Memorandum submitted by Professor Sayantan
Ghosal, Dr Emanuel Kohlscheen and Professor Marcus Miller, University
of Warwick and CSGR
RESERVE POOLING, MORAL HAZARD AND SUSTAINABILITY:
THE ROLE OF THE IMF
This paper relates to the first topic for investigation
in the inquiry, namely the impact of globalisation on the current
and future roles of the IMF.
In considering the issues of (a) Reserve Acquisition
and Reserve Pooling; (b) Moral Hazard and Political Economy; and
(b) Sustainability and Debt Swaps, we draw on three papers we
are to present at the Easter Meetings of Royal Economic Society,
(1) Emanuel Kolscheen: "International
Liquidity Swaps: an Assessment of the Chiang Mai Initiative";
(2) Sayantan Ghosal and Kannika Thampanishvong
"Optimal Sovereign Debt Write-downs"; and
(3) Sayantan Ghosal and Marcus Miller: "Growth,
sustainability and delay: a
model of sovereign debt restructuring".
1. Introduction: Setting up the IMF
When the IMF was established at Bretton Woods
in 1944, it was to oversee and help manage a global system of
"fixed but adjustable" exchange rates. Countries were
only meant to change exchange rates in "fundamental disequilibrium";
and one role of the Fund was to judge whether or not this was
Another key role for the Fund was to provide
short-term financial support to countries experiencing payments
difficulties. With speculative flows held in check by controls
on international capital movements, it was anticipated that the
resources of the Fund would be used for temporary assistance to
members in financing balance of payments on current account, Crockett
(1977, p 217).
The contribution of each member to the IMF's
financial resources was (and is) determined by its assigned "quota",
broadly determined in relation to a country's size and its opennessa
quarter paid in the form of widely accepted hard currency, the
rest in own currency. A country's quota determines both its borrowing
rights and effectively the weight it has in the weighted majority
voting of the Executive Board, where decisions require a simple
majority, except for constitutional issues where a supermajority
of 85% is necessary.
The size of the Fund itself is decided by a
vote of the members in the Annual General Meeting of Governors,
every five years at least. When it was first established in the
1940s with 45 member countries, total quotas subscribed amounted
to about two percentage points of global GDP.
Though the 1998 quota review saw a substantial increase in the
value of quotas to about $300 billion, the review of early 2003
resulted in no change in quotas. Despite the increase of membership,
total quotas are now only about one percent of global GDP.
Capital Account Liberalisation
Increased freedom of transactions on capital
account have wrought major changes, however. In the first place,
with markets able to anticipate adjustments in pegged exchange
rates by speculative flows, the Bretton Woods system of fixed-but-adjustable
exchange rates was no longer viable: and it came to an end in
the early 1970s, when countries were free to adopt floating exchange
rates. Secondly, potential calls on IMF resources could now be
driven by adverse capital flows, with the need to speed up IMF
lending in response to capital account reversals being discussed
as early as 1972 (De Vries, 1985).
As capital account liberalisation accelerated
in the 1990's, "capital account crises" became widespread,
beginning with that of the European Exchange Rate Mechanism in
1992-93. After an initial widening of currency bands, the European
response has been closer monetary and political integration, leading
to the abolition of many national currencies and the creation
of the euro. But emerging market countries with implicit or explicit
pegs against major currencies have remained exposed to repeated
financial crises, beginning with the Mexican peso crisis of 1994-95,
and soon followed by the East Asian crises of 1997-98 which the
economist Jeffrey Sachs likened to a "bank-run".
The rise in sovereign spreads consequent upon
the Russian default of 1998 exposed many Latin American and other
emerging markets with substantial external debts in dollars to
self-fulfilling runs. Indeed, in a recent IMF Discussion Paper
advocating a Country Insurance Facility, Cordella and Levy-Yeyati
(2005, p 3) conclude that there is "an increasing consensus
that self-fulfilling liquidity runs have triggered many of the
recent financial crises, as sudden increases in perceived rollover
risk set off an escalation of interest rates that rendered otherwise
sustainable debt levels unsustainable".
The initial reaction by the IMF to such events
was to assemble financial packages ("bailouts") involving
levels of assistance much larger than the two-times-quota seen
as an appropriate limit for current account crises. IMF disbursements
were about five times quota for Mexico (1995), 18 times for Korea
(1998), 7.7 times for Brazil (2001 and 2002 combined) and 17 times
for Turkey (1999-2001 and 2002 combined).
Latterly, however, the IMF has sought to limit its exposure by
recommending policy changes by the debtorwith respect to
exchange rate for exampleand debt restructuring or "bail-ins"
The reaction by many emerging economies has
been self-insurance by the accumulation of international reserves.
For a sample of 35 emerging markets, reserve holdings measured
on this basis has doubled over the last decadefrom 8.9%
of GDP in 1992 to 18.1% in 2002.
Korea and China hold reserves valued at 22% of GDP. There have
been moves towards co-insurance at a regional level: and suggestions
for novel co-insurance provisions by the IMF.
Key issues for debate
Ex ante mechanisms to help prevent capital
(a) a shift towards floatingbut managedexchange
(b) the increase in owned-reserve positions
since the East Asian crises of 1997-98, particularly by countries
in the region (Jeanne and Ranciere, 2005);
(c) the development of intra-regional co-insurance
mechanisms such as the Chiang Mai agreement of SE Asia (discussed
in section 2 below);
(d) the creation of an automatic Country
Insurance Facility (CIF) as a complement to existing IMF arrangements
( Cordella and Levy-Yeyati, 2005); and
(e) possible use of capital controls (Levy-Yeyati,
Ex post mechanisms for crisis resolution:
(a) Collective Action Clauses (see, for example,
Sayantan Ghosal and Kannika Thampanishvong, 2005a);
(b) The risks that governance problems pose
for IMF lending and implications for conditionality (discussed
in section 3 below); and
(c) Improvements in bilateral bargaining
procedures (discussed in section 4).
2. Ex ante co-insurance: Regional Agreements
A new form of coinsurance has emerged in East
Asia: the network of bilateral swap agreements (BSAs) launched
under the Chiang Mai Initiative. A BSA is a bi-lateral arrangement
that gives either signatory access to a fixed amount of the co-signatory's
international reserves, upon request. China, Japan and South Korea
have now signed such agreements with each other and also with
Indonesia, Malaysia, the Philippines and Thailand. The total amount
of these agreements (U$ 37.5 billion) may be small compared to
the volume of international reserves accumulated by the larger
economies of the Far East in the last few years: but for smaller
the amounts involved are significant and already correspond to
about four times the reserve tranches of these countries at the
The network has already been seen by some observers
as the first step towards an Asian Monetary Fund. While there
is no evidence that BSAs have so far affected bond valuations,
Kohlscheen (2005) finds that the network reflects the potential
gains from co-operation in that larger agreements have been signed
between countries for whom international reserves tend to move
in opposite directions. He points out that the same economic logic
implies that BSAs could be signed between countries of different
regions (eg Asia and Latin America).
The increase in the prevalence and magnitude
of the BSAs (the Chinese government has already pledged to double
its commitments) brings important politico-economic questions
to the forefront. Although they have yet to be tested, the existence
of regional networks that can quickly supply liquidity may provide
an additional line of defence and help shield signatories from
the perils of sudden capital flow reversals. Bilateral or regional
swap agreements could lead to a more efficient use of existing
international reservespossibly reducing the aggregate demand
for international reserves. This may change demand for IMF resources
Could the prevalence of such regional arrangements
lead to a "moral hazard" problem as countries that are
members of such networks underinsure themselves? Given the substantial
accumulation of own reserves, this does not appear to be an issue.
Note too that 90% of the disbursements are conditional on the
existence of an IMF programme.
3. Governance issues
Provision of ready access to foreign exchange
to help insure against crises may seem a clear welfare improvement:
and if debtor countries are well-governed and the use of loan
facilities subject to explicit "conditionality", the
possible perverse effects on incentivesthe problem of debtor
moral hazardmay not seem an issue. But what if there are
problems of governance and there is an interaction between international
bailouts and domestic policies?
In these circumstances, as two economists from
the IMF have argued, "The international financial safety
net increases the scope for bad policies as well as good policies.
Increasing the borrowing capacity in such circumstances may lead
to worse outcomes, not better. International bailouts can help
countries implement good policies, but they could also make the
consequences of bad domestic policies much worse." Jeanne
and Zettlemeyer (2001, p 411).
In the paper they are to present at the RES
conference, Ghosal and Kampanishvong (2006) consider the situation
when government is by an elite which is not representative of
the population as a whole. Instead of providing public goods,
such as health, education and social infrastructure, the elite
pursues its own idiosyncratic objectives with an emphasis on short-run
payoffs. It may, for example, use public funds for ostentatious
investment and conspicuous consumption: additionally, public borrowing
offshore may be diverted to finance the acquisition of private
bank accounts in safe havens such as Switzerland, where ownership
of accounts is not divulged.
Where poor governance involves domestic distributive
conflict and myopia, the authors argue that conditionality imposed
by external funding agencies can and should be designed strategically
to help deal with such issues. Debt write-downs can be made conditional
on government's provision of public goods, for example; and limits
may be imposed on new issuance of public debts and on taxes levied
on the non-elite.
If not, external support may simply provide more resources for
the elite to misuse. The paper cited focuses on African countries,
but the problems of governance they discuss are much more widespread.
In emerging markets of East Asia and Latin America,
where private external finance is readily available, the provision
of financial bailouts may have perverse incentive effects on creditors.
For if creditors know they have an exit option, they will have
less incentive to monitor the uses to which their funds are being
put; and will be willing to lend at lower rates. This problem
of creditor moral hazard will be reflected not in high sovereign
spreads but by a "rush for the exit" when prospects
take a turn for the worse.
If the creditors do not charge high spreads
for the risk of default, who pays? It is a popular fallacy that
the international community pays, via the lending programmes of
the IMF and World Bank. But it has been shown beyond any reasonable
that IMF lending does not mask economic transfers: except for
some poor countries, the Fund lends short term and recovers its
money. A key feature of the economic programmes that accompany
IMF packages is the requirement that the debtor does repay the
IMF in full: otherwise the IMF would be driven out of business.
In the absence international transfers and of
a significant write-down of private debt, it is therefore the
debtor that paysand this will typically mean the non-elite.
By allowing for exit by the private lenders, but insisting that
it is paid in full, the IMF must needs use the conditionality
of its lending to extract ex post what capital markets
would have charged ex ante: and it is often the non-elite
who bear the burden of earlier excess borrowing.
4. Ex post resolution: Delay in Debt Swaps
In the late 1990s the IMF took steps to extricate
itself from the invidious position of acting as debt collector
for careless creditors; the idea being to "bail in"
the creditors rather than bail them out. The high point of this
attempt came in 2001 when Ann Krueger proposed an official Sovereign
Debt Restructuring Mechanism, based on Chapter 11 of the US bankruptcy
code, to oversee debt write downs ex post.
This initiative was rejected, however. Instead,
as the case of Argentine has shown, write-downs may need to be
negotiated in face-to-face bargains between debtor and creditors,
without the intermediation of the IMF.
(In future, such negotiations should be expedited by the widespread
incorporation of Collective Action Clauses into sovereign debt
While creditors have been "bailed- in"
in several recent cases, negotiations proved to be fairly lengthy.
Default by Ecuador, for example, which led to a 60% "haircut"
(write-down of face-value) took a year to resolve. The defaults
by Russia and Argentina, which involved the largest amounts of
debt (around $30 billion and $80 billion respectively) and resulted
in the largest haircuts (closer to 70%), both took a year and
a half of negotiation.
In the paper they are to present to the RES,
Ghosal and Miller (2006) focus on reasons for delay in bargaining,
with special reference to the case of Argentina.
A model of bargaining is used to show that settlement may well
be delayed until the economy recovers: and delay occurs when growth
prospects exceed the rate of discount. In addition, however, the
need to ensure a sustainable settlement appears to have played
a key role in the Argentine debt negotiations, as the debtor was
desperate to avoid a repeat experience of default.
When bargaining is adapted so to reflect concern
for sustainability, this tends to increase in the debtor's bargaining
power; but asymmetric information about the debtor's situation
increases the likelihood of delay. The debtor claims that the
write-down is needed to ensure economic viability, but the creditor
isn't sure if that is true: delay is one way that the debtor can
make the point. Could this been why the Argentine episode involved
the longest delayand largest write-downin recent
sovereign debt restructuring?
If sustainability is a key factorbut
creditors are unsure about the detailsthe IMF could help
to reduce delay by providing reliable information: where sustainability
conditions are made common knowledge, there is no need for the
debtor to use delay as a signalling device.
A potentially serious challenge to carrying out this informational
role is that the IMF, as senior creditor, faces a "conflict
of interest": there is presumably an incentive to exaggerate
sustainability requirements in favour of the debtor so as minimise
other claims on the debtor's resources. Perhaps such induced compassion
for debtors would be checked by its creditor-dominated Executive
Board? If not, this informational task could be delegated elsewhere:
to the IADB for example.
Increased mobility of capital has opened a Pandora's
box of issues for discussion. In this submission we have focussed
on three that have implications for the IMF: the provision of
ex ante co-insurance through regional swaps that could reduce
the calls on IMF resources; governance problems in borrower countries
which may limit the IMF's incentives to provide co-insurance;
and the delays that may attend ex post resolution in the absence
of agreement of the parameters involved. The last of these suggests
a role for the IMF as provider of information and judgement on
issues of debtor sustainabilityso long as its credibility
is not compromised by conflict of interest.
T and E Levy-Yeyati (2005) "A (new) Country Insurance Facility".
IMF Working Paper No 05/23.
Cohen, D and R Portes (20040 "Towards a Lender
of First resort" CEPR DP. No 4615, London: CEPR.
Crockett, A (1975) International Money. Sunbury-on-Thames:
De Vries, M (1985) The International Monetary
Fund, 1972-1978. Washington:IMF.
Dhillon, A, J Garcia-Fronti, S Ghosal and M Miller
(2006) "Debt Restructuring and Economic Recovery: Analysing
the Argentine Swap". The World Economy, (forthcoming).
Ghosal, S and M Miller (2006) "Growth, sustainability
and delay: a model of sovereign debt restructuring". Mimeo.
University of Warwick.
Ghosal, S and K Thampanishvong (2005) "Bargaining,
Moral Hazard and Sovereign Debt Crisis" Mimeo. University
Ghosal, S and K Thampanishvong (2006) "Optimal
Sovereign Debt Write-downs" Mimeo. University of Warwick.
Irwin, G, A Penalver, C Salmon and A Taylor (2005)
"Dealing with country diversity: challenges for the IMF coinsurance
model" Paper presented at LACEA 2005, Paris.
Jeanne, O and R Ranciere (2005) "The optimal
level of international reserves for emerging market countries:
formulas and applications" Mimeo, IMF.
Jeanne O and J Zettlemeyer (2001). "International
bailouts, moral hazard and conditionality" Economic Policy,
vol 16, Issue 33, pp 409-32.
Kohlscheen, E (2005) "International Liquidity
Swaps: an Assessment of the Chiang Mai Initiative" Mimeo.
University of Warwick.
Levy Yeyati, Eduardo (2005), "Tras el canje
de deuda vuelve el control de capitales version 2005", Cronista
Comercial. Buenos Aires. 26 February.
Rochet, J-C. (2005) "Optimal Sovereign Debt:
An analytical approach". Mimeo. University of Toulouse.
Roubini N and B Setser (2004a), Bailouts or Bail-ins?
Responding to Financial Crises in Emerging Economies. (Washington
DC: Institute for International Economics).
Roubini N and B Setser (2004b), "The Reform
of the Sovereign Debt Restructuring Process: Problems, Proposed
Solutions and the Argentine Episode" Journal of Restructuring
Finance, 1, 1, 173-84.
Porzecanski, A C (2005) "From Rogue Creditors
to Rogue Debtors: Implications of Argentina's Default". Chicago
Journal of International Law, 6(1) pp.311-32.
54 The papers will posted on the Royal Economic Society
website; (1) and (3) will feature in a Special Session at the
Though votes are seldom taken, the voting rules are nonetheless
important: the fact that the US has a 17% share of the subscriptions
means that it has a veto on any change of the Fund's Articles,
for example. Back
Irwin et al (2005). Back
World GDP was valued at $32, 300 for 2002, The Economist, Pocket
World in Figures, 2005 Edn. Back
See Cohen and Portes (2004) for a similar diagnosis. Back
In the case of Korea, however, when outflows threatened to overwhelm
the $41 billion assembled as first and second lines of defence
in this way, G7 countries engineered a successful "bail in":
instead of raising fresh funds to finance the exit of banks, they
persuaded their banks to roll over their lending to the country. Back
See Roubini and Setser (2004a). Additionally, in most cases except
Turkey, the amounts made available were substantially augmented
by other bilateral commitments mainly supplied by G7 countries
(Roubini and Setser 2004a, Table 4.1). Back
Roubini and Setser (2004a) Chapter 4 "Experience with Bailouts
and Bail-ins" provides a good review. Back
Cordella and Levy-Yeyati (2005). Back
Such as the last Indonesia, Malaysia, the Phillipines and Thailand. Back
For the Philippines, for example, the $5.3 billion made available
through BSAs amounts to about a third of its stock of international
In similar vein, Rochet (2005) argues that crisis lending by the
IMF may need to be accompanied by a constitutional amendment limiting
debt creation by the government. Back
See Jeanne and Zettlemeyer (2001). Back
See Dhillon et al (2005) for a bargaining approach applied to
this case. Back
The delay between Argentine default and restructuring (40 months)
was a lot longer than the period of negotiation (19 months): this
is because the interim administration appointed in early 2002
regarded debt restructuring as outside its competence so negotiations
did not commence until President Kirchner was elected in mid-2003
with a mandate to negotiate. Back
Some observers argue that the Argentine swap has a new era of
the rogue debtor, Porzecanski (2005). If so, one would expect
sovereign spreads to rise sharply as creditors "price in"
the cost of future write-downs. As yet, there is no sign of that. Back
In their assessment of the reasons for delay, add that the IMF
can play a key role in coordinating creditors Roubini and Setser
(2004b): they also argue that the IMF could orchestrate "debtor
in possession" finance, Roubini and Setser (2004a). Back