times, the timing of reserve needs will differ as well.77 The same quantity of

75
The desire to accumulate reserves is only one reason, of course, why some countries are inclined to
maintain highly competitive real exchange rates and run chronic external surpluses. Rodrik (2008) argues
for example that so-called undervalued exchange rates are associated with rapid economic growth because
they encourage manufacturing employment. To the extent that these other motives prevailed, the perverse
“income effect” would not dominate.
76
I cannot resist observing that an independent IMF could react to events even more quickly.
77
Insofar as shocks have a strong regional component – different countries in a region tend to suffer them
at the same time – regional reserve pooling is second best to global reserve pooling. On the regional
dimension of crises see Glick and Rose (1998).

38
reserves can go further if pooled, and effective pooling will reduce the pressure to run

large surpluses in order to accumulate more. It will also minimize the other costs of

reserve accumulation which range from the risk of capital losses on foreign currency

holdings to foregoing higher levels of consumption and investment.

The Chiang Mai Initiative, now the Chiang Mai Initiative Multilateralization

(CMIM), is the most highly developed example. In May 2009 ASEAN+3 finance

ministers agreed to transform their $120 billion of bilateral swaps and credits into a

reserve pool. Operational decisions will be by simple majority, where countries will have

votes roughly in proportion to their contributions. China and Japan will both contribute

32 per cent, Korea 16 per cent, ASEAN the remainder. The agreement also included a

commitment to establish a regional surveillance unit, although there is no consensus on

where to situate it or how to staff it.

But disbursing more than 20 per cent of the credits available to a country still

requires that it reach an agreement with the IMF. 20 per cent of a country’s entitlement is

actually less than it contributes to the pool. This nullifies the purpose of the arrangement,

which is to provide an alternative to the IMF. While there is a plan to first raise and then

eliminate the 20 per cent threshold, this is left for some unspecified date.

The reason is straightforward. Countries want assurances that their resources will

not be used frivolously. They want to know that they will be repaid. But regional

neighbors find it hard to criticize one another’s policies and demand adjustment. Political

sensitivities run high in Asia. But even in Europe, with its long history of cooperation,

surveillance and conditionality are outsourced to the IMF. Revealingly, the Fund and not

39
the EU has taken the lead in negotiating emergency assistance packages for Hungary and

Latvia.

Delinking the CMIM from the IMF will require Asian countries to undertake

hard-hitting reviews and demand difficult policy adjustments. One solution, again, would

be to give both surveillance responsibilities and the actual power to disburse funds to an

independent board. Its members, enjoying statutory independence and long terms in

office, could function like the monetary policy committee of a central bank. They could

issue a Financial Stability Report that bluntly flags weak policies and vulnerabilities.

And they could demand policy adjustments as a condition for disbursing funds. The

CMIM could then be delinked from the IMF.

Then there are a variety of proposals for reforming the international monetary

system. The current system already includes the one feature that is most useful for

correcting imbalances, namely exchange rate flexibility. This permits surplus countries

increasing spending to raise the relative price of locally produced goods without suffering

inflation and deficit countries doing the opposite to avoid significant deflation. It is all to

the good that we are unlikely to see changes in this exchange rate system as a result of the

crisis.

The other relevant aspect of the international monetary system is the supply of

international reserves. Here one encounters a variety of proposals for replacing the dollar

with another unit. These are based on the argument that allowing a national currency to

constitute the dominant share of international reserves requires the country issuing it to

run the current account deficits that are at the root of the imbalances problem.78 It is

78
There are also other arguments, such as the desirability of substituting another unit, say the SDR, for
existing dollar holdings to relieve reserve holders of the risk of capital losses on those existing dollar

40
important to understand that the “requires” part does not follow. To see this, observe that

the euro has gained ground as a reserve currency even though the euro area has not run

significant current account deficits in recent years. Or recall that countries accumulated

dollar reserves under the original Bretton Woods System even though the U.S. had a

balanced current account and even substantial surpluses for the vast majority of the

period. All that is required is that the reserve-currency country running the balanced

current account should invest abroad at least in an amount equal to the incremental

demand for reserves in the rest of the world.

Hence the argument that being the sole supplier of reserves creates a tendency for

a country to run chronic deficits must be a different one. It must be that the desire of

other countries to accumulate reserves reduces the incentive for the reserve issuer to run a

balanced current account. Knowing that other countries demand additional reserves and

will willingly finance the reserve center’s current account deficit, policy makers in the

reserve-issuing country must have less incentive to adopt painful policies that raise

national savings to the level of national investment.79 Think of it as a problem of moral

hazard.

To the extent that this moral hazard is present, the question is what to do about it.

One idea is ongoing issuance of Special Drawing Rights to provide a non-national source

of incremental reserves. The IMF would issue SDRs on a regular basis in amounts equal

to the increase in global reserve demand. The problem here is that SDRs can be used only

balances. This is the idea of creating a new “Substitution Account.” I do not consider this here (except in a
couple of footnotes down) for reasons of space and because it is concerned with the financial legacy of past
imbalances rather than the question of how to prevent future imbalances.
79
This is a conceivable result, although not a necessary one. Still, it is not implausible that this was part of
the explanation for the imbalances problem of recent years. In other words, there is a high probability that
the United States would have adopted policies more closely equalizing the country’s saving and
investment—or that the market would have brought about this result through a decline in the dollar—had
there not existed a strong central bank demand for dollar reserves.

41
for transactions with the IMF and among consenting governments. Unlike national

currencies they cannot be used for foreign exchange market intervention and other

transactions with market participants. For central banks and governments that see

reserves as insurance – that anticipate actually having to use them – this illiquidity

renders SDRs unattractive.80

Making SDRs attractive would require making them liquid. This would mean

developing private markets on which SDR claims can be bought and sold. It would be

necessary to build broad and liquid markets on which governments and, for that matter,

financial and nonfinancial firms can issue SDR bonds at competitive cost. Banks would

have to find it attractive to accept SDR-denominated deposits and extend SDR-

denominated loans. The pension funds and insurance companies that are the dominant

sources of private demand for bonds would have to be attracted to holding bonds

denominated in a basket of currencies despite the fact that their liabilities tend to be

dominated in a single national currency.81 It would be necessary to restructure foreign

exchange markets so that traders seeking to buy, say, Korean won for Thai baht first sold

baht for SDRs (before buying won) rather than first selling baht for dollars. While all

this is possible, it would not be easy. It is worth recalling that there was a previous

attempt to commercialize the SDR in the 1970s that never really got off the ground.

80
Just why Chinese, Russian and Brazilian officials have been pushing the SDR option is an interesting
question. It could be that they see it as a stalking horse for a Substitution Account – as a way of getting
existing dollar balances off their balance sheets as opposed to an alternative for accumulating future
reserves. It could be that they see this as a way of demonstrating their desire to be players in discussions of
international monetary reform. Conference participants may have a better answer to this question than I.
81
They could swap out the currency risk, but this would be an additional cost of the investment strategy,
which would presumably render it unattractive – or require an interest-rate premium of the issuer, which
would make issuance less attractive.

42
Succeeding this time would take decades rather than years.82 We can discuss it at the San

Francisco Fed’s 10st biannual Asia-U.S. conference.

As part of this effort, the IMF would have to be authorized to issue additional

SDRs in periods of shortage, much as the Fed provided dollar swaps to provide dollar

liquidity in the second half of 2008. At the moment countries holding 85 per cent of IMF

voting power must agree before SDRs can be issued, which is not exactly a recipe for

quick action. IMF management would have to be empowered to decide on emergency

SDR issuance just as the Federal Reserve can decide to offer emergency currency swaps.

For the SDR to become a true international currency, in other words, the IMF would have

to become more like an independent global central bank. The idea of an independent

IMF has its advocates, as I have made clear above, but it is not clear that China, Russia,

Brazil and other advocates of replacing the dollar with the SDR are aware that this is the

implication of their proposal.

The other approach to reducing the dominance of the dollar would be to diversify

the sources of international reserves. The moral hazard felt by any one nation’s policy

makers would then be limited. Imagine 20 years from now three economies of roughly

comparable size, each with a convertible currency traded on liquid markets that can be

used to satisfy the incremental demand for reserves. No one of them will be able to

reduce its saving relative to its investment by a substantial margin simply because the

global demand for reserves is growing. One way of understanding how global

imbalances grew so pronounced in recent years is that the incremental demand for

reserves was increasingly large while the share of the reserve-issuing country in the

82
The current crisis itself is a reminder that building liquid markets in a new, novel asset is not something
that occurs overnight.

43
global economy was unusually small. So it was that the United States came to account

for some 75 per cent of global current account deficits. With the U.S., the euro area and

China all issuing reserves (to reveal the identities of my three plausible candidates for

reserve center status 20 years from now), such imbalances would be less. Given the

existence of alternatives, an issuer prone to excessive deficits would quickly see other

countries accumulating reserves in currencies other than its own. That, in turn, would be

a source of external discipline.

This, I have argued elsewhere, is the direction we are heading.83 The euro’s share

of global reserves has risen since the new European currency was created in 1999. And

Chinese officials have clearly mounted a campaign to transform the renminbi into an

international currency, encouraging domestic and foreign firms to settle their transactions

in renminbi, signing agreements with foreign governments to do likewise, extending

renminbi swaps to foreign central banks, and relaxing restrictions on the ability of foreign

financial institutions to issue renminbi debt in Hong Kong.

But, again, the euro and the renminbi will match the dollar as an attractive form of

reserves only when they possess equally deep and liquid markets. The market in U.S.

treasury debt remains far and away the most liquid in the world. Europe and China may

eventually succeed in creating equally liquid markets in debt securities denominated in

their currencies, but the relevant time frame is measured in decades, not years. Europe’s

problem is that the stock of government debt securities is not homogeneous. Different

government bonds differ in their risk, returns and liquidity. German bunds have a

reputation for stability, but since they tend to be held to maturity by institutional investors

the market in them lacks liquidity. Other euro area countries with plenty of bonds have
83
See Eichengreen (2009d).

44
deep financial problems as a result of past policies and the crisis. Italian government

bonds are in fact the most important euro area debt securities by value, but the country’s

problems mean that they are not attractive as reserve assets. The crisis has encouraged

talk of issuing euro area bonds and putting the full faith and credit of the entire set of

members, starting with Germany, behind them. Were this done on a significant scale and

were such debt to replace the national debt securities of the member states, the euro area

would possess something more closely resembling the U.S. treasury market. But this

kind of radical fiscal federalism is not something to which the German government is

likely to agree anytime soon.

For the renminbi, an important precondition is full capital account convertibility,

and even that is only necessary, not a sufficient, for market liquidity. Chinese officials

have targeted 2020 as the date by which Shanghai should be transformed into an

international financial center, meaning that its markets are open to foreign investors free

of capital-account restrictions. At that point the process of building truly liquid markets

can commence.

Someday we will have a multiple reserve-currency system not unlike the one that

existed before 1913 that limits the problem of global imbalances. But not tomorrow.

Conclusion

Financial crises are complex. Our recent crisis is one such complex event whose

causes can be broadly grouped under two headings: lax regulation combined with skewed

incentives in financial markets; and accommodating monetary policy combined with

global imbalances that fueled an unsustainable housing and credit boom.

45
 That crises rarely have a single cause means that avoiding them can rarely be

achieved by a single policy reform or set of reforms. This paper has therefore provided

two lists of reforms designed to address the two sources of instability contributing to our

recent crisis. Both lists are long. Neither will be easy to implement. In both cases

powerful stakeholders will resist reform. In both cases important details remain to be

worked out. In both cases the extent of intellectual agreement on what must be done may

be less deep and broad than I have made out in this paper.

Be that as it may, now that the worst of the crisis has passed it is important that
the sense of urgency attached to reform, and the willingness to collaborate internationally
in its pursuit, not also be allowed to pass.

46
References

Balls, Edward (2003), “Preventing Financial Crises: The Case for Independent IMF
Surveillance,” Remarks at the Institute of International Economics (6 March),
www.piie.com.

Bank of England (2009), “Building a More Resilient Financial System,” Chapter 3 in

Blinder, Alan (2009), “It’s Brok, Let’s Fix It: Rethinking Financial Regulation,” Paper
prepared for the Federal Reserve Bank of Boston Conference, Chatham, Mass. (23
October).

Buiter, Willem H. (2009), “Lessons from the Global Financial Crisis for Regulators and
Supervisors,” unpublished manuscript, London School of Economics (May).

Calomiris, Charles (2009), “Financial Innovation, Regulation and Reform,” Cato Journal
29, pp.65-91.

Cohen, Rodgin and Morris Goldstein (2009), “The Case for an Orderly Resolution
Regime for Systemically-Important Financial Institutions,” Pew Charitable Trusts,
Financial Reform Project Briefing Paper no.13 (October).

Clarida, Richard, Jordi Gali and Mark Gertler (1998), “Monetary Policy Rules in
Practice: Some International Evidence,” European Economic Review, pp.1033-1068.

Committee on Capital Markets Regulation (2009), “The Global Financial Crisis: A Plan
for Regulatory Reform,” copyright, Committee on Capital Markets Regulation (May).

DeGregorio, Jose, Barry Eichengreen, Takatoshi Ito and Charles Wyplosz (1999), An
Independent and Accountable IMF, London: CEPR.

de Larosiere, Jacques et al. (2009), “Report of the High-Level Group on Financial

Dooley, Michael and Peter Garber (2005), “Is it 1958 or 1968? Three Notes on the
Longevity of the Revived Bretton Woods System,” Brookings Papers on Economic
Activity 1, pp.147-187.

Duffie, Darrell (2009), “Policy Issues Facing the Market for Credit Derivatives,” in John
Ciorciari and John Taylor (eds), The Road Ahead for the Fed, Stanford: Hoover
Institution Press, pp.123-136.

Eichengreen, Barry (2009a), “Lessons of the Crisis for Emerging Markets,” Paper
prepared for the George Mason University/Claremont Graduate School Conference on
the Financial Crisis and Asia.

47
Eichengreen, Barry (2009b), “The Last Temptation of Risk,” National Interest 101
(May/June), pp.8-14.

Eichengreen, Barry (2009c), “Out of the Box Thoughts on the International Financial
Architecture,” IMF Working Paper no. WP/09/116 (May).

Eichengreen, Barry (2009d), “The Dollar’s Dilemmas,” Foreign Affairs

Eichengreen, Barry and Andrew Rose (1999), “The Empirics of Currency and Banking
Crises,” NBER Reporter, http://www.nber.org/reporter/winter99/eichengreen.html.

Elliott, Douglas (2009), “Preemptive Bank Nationalization would Present Thorny

Financial Stability Forum (2008), Report of the FSF on Enhancing Market and
Institutional Resilience, Basel: FSF (7 April).

Financial Stability Forum (2009), “FSF Principles for Sound Compensation Practices,”
Basel: FSF (2 April).

Glick, Reuven and Andrew Rose (1998) “Contagion and Trade: Why Are Currency
Crises Regional?” NBER Working Paper no. 6806 (November).

Goldman Sachs (2009), “Avoiding another Meltdown,” Global Markets Institute, New
York: Goldman Sachs (March).

Goodhart, Charles and Dirk Schoenmaker (2009), “The de Larosiere Report: Two Down,
Two to Go,” VoxEU, www.vox.eu (30 March).

Hallerberg, Mark, Rolf Ranier Strauch and Juergen von Hagen (2009), Fiscal
Governance in Europe, Cambridge: Cambridge University Press.

International Monetary Fund (2007), “IMF Executive Board Discusses Multilateral

Kashyap, Anil (2009), “A Sound Funeral Plan can Prolong a Bank’s Life,” www.ft.com
(29 June).

Persaud, Avinash (2009), “Macroprudential Regulation: Fixing Fundamental Market (and

48
Prasad, Eswar (2009), “Global Macroeconomic Imbalances: G20 Leaders Must Back Up
Their Rhetoric with Deeds,” Financial Times Economists’ Forum, blogs.ft.com (13
October).

Rodrik, Dani (2008), “The Real Exchange Rate and Economic Growth,” Brookings
Papers on Economic Activity (Fall), pp.365-412.

Rose, Andrew and Mark Spiegel (2009), “Cross Country Causes and Consequences of
the 2008 Crisis: Earning Warning,” CEPR Discussion Paper no. 7484 (July).

Shih, Victor (2009), “China Takes the Brakes Off,” Wall Street Journal, www.wsj.org
(23 July).

Taylor, John (2009), Getting Off Track: How Government Actions and Interventions
Caused, Prolonged and Worsened the Financial Crisis, Stanford: Hoover Institution
Press.

Tucker, Paul (2009), “Regimes for Handling Bank Failures – Redrawing the Banking
Social Contract,” Remarks at the British Banks’ Association, London, www.bis.org (30
June).

United States Treasury (2009a), Financial Regulatory Reform: A New Foundation,

United States Treasury (2009b), “Principles for Reforming the U.S. and International
Regulatory Capital Framework for Banking Firms,” Washington, D.C.: U.S. Treasury
(September).

Visco, Ignazio (2009), “The Global Crisis – The Role of Policies and the International
Monetary System,” Presentation to the G20 Workshop on the Global Economy, Mumbai,
www.bis.org (24-26 May).

Warnock, Frank and Virginia Warnock (2006), “International Capital Flows and U.S.
Interest Rates,” International Finance Discussion Paper no. 860, International Finance
Division, Board of Governors of the Federal Reserve System (September).

49