a framework for restructuring debt in developing countries through the creation of special sovereign...

7/27/2019 A Framework for Restructuring Debt in Developing Countries Through the Creation of Special Sovereign Borrowing

1/34

2013 Cambridge Business & Economics Conference ISBN : 9780974211428

A FRAMEWORK FOR RESTRUCTURING DEBT IN DEVELOPING

COUNTRIES THROUGH THE CREATION OF SPECIAL

SOVEREIGN BORROWING ENTITIES1

Roger W. Clark2

Austin Peay State UniversitySchool of Business

Clarksville, TN 37044, USAPhone: 931-221-7574Fax: 931-221-7355

Email: [email protected]

And

George C. Philippatos

University of Tennessee432 Stokely Management CenterKnoxville, TN 37996-0540, USA

Phone: 865-690-9684Email: [email protected]

And

David J. MooreCollege of Business Administration

California State University, Sacramento

Tokoe Hall 21176000 J Street

Sacramento, CA 95819-6088Email:[email protected]

First draft January, 2003Revised May, 2006Revised March, 2007Revised November 2010Current Draft April 2013

1 Earlier abbreviated versions of this research were presented at the fifth Encuentro de FinanzasInternacional USACH, Via del Mar, Chile, January 2003, The Durham University Finance Conference,June 27, 2006, Durham, UK and at several universities. The authors wish to thank participants at theseconferences for their suggestions. Needless to say, the authors take full responsibility for any errors stillremaining in this paper.2 Corresponding AuthorJuly 2-3, 2013Cambridge, UK 1
mailto:[email protected]:[email protected]


2/34


A FRAMEWORK FOR RESTRUCTURING DEBT IN DEVELOPING

COUNTRIES THROUGH THE CREATION OF SPECIAL

SOVEREIGN BORROWING ENTITIES

By Roger W. Clark, Austin Peay State University, P.O. Box 4415, Clarksville, TN 37044, USA; andGeorge C. Philippatos, University of Tennessee 432 Stokely Management Center Knoxville, TN 37996-0540, USA; and David J. Moore, California State University, Sacramento, CA. 95819-6088, USA

ABSTRACT

The recent Global Financial Crisis (2007- present) has renewed the interest of theInternational Financial Institutions (IFIs) in finding new methods of accommodating thecontinuing needs of developing countries for outside capital.

This paper proposes the creation of a special sovereign borrowing entity (SSBE)under the auspices of the International Monetary Fund (IMF) and other InternationalFinancial Institutions (IFIs) that would guarantee bond issuances by developing nations,package them in relatively small denominations of $50,000 - $500,000 US, and auctionthem to the public. Should a nation fail to pay its debts the SSBE may raise fundsthrough a punitive tariff on all exports from the developing country, administered by allnations that are members of the IMF.

An extensive illustration that contains computations and a Monte Carlo simulationis also provided to enhance the analytical material in this paper regarding the feasibilityand viability of the proposed Special Sovereign Borrowing Entity (SSBE)

Key Words: special sovereign borrowing entity, sovereign debt, bankruptcy, default,debt restructuring, Monte Carlo simulation, IMF, IBRD

JEL Code: F18, F34

July 2-3, 2013Cambridge, UK 2


3/34


1. INTRODUCTION

The financial arrangements of the current system of sovereign debt issuance are

similar to financing in the nineteenth century. Loans are made on general credit; the

creditors do not have full access to the debtors books, and in many cases they do not

have real control over how the money is spent or how it is to be repaid. The remedy for

default is the sovereign equivalent of debtors prison; that is, the drying up of trade

credits and restrictions on trade with the debtor country. To the chagrin of the global

financial system, these sanctions have not aided the debtor nations in paying off their

debts. Starting around 2004, Great Britain agreed to pay off 10% of the debt of

developing countries1, but this does not reach the underlying problem of their growing

needs for outside capital.

The present paper proposes the creation of a special sovereign borrowing entity

(SSBE) under the auspices of the International Monetary Fund (IMF) and other

International Financial Institutions (IFIs), that would guarantee bond issuances by

developing nations, package them in relatively small denominations of $50,000 -

$500,000 US, and auction them to the public. Should a nation fail to pay its debt the

SSBE may raise funds through a punitive tariff on all exports from the developing debtor

country, administered by all nations that are members of the IMF.



4/34


2. NATIONAL DEBT VS CORPORATE DEBT

Bulow, et. al. (2002) have likened debt of developing countries to corporate junk

bonds, the main differences being collateral and control. Corporate bonds generally have

the assets of the firm for backing. Further, should a corporation have poor governance,

the trustee in bankruptcy has the option to replace completely the management team.

Neither of these two options is operationally feasible with national debt. The markets

also rate the debt of developing countries at junk-bond levels, viewing them as more than

likely to default, generally a 50% writedown is anticipated over the next decade. (Bulow,

et. al. (2002))

A further problem exists in the size of the total national debt and its uses. While it

has been observed that in general the total national debt of a developing country amounts

to only three to six months of its GDP, this is deceiving (Bulow, et. al., 1997). Sachs

(2002) has developed this concept further with his so-called poverty trap. He notes the

large number of small developing countries (SDCs) that have been in a continual state of

financial crisis for many years and that the only ones that have been cured of high debt

were the countries with larger gross domestic incomes (GDIs) initially. From this he

infers that, in contrast to the neo-Classical view, savings and capital accumulation rates

are non linear. Low income countries will be spending more on basic survival rather than

the building of capital. As a consequence, these small countries will almost never be able

to lift themselves out of poverty. Also the problem with lending from multinational

agencies such as the IMF and the World Bank is that they have been prone to use

arbitrary formulas for lending (and lending only enough to avert disaster, rather than



5/34


solve the basic problem). Sachs proposes that debt renegotiations take into consideration

the fact that maximization of debt repayments is only one of a series of objectives. Other

objectives are: (a) giving the insolvent country a fresh start, (b) preserving its public

functions, and (c) achieving its development objectives. This calls for the creditor

countries and International Financial Institutions (IFIs) making independent plans and

estimates for the debtor countries development, and for pulling them out of the poverty

trap.

All of the above literature, however, begs the question of why a nation would

bother paying its debts. The age of gunboat diplomacy, where a powerful nations fleet

could sail into the harbor of a defaulting debtor nation and seize its assets, is past. Also,

most developing nations do not have assets of high value outside of their borders subject

to seizure. Still, nations that renounce their debt entirely are few or almost non-existent.

Nations in financial trouble usually undergo financial reorganization under the auspices

of the IMF and renegotiation of debt under the Paris and London Clubs. Again, why

would they go to this trouble if they can simply renounce their debt?

Eaton and Gersovitz (1981) have advanced the Reputational Capital model. In

their framework the potential loss of future borrowing power is greater than the amount

of debt renounced. However, this has been questioned by Bulow and Rogoff, (1989) who

argue that loss of reputational capital alone cannot account for the full value of the debt

renounced. Rogoff and Zettelmeyer (2002) have noted that the literature of the 80s does

not give easy solutions. It would appear the true economic loss is both loss of

reputational capital and the possibility that individual lawsuits against a nation would



6/34


trigger cross-default clauses in all contracts of the country. This would entail a rush to

the courthouse by all creditors and imperil ongoing IMF loans.

Another cost of default would be the drying up of trade credits for both imports

and exports (Cohen, 1991, Kaltesky, 1985, Rogoff, 1999). This could effectively cut off

the debtors trade with the outside world and bankrupt merchants in the debtor country

that depended on this trade. Also, Rose (2002) has noted that a countrys problems with

portfolio creditors may have an adverse impact on foreign trade. In his paper Rose notes

that a typical debtor countrys outside trade shrank on average 8% a year and the effects

lasted for fifteen years after renegotiation of debt under the Paris club [Rose (2002) p.

19]. One question that was not explored was whether this was a punishment by the

creditor countries for defaulting or a natural consequence of poor economic conditions in

the debtor countries examined.

In a paper released later the same year Rose (2002) finds a high correlation

between the amount of loans a creditor country makes to a debtor country and amount of

trade with the debtor country. This seems to add weight to the hypothesis that trade can

be used as a lever to obtain favorable terms by the creditor. Should this be the case, it

would appear the market for government bonds of developing countries is limited to the

trading partners of the debtor country. Potential creditors (e.g. for example portfolio

investors) in other countries may not have the power to enforce payment under present

arrangements.



7/34


3. RENEGOTIATING NATIONAL DEBT

When a nation finds itself unable to make timely payments on its debt it must first

go to the IMF to obtain temporary financial relief and submit to a program approved by

the IMF for resolving the underlying problem. Once the country has agreed to the IMF

program it petitions the Paris Club to renegotiate its sovereign debt. The Paris Club is

an (ad hoc) informal organization of nations that work to reschedule loan repayments on

terms the debtor nation can afford to pay. It can lengthen the terms and/or change the

interest rates, depending on the needs of the nation. This club has no formal enforcement

powers and works by consensus. The process usually takes six to eight months.

Afterwards, the debtor nation goes to the London club to work out its private loans. In

general, the London Club reschedules loans using the same terms as the Paris Club. This

arrangement has generated some criticism. Typically, the criticism is that the IMF has a

one size fits all policy that does not take into full consideration the specific problems in

each country (Stiglitz, 2002). Also, the mere existence of supra national organizations

may create a moral hazard problem.2 Additionally, the fact that the Paris Club operates

by consensus without enforcement authority may create a free rider problem3 (supra).

Finally, the entire process of negotiation engenders serious problems with informational

asymmetries and adverse selection.4

The presence of such informational asymmetries may affect asset values through

a nexus of economic distortions that manifest themselves as agency costs, signaling

costs, moral hazard costs, free rider costs, and adverse selection costs, among

others. In general the presence and costs of informational asymmetries are higher in the



8/34


mixed market economies of the developing (debtor) nations than they are in the market

economies of the G8 (creditor) countries. Much of the problem can be traced to lax

accounting standards, poor record keeping and internal auditors controlled by those being

audited.5 (Johnson, 1992).

Should a country actually have a healthy, corruption free system it will find its

official statistics not much different from a corrupt country that also misrepresents its

own statistics. Here we have a problem with informational asymmetry and adverse

selection. It is to the creditors interest to have as much information as possible

concerning a nations ability to repay its debts. However, as Stiglitz (1975, 2002) has

suggested, it is not in the creditors interest to have this information made public

knowledge, so that he/she can take advantage of this informational asymmetry to demand

higher interest rates. All of this, of course, leads to the problem of adverse selection.

Creditors, without the benefit of credible public regulatory agencies that promote and

enforce full disclosure, will be forced to rely on private, often unreliable, channels of

information. This will force interest rates up to offset perceived debtor risk. Debtor

nations will, in turn, only borrow if they have high positive NPV projects or when they

have no intention of repaying the full amount of the debt. As a consequence of this, debt

from developing countries has consistently stayed at junk bond levels (see Bulow, 2002).

Informational asymmetry also arises in the prospect of a nation having its debt

repurchased by the larger industrial countries. The problem here is that the entire

premise of this debt repurchase rests on the value of the debt remaining at the market

before the repurchase is announced. When it is discovered the debt will be repurchased



9/34


by the larger industrial nations, the debt value will rise to approximate the value of the

larger countries bonds. (Bulow and Rogoff, 1990 )

One possible solution to this problem of informational asymmetry may be in the

syndicated loan market. Sufi (2005) finds informational asymmetry is greatest with

companies new to the loan market. Here due diligence and monitoring are most needed.

Lead bankers that arrange the loans usually overcome this fear of informational

asymmetry by taking a large position in the loan amount. The reputation of the lead

lenders may mitigate, but not reduce significantly, the effects of informational

asymmetry.

Additional support along the efforts to mitigate informational asymmetry through

the syndicated loan market is offered by Dailami, et. al.. Using data from J P Morgan

Emerging Market Bond Index Plus (EMBI+) spreads, they studied the secondary

markets of Brady Bonds and Eurodollar issues for all major sovereign borrowers (17

countries), by separating the public data into crisis and non-crisis periods for

countries that are close to the borderline of solvency and those that are not. The

authors concluded as follows, there is evidence that investors are much more able to

discriminate among borrowers, and less likely to infer that problems in one country signal

problems in others (p.27). They further offer clinical evidence of the Argentinean

debt default in 2002 the largest known default in recent history which did not cause

much disruption in world capital markets, nor did its neighboring countries suffer major

increases in their spreads6.



10/34


3.1 Moral hazard

In general, the concept of moral hazard stems from the behavioral perceptions that

insurance against some unforeseen event may actually encourage it to occur. As applied

to multinational finance, the IMF has arranged standby funding packages for several

emerging-market countries in financial crisis. To wit, some parties have expressed

concern that this could have long-term adverse effects on the world economy. For

creditors it encourages lending to poor countries without factoring fully the importance of

financial risk. Thus, some debtor countries may obtain financing without undertaking

needed economic reforms.

Another case of moral hazard exists when we witness the private creditors egging

the IFIs on to bail out the bankrupt debtor countries, although we know that it would

be better, from an efficiency standpoint, if they could precommit not to bail out the

countries in distress. Bulow (2004), further states the IFIs are probably making

loans that earn zero economic profits. Therefore the IFIs facilitate excessive lending.

(ibid, p.6) Indeed, in many cases, the IFIs make loans that are below the break even

point. Therefore the cost to the IFIs must be greater than or equal to the

inefficiencies created.7

The overall question is whether we have empirical evidence of the presence and

the extent of moral hazard in sovereign debt financing. Kamin (2002) has studied moral

hazard using the spreads before the 1995 Mexican bailout as a benchmark. This was the

first major bailout of a large country in financial trouble and, he reasons, before this,

countries would have no reasonable expectations of a major bailout in the event of a



11/34


financial crisis. He finds no statistically significant increase in the spreads before and

after the Mexican bailout except in the period of 1996 through mid 1998, which he

attributes to financial market exuberance. Thus, Kamin (2002) concludes that there is

little evidence for moral hazard.

DellAriccia, Schnabel, and Zettelmeyer (2002) discovered possible evidence for

moral hazard in the aftermath of the Russian lending crisis of 1998. Here the authors

regressed the spreads of sovereign bonds against country fundamentals both before and

after the Russian crisis. The results were that these spreads increased significantly after

the crisis when it was perceived that country bailouts were not going to be as common as

in the past. Further, these spreads were largest in emerging market countries8. They

point out, however, that this evidence for moral hazard rests on the premise that country

bailouts do not significantly reduce true economic risk. Should this premise be rejected

the increase in spreads reflects the increased risk in a world without these safeguards.

Rogoff (2002) notes in an article commenting on moral hazard and the IMF that to date,

the IMF loans have, in general, been paid in full. Further, he notes that, even if there

were a default, the countries involved are so small that the loan would amount to a

miniscule portion of the global GDP. Rogoff further likens the IMF loans to the

performance of the Federal Reserve System in the 1930s. This system performed well

until the 1980s before there was any appreciable loss from moral hazard in the form of

the Savings and Loan (S&L) crises. This should be balanced against the roughly fifty

years of relative stability in the U.S. banking system.



12/34


3.2 The free rider problem

Sachs (1986) outlined the free rider problem of renegotiating debt in the current

environment. When a country defaults on debt involving several creditors it is to a

creditors advantage to hold out in any renegotiations, rather than join in with the rest of

the creditors. In this way the creditor may sue to enforce their legal rights and obtain a

better settlement than other creditors. This can cause a destructive race to force a debtor

to pay through the courts. Balanced against this is the problem of jurisdiction. The fact

is, that these lawsuits are brought in a country other than the debtor country. Unless the

debt is secured by collateral located in another country the debtor nation can simply

refuse to pay.

3.3 Off balance sheet financing and SSBEs for

sovereign borrowing.

With the Enron and World Com scandals the Special Purpose Entity (SPE) has

acquired a rather questionable reputation. Much of this is undeserved. The SPE was

originally conceived as a method of allowing private firms to avoid capitalization of

certain types of leases under Financial Accounting Standard (FAS) 13. Under the

original SPE rules, outside investors would contribute at least 3% (now 10%) of a

companys capital while the originator company could guarantee as much as 90% of the

SPEs debt. In many cases this allowed the SPE to obtain lower financing rates than the

originator companys normal borrowing rate. Additionally, the SPE had the cosmetic



13/34


effect of keeping the liabilities associated with the project off the originator companys

books9.

The SPE is still used broadly in private industry and in some cases in

governmental entities. In the United States the Federal National Mortgage Association

(Fannie Mae) and the Governmental National Mortgage Association (Ginnie Mae),

among others, function as SPEs. In addition Japan has formed offshore SPEs to handle

non-performing loans and enhance the income statements and balance sheets of their

commercial banks.

The international financial institutions (IFIs) such as the World Bank or the IMF

also function as a form of SPE for the United Nations and the world community. They

generally couch their aid to developing countries in the form of loans or loan guarantees.

This allows the developed countries to utilize leverage in aiding less developed countries.

As an example, the World Bank is capitalized at $189.5 billion while only $11.5 billion

has been paid in by the subscribing countries (Bulow, et.al. 2002)10.

3.4 Comparative Statics of Known U.S. Government

SPEs and Proposed Special Sovereign Borrowing

Entity (SSBE)

By far the best known United States Governmental SPEs are Ginnie Mae

(GNMA), Freddie Mac (FRE), Fannie Mae (FNMA) and Sallie Mae (SLM). All four are

institutions founded by the U.S. Government11. Their responsibility is to provide



14/34


liquidity for the housing and student loan markets where they rate, grade, securitize, and

frequently auction underlying claims (mortgages and/or loans).

The main focus of Ginnie Mae is to ensure liquidity for U.S. Government insured

mortgages, including those insured by the Federal Housing Administration (FHA), the

Veterans Administration (VA), and the Rural Housing Administration (RHA). The

majority of mortgages securitized as Ginnie Mae (MBSs) are those guaranteed by the

FHA. FHA mortgagees are typically first time home buyers and low income borrowers.

Sallie Mae, on the other hand, guarantees the timely payment of interest and principal for

college education (student) loans, while Fannie Mae resembles a hybrid of Sallie Mae

and Ginnie Mae. , below, summarizes the attributes of the proposed sovereign borrowing

entity (SSBE) and Sallie Mae, which appears to be the closest equivalent among the U.S.

Government SPEs.

Tables 2 presents some descriptive statistics from another well-known government SPE,

Freddie Mac (FRE) for the 15 year period 1990-2004. From the statistics it transpires

that mortgage purchases during this period increased by a multiple of 11.25 times, while

the issuance of mortgage backed securities (MBSs) increased by a factor of 9.8 times. As

can be seen from Charts 1 and 2 the delinquency rates of all loans, which include Freddie

Mac loans, is higher than the delinquency and foreclosure rates of conventional loans.

This is expected as the buyers under Freddie Mac are usually first time buyers making

comparatively low down payments.



15/34


The default rates of Sallie Mae student loans are not published in detail, but Sallie

Mae has estimated that Consistent with historical experience, the assumed average

lifetime default rates are 15 percent for borrowers attending four-year schools, 30 percent

for borrowers attending two-year schools, and 10 percent for borrowers attending

graduate or professional schools12.

4. A MODEL FOR SOVEREIGN COUNTRY BORROWING

UNDER AN SSBE

Let us assume there is one country, , that wishes to borrow money. It may

choose to borrow directly from a creditor country, , or borrow through the SSBE.

Should borrow through and finance it with the issuance of sovereign bonds, it will

face the penalty of trade restrictions in the event of default. Rose (2002) and Rose and

Spiegel (2002) have demonstrated that these restrictions can be and apparently are used

in such circumstances. They do not seem particularly effective as the punishment hurts

the creditor in trade as well as the debtor, without actually remedying the default.

Apparently the market does not think highly of this as a remedy; hence the trading of

these obligations at junk bond levels.

Should the country choose to go through the SSBE, it will receive a loan amount

of L from the SSBE. We begin with a two-period model in which L will be received in

period one and repaid in period two at D where D = (1+r)L; and r is the debtor

countrys cost of capital. Loan proceeds will be used to enhance trade with other

countries, thus fostering diversification of trading partners and giving the debtor country



16/34


revenues in the second period of T; where, T is the total amount of exports from the

debtor country. It is further assumed the prices of the exports are inelastic. The utility

function in country will satisfy the following condition:

E1(U) = U(C1) + E(U(C2)) (1)

where U >0, U r, and r

represents the new discount rate of the country in default, adjusted for costs of

reputational capital and loss of foreign trade.

From equations (3) and (4) it is evident the utility maximizing choice of the

debtor country will be to pay its debts under all circumstances.



17/34


5. AN ILLUSTRATION OF THE PROPOSED

SPECIAL SOVEREIGN BORROWING ENTITY

VIABILITY AND FEASIBILITY

The Special Sovereign Borrowing Entity (SSBE) is expected to operate in a

higher default rate environment than U.S. government-sponsored SPEs (US-SPEs). In

such an environment interest will be collected from a smaller percentage of loans.

However, the combination of higher interest rates (to compensate for higher risk) and the

presence of credit enhancements13may partly offset the effects of increased defaults.

At this stage we identify several preliminary conjectures regarding the proposed

special sovereign borrowing entity (SSBE) viability and feasibility:

Profitability sensitivity conjecture. The profit margin of firms that securitize

loans is insensitive or positively related to the default rate of underlying loans.

Effective spread conjecture. The effective spread14 of a loan securitization SSBE

is not adversely affected by increased default rates.

Feasibility conjecture. The cash reserve requirements of the proposed SSBE are

sufficiently small to be accommodated by available IMF/World Bank funds.

To test these conjectures we analyze data obtained from COMPUSTAT, Reuters

Corporate Spreads for Industrials and annual reports of Fannie Mae, Freddie Mac and

Sallie Mae, for various sub-periods between 1968 and 2004 (whenever they are

available). Using these databases, we compute the profit margin of these US sponsored

SPEs over time and estimate the effect of risky debt on such margins. We then employ



18/34


these computations to draw inferences about the proposed Special Sovereign Borrowing

Entity (SBE). From here we estimate the relationship between default rate and spread

and incorporate the results into a Monte Carlo simulation of SBE operations to arrive at

the effective spreadmeasure.

5.1 Profitability sensitivity

5.1.1 Model

It is important to illustrate the connection between profit margin and default rate,

given the likelihood of higher default rates by poorer LDCs. In addition, delinquencies

have the potential to impact earnings adversely through increased servicing costs,

collection costs, and account charge-offs. Therefore, delinquency rates capture much of

the cost of carrying risky debt as do default rates. Hence, given the availability of a

sufficient number of observations, we should be able to infer the impact of default rates

on profit margins using delinquency rate data. This was shown in Table 2, which

presents the U.S. delinquency and foreclosure rates (as a percentage of outstanding loans)

for all loans and for conventional loans during the period 1990 2004. However, the

number of observations available is not large enough to yield statistically significant

results.

Nevertheless, since the primary purpose of this exercise is to provide a practical

illustration, we shall proceed with the empirics and rely on quarterly delinquency data

from the three U.S. sponsored SPEs (FNM, FRE and SLM). To test the profitability



19/34


sensitivity conjecture, we apply the asset pricing model in Equation 5 to profit margins

and include delinquency rate DELRATE as follows:

PMt

0

1PMM,t

2DELRATEt

ut (5)

where profit margin PM is defined as net income divided by sales. If the coefficient

2 is positive and statistically or economically insignificant, we will have

demonstrated lack of profitability sensitivity to delinquency rates.

5.1.2 Model inputs

In order to run the regression specified in equation (5), three data items are needed: i) the

profit margin over time for each US sponsored SPE; ii0 the market profit margin over

time, and ii) delinquency rates over time. US sponsored SPE profit margin data are

extracted from COMPUSTAT for FNM, FRE, and SLM for the periods 1968 2002,

1986 2004 and 1981 2004, respectively. The market profit margin is computed on a

year by year basis as the equal-weighted profit margin for all companies with nonzero

sales (to avoid division by zero). Default rate data are extracted from the U.S.

delinquency and foreclosure rate for all loans during the period 1990-2004 data

previously shown in Table 2. Although default rate is not explicitly published, we

consider the foreclosure rate as synonymous with default rate.

5.1.3 Results

Regression (5) is run for each of the three US sponsored SPEs and results are reported in

Table 3below with standard errors in parentheses.



20/34


5.2 Effective Spread

5.2.1 Model

A percentage ( CER ) of the SSBE loan portfolio will likely have credit

enhancements while the remainder (1 CER ) will not be credit-enhanced. The typical

spread of 200 basis points associated with less developed country debt is assigned to the

spread of the credit-enhanced portfolio SCE . By construction of credit-enhancements,

all interest will be collected from the credit-enhanced (CE) portfolio irrespective of the

default rate experienced. However, the amount of interest collected from the non-credit-

enhanced (NCE) portfolio will depend on the associated spread SNCE and default rate of

NCE loans DEFRATENCE . Therefore, the effective spread ES is as expressed in

Equation (6) below:

ES CER SCE R

AAA 1 CER 1 DEFRATENCE SNCE RAAA RAAA (6)

where RAAA represents the rate of a AAA rated corporate bond. A Monte Carlo

simulation of (6) is run and the value of ES is compared to the average spread associated

with US-SPEs.

5.2.2 Model inputs

This section describes the data used and construction of requisite inputs to the Monte

Carlo simulation of equation (6).



21/34


a) Spread-default rate relationship for CE portfolio

We estimate the relationship between the default rate and spread based on credit

ratings by utilizing data from: (1) Reuters Corporate Spreads for Industrials (June, 2004);

and (2) Moodys Investor Service Global Credit Research (February, 2002). From these

data we estimate the relationship between default rate and spread. The data reveals that

the spread between a bonds rate and the 10 year U.S. Treasury security Spread

increases as the bond rating decreases.15

b) Credit-enhancement ratio

In computing the credit-enhanced percentage of the total mortgage portfolio, Freddie Mac

excludes non-Freddie Mac agency16 and non-agency17 mortgage related securities.

However, delinquency rates are subdivided for credit-enhanced and non credit enhanced

portions of total mortgage portfolio. The general aim of this model is to connect the

spread between loans to sovereign countries and bond coupon payments in the context of

higher default rates and mortgage guarantees. Therefore, the exclusion of agencies and

non-agencies is acceptable given the equivalent guaranteeingfunction embodied in

credit enhancements.

Based on Freddie Mac Data. a random variable for CER was generated that has a

triangular18 distribution with lower limit, modal value, and upper limit equal to the

minimum, mean, and maximum values ofCER , respectively.



22/34


c) Spread-default rate relationship for the NCE portfolio

Freddie Mac reports delinquencyrate data for its credit enhanced and non credit

enhanced portfolios while our Monte Carlo model (6) requires the overall defaultrate.

The relationship between delinquency rates of CE loans and NCE loans can be estimated

using Freddie Mac delinquency data. From this relationship we infer the NCE/CE default

relationship and thus mitigate the shortcomings of default rate data unavailability. A

random variable for the NCE default rate is formed based on the CE default rate and the

NCE/CE relationship.

Finally, now that we have a value for the NCE portfolio default rate, we compute the

associated spread by utilizing the spread/default relationship obtained earlier.

5.2.3 Simulation Results

Referring to Table 2, the average default rate (foreclosures started) for all mortgages

in the U.S. is 0.34% over the period 1990 to 2004. Using the previously determined

default rate relationship, this translates into a spread of approximately 80 basis points. In

the neighborhood of this spread, the three US-SPEs have persisted (i.e., have remained

viable). Error: Reference source not found reflects the effective spread of the SSBE

given for various values for the credit-enhanced loan spread. As you can see, the

effective spread at the proposed 200 basis point value of SCE is well above the 80 basis

point US-SPE spread. In other words, the proposed SSBE is able to realize a higher

effective spread than that of US sponsored SPEs. What this means is the higher costs of



23/34


default (including lost interest payments) are more than offset by the additional revenue

generated from having a larger spread.

5.3 Feasibility Conjecture

To investigate the feasibility of creating the proposed SSBE, we consider i) the cash to

loan ratio of US sponsored SPEs (USCTLR) and ii) the anticipated amount of SSBE

loans. On average, the loan balance for US sponsored SPEs is approximately 80% of

total assets. Therefore USCTLR can be calculated as:

USCTLR cash

0.80* totalassets (11)

Using this calculation the average cash-to-loan ratios for FNM, FRE, and SLM are 6.5%,

11.5%, and 16.0%, respectively. Should the anticipated amount of loans handled by the

proposed SSBE be designated as $ 50 billion, the initial amount of IMF/World Bank cash

needed to seed the SPE is $3.25 to $8 billion dollars.

5.4 Tentative Conclusions

Based on the analytics, empirics, and Monte Carlo simulations of Sections IV and V, we

can infer the following:

1. SSBEs, which securitize loans, can be profitable even in the presence of higher

default rates;

2. In addition, only a small amount ($3.25 to $8 billion dollars) of available IMF

funds is needed to fund the start up costs associated with the proposed Sovereign

Borrowing Entity (SSBE) for an anticipated amount of loans of $50 billion.



24/34


3. Finally, in light of the above it is evident that the utility maximization choice will

motivate borrowing nations to pay rather than default of their SSBE sponsored

loans.

6. RECOMMENDATIONS

This paper proposes a new type of sovereign bond issuance where risk is largely

determined at the time of issuance and greater transparency is given to the purchasers.

The International Monetary Fund will create and fund a special sovereign borrowing

entity (SSBE) that will be modeled after the Federal National Mortgage Association

(Fannie Mae) in the United States. They would examine the debt capacity of countries

interested in offering bonds on the market using methodologies similar to that used by

Moodys or Standard and Poors or an investment bank deciding whether to underwrite a

private bond offering19. An acceptable price would be established commensurate with the

risks involved. The SSBE would then package the issuance in a manner similar to the

Fannie Mae bonds in the United States and sell them at auction on a regular basis. They

will guarantee the bonds at a specific interest rate. Should a nation fail to pay when due,

the agency will have power to levy a tax or tariff on the exports of the debtor country.

This will give the agency enforcement capabilities through its member nations. In

addition, the debtor nation will make its books available to the SSBE and funds will be

released only as the projects, such as roads or schools, are completed. This arrangement

should open the markets for developing country bonds to a worldwide group of investors.

Since there will be no one nation heavily invested in the debtor country (diversification



25/34


effect), the temptation to restrict trade will be minimal. Instead, a tariff can be better

tailored to collect revenues with minimal effects on the debtor countrys external trade.

In perspective, our recommendations for an SSBE based sovereign debt structure

contain implicitly several advantages and financial safeguards, as follows:

1. The auctioning of the SSBE-claims on a regular basis will foster some

semblance of practical diversification, which will enable claim-holders to develop

risk-adjusted measures of return for these SSBE bonds.

2. The guarantee provisions of SSBE claims by the IMF will likely upgrade

these claims eventually to investment grade and improve their liquidity and

marketability. In the long-run, as the default probabilities of such bonds are

reduced (or, at least, bounded), some financial institutions may wish to employ

them as instruments of portfolio immunization along with other investment

quality claims.



26/34


Tables and Charts

Table 1 Sallie Mae and proposed sovereign borrowing entity attributes

Similarities Between a U.S. Government SPE (SLM) and Proposed Sovereign Borrowing Entity(SBE)

SALLIE MAE (SLM) Proposed SOVEREIGN BORROWING ENTITY(SBE)

Rates, Grades, Securitizes and auctions off loans ofCollege Students

Rates, Grades, securitizes and auctions off the loansof Less Developed Countries (LDCs)

Funds are released as academic work progresses (ona semester basis)

Funds will be released as projects progress (on aquarterly basis)

Loan repayments are typically made. Studentborrowers will repay old loans by taking new loans(refinancing)

Loans typically will not be paid off, rather they willbe replaced with new loans. This will allow moreinterest to be earned (from refinancing).

Default is mitigated by a screening (monitoring)process and the refinancing of loans.

Default will be mitigated by a screening/monitoringprocess (rating, grading, and frequent auctioning)

and the refinancing of loans.



27/34


Table 2 U.S. delinquency and foreclosure rates frame

Percent Of Outstanding Loans

All Loans Conv Loans

Delinquency RatesForeclosures

Started

Delinquency RatesForeclosures

Started60-90 Days > 90 Days 60-90 Days >90 Days

1990 0.82 0.78 0.29 0.51 0.44 0.26

1991 0.83 0.86 0.35 0.52 0.5 0.22

1992 0.7 0.8 0.34 0.44 0.47 0.26

1993 0.65 0.79 0.31 0.39 0.43 0.22

1994 0.69 0.76 0.33 3.43 0.43 0.22

1995 0.72 0.73 0.33 0.47 0.41 0.25

1996 0.73 0.63 0.33 0.44 0.32 0.24

1997 0.75 0.65 0.37 0.46 0.35 0.27

1998 0.73 0.64 0.37 0.44 0.32 0.27

1999 0.67 0.6 0.3 0.37 0.28 0.212000 0.79 0.66 0.31 0.44 0.28 0.23

2001 0.87 0.78 0.38 0.55 0.35 0.28

2002 0.79 0.82 0.37 0.5 0.42 0.27

2003 0.71 0.8 0.4 0.84 0.3 0.2

2004 0.66 0.73 0.4 0.35 0.29 0.2

*Percent of outstanding loans

Note: data are as of year-end except for 2004 (1st quarter)

Source: Mortgage Bankers Association

As quoted in Freddie Mac Factbook 2004

Table 3 - Regression results for quarterly data

Firm 0

1

2

N R2

adjR2

FNM 8.98 (12.35) -1.19 (1.15) 17.30 (22.50) 32 0.05 -0.01

FRE 98.83 (42.89) -13.01 (3.74) 183.26 (77.34) 17 0.60 0.54

SLM 0.49 (2.20) -2.63 (1.58) 2.84 (2.05) 17 0.17 0.05



28/34


Chart 1 60-90 day U.S. delinquency rates Chart 2 - >90 day U.S. delinquency rates

U.S. Delinquency Rates 60-90 days

0

0.2

0.4

0.6

0.8

1

1990

1992

1994

1996

1998

2000

2002

2004

Percentoutstanding

loans

All Loans

Conv Loans

U.S. Delinquency Rates >90 days

0

0.2

0.4

0.6

0.8

1

1990

1992

1994

1996

1998

2000

2002

2004

Percentoutstanding

loans

All Loans

Conv Loans

REFERENCES

Bulow, Jeremy, 2002, First World Governments and Third World Debt.Brookings

Papers on Economic Activity, Spring 2002 p229(27)

Bulow, Jeremy, and Kenneth Rogoff, 1988a, Multilateral Negotiations for Rescheduling

Developing Country Debt: A Bargaining-Theoretic Framework,IMF Staff Papers, Vol.

35 (December), pp. 64457.

Bulow, Jeremy, and Kenneth Rogoff, 1988b, The Buyback Boondoggle,Brookings

Papers on Economic Activity: 2,

pp. 64598.

Bulow, Jeremy, and Kenneth Rogoff,1989, Sovereign Debt: Is to Forgive to Forget?

American Economic Review, Vol. 79, (March), pp. 4350.

Bulow, Jeremy, and Kenneth Rogoff, Sovereign Debt Repurchases; No Cure for

Overhang, The Quarterly Journal of Economics, Vol 106, No 4 (Nov., 1991), 1219-1235



29/34


Bulow, Jeremy, and Kenneth Rogoff, Cleaning Up Third World Debt Without Being

Taken to the Cleaners, The Journal of Economic Perspectives, Vol 4, No 1 (Winter,

1990), 31-42

Bulow, Jeremy, Kenneth Rogoff, and Alfonso S. Bevilaqua. 1992. "Official Creditor

Seniority and Burden-Sharing in the Former Soviet Bloc." BPEA, 1:1992, 195-222.

Cohen, Daniel (1991)Private Lending to Sovereign States (Cambridge: MIT Press).

Dailami, Mamsoor,, Paul R. Masson and Jean Jose Padou, 2005, Global Monetary

Conditions versus Country-Specific Factors in the Determination of Emerging Market

Debt Spreads, World Bank Policy Research, Working Paper 3626, June, 2005

Dugger, Celia, World Bank Challenged: Are Poor Really Helped? The New York

Times, July 28, 2004, Section A, page 3, column 4.

Eaton, Jonathan, and Mark Gersovitz, 1981, Debt with Potential Repudiation:

Theoretical and Empirical Analysis,Review of Economic Studies, Vol. 48 (April), pp.

289309.

Jeanne, Olivier, and Jeromin Zettelmeyer. 2001. "International Bailouts, Moral Hazard,

and Conditionality." Economic Policy: A European Forum 16(33): 407-24.

Johnson, Daphne M., Professionalism in the Third World, Internal Auditor, Oct. 1992

v49 n5 p51(4)

Kaletsky, Anatole, 1985, The Costs of Default(New York: Priority Press).



30/34


Masson, Paul, 2003, Empirical Regularities in Emerging Market Spreads, mimeo,

Brookings Institution.

Rogoff, Kenneth, 1999, International Institutions for Reducing Global Financial

Instability.Journal of Economic Perspectives 1304, 21-42.

Rogoff, Kenneth, and Jeromin Zettelmeyer, 2002, Bankruptcy Procedures for

Sovereigns: A history of Ideas, 1976-2001, IMF Staff Papers, Vol 49, No 3.

Rose, Andrew K., 2002, One Reason Countries Pay Their Debts: Renegotiation and

International Trade, NBER Working Paper 8853, March.

Rose, Andrew K., and Mark Spiegel, 2002, A Gravity Model of Sovereign Lending:

Trade, Default and Credit, NBER Working Paper 9285, October.

Sachs, Jeffrey D., 2002, Resolving the debt crisis of low-income countries,Brookings

Papers on Economic Activity, Spring 2002 p257(30)

Sachs, Jeffrey D., 1986a, Managing the LDC Debt Crisis,Brookings Papers on

Economic Activity: 2, pp. 397431.

Sachs, Jeffrey D., 1986b, Testimony to the Subcommittee on International Trade of the

Committee on Finance, United States Senate, May 13, 1986 (Washington: U.S.

Government Printing Office), pp. 5890.

Stiglitz, Joseph E., The Theory of Screening, Education, and the Distribution of

Income, American Economic Review, June, 1975, 65(3), pp. 283-600



31/34


Stiglitz, Joseph E., 2002, Globalism's discontents: integration with the global economy

works just fine when sovereign countries define the terms. It works disastrously when

terms are dictated The American Prospect, Jan 1, 2002 v13 i1 pA16(6)

Stiglitz, Joseph E., Information and the Change in the Paradigm in Economics, The

American Economic Review, Vol. 92, No. 3, June, 2002, pp. 460-501

Sufi, Amir, Information Asymmetry and Financing Arrangements: Evidence from

Syndicated Loans Working paper, presented at Midwest Finance Association, Eastern

Finance Association, and Western Finance Association annual meetings. , April 2005

The paper may be found on the Internet at

http://gsbwww.uchicago.edu/fac/amir.sufi/sufisl0425.pdf

ENDNOTES

http://gsbwww.uchicago.edu/fac/amir.sufi/sufisl0425.pdfhttp://gsbwww.uchicago.edu/fac/amir.sufi/sufisl0425.pdf


32/34


1UK Pledges Intl Debt Relief; Urges Other Lenders Do Same The Wall Street Journal. September 26, 2004 Found at

http://online.wsj.com/article/0,,BT_CO_20040926_001724,00.html. The U.K. pledge was augmented by more concrete

pledges made at the G8 Glenn Eagle summit, held in June, 2005, Scotland, U.K. See IMF Survey, Vol. 34, No. 13, July 18,

2005, p. 2003

2 In general, the concept of moral hazard stems from the idea that insurance against some unforeseen event may actually

encourage it to occur. As applied to multinational finance, the IMF has arranged standby funding packages for several

emerging-market countries in financial crisis. Several experts have expressed concern that this could have long-term

adverse effects on the world economy. For creditors it encourages lending to poor countries without factoring fully the

importance of financial risk. Thus, some debtor countries may obtain financing without undertaking needed economic

reforms.

Another case of moral hazard exists when we see the private creditors egging the IFIs on to bail out the bankrupt

debtor countries, although we know that it would be better, from an efficiency standpoint, if they could precommit

not to bail out the countries in distress. Bulow (2004), further states the IFIs are probably making loans that earn zero

economic profits. Therefore the IFIs facilitate excessive lending. (ibid, p.6)

3 Sachs (1986) has outlined the free rider problem of renegotiating debt in the current environment. When a country

defaults on debt involving several creditors it is to a creditors advantage to hold out in any renegotiations, rather than join

in with the rest of the creditors. In this way the creditor may sue to enforce their legal rights and obtain a better settlement

than other creditors, which can cause a destructive race to force a debtor to pay through the courts. Balanced against this is

the problem of jurisdiction. The fact is, that these lawsuits are brought in a country other than the debtor country. Unless

the debt is secured by collateral located in another country the debtor nation can simply refuse to pay.

4 Informational asymmetries arise when the assumption of perfect information of neoclassical economics theory is violated

by the presence of differential information among the market participants engaged in negotiations and/or transactions. The

presence of such informational asymmetries may affect asset values through a nexus of economic distortions that manifest

themselves as agency costs, signaling costs, moral hazard costs, free rider costs, and adverse selection costs,

among others. In general the presence and costs of informational asymmetries are higher in the mixed market economies of

the developing (debtor) nations than they are in the market economies of the G7 (creditor) countries. Much of the problem

http://online.wsj.com/article/0,,BT_CO_20040926_001724,00.htmlhttp://online.wsj.com/article/0,,BT_CO_20040926_001724,00.htmlhttp://online.wsj.com/article/0,,BT_CO_20040926_001724,00.html


33/34


can be traced to lax accounting standards, poor record keeping and internal auditors controled by those audited. The

consequences of this agency-like phenomenon are compounded by the confounding of the roles of those being audited

(agents) and those performing audits (principals) (Johnson, 1992).

5 The consequences of this agency-like phenomenon are compounded by the confounding of the roles of those being

audited (agents) and those performing audits (principals). The reader is referred to the alleged involvement of Goldman

Sachs in restructuring the books of the Greek Government. This, in effect, handed Greece a $1 billion loan from the

eurozone in exchange for $300 million in fees for the restructuring efforts of the investment bankers (see Michael Lewis,

Boomerang, W.W. Norton company pages 12-21).

6 Dailami, et al, also cite a paper by Masson (2003) who found that co-movement of EM spreads was lower in crises

subsequent to the Asian crisis, indicating greater differentiation among countries. Ibid, p.27

7 The above arguments notwithstanding, Bulow, Rogoff and Bevilaqua (1992) inform us that Hardly any debt to the IMF

or the World Bank has ultimately defaulted. However, when there is a loan default to the IFIs, as Jeanne and Zettelmeyer

argue, the moral hazard is born by the domestic taxpayer in the debtor countries.

8 Since Russia has recently become a free-market republic these results should be viewed with healthy skepticism.

9 It should be noted here that since the early 1970s many science and technology based companies have employed SPEs to

finance their research and development activities. Known variably as SWORD or R&D limited partnerships (LPs) these

spun-off entities have become the financing vehicles of choice for chemical and pharmaceutical companies, where R&D

activities are characterized by high costs of financing, long horizons, great uncertainty and lengthy regulatory approval

processes. Such R&D SPEs involve an additional type of agency problem, i.e. between the investors of the SPEs and the

management of the parent company. Of course, the ultimate test for SWORDS and R&D LPs as financial innovations is

whether they enable parent companies to raise capital at a lower cost for a unique class of assets like R&D or to raise capital

that could not be gathered through other traditional financing modalities.

10This enables countries such as the United States to guarantee $30 billion in loans while showing on the books that the

World Bank membership has only cost them $2 billion throughout its history (Bulow, et.al (2002) p.236).



34/34


11 While we realize that the problem associated with the sub-prime loan crisis have put these companies in the spotlight we

also believe that their long and successful record will ensure their future survival in some enhanced organizational forms.

12 CBO Papers, The Experience of the Stafford Loan Program and Options for Change, Congressional Budget Office, 1991

13 Credit-enhanced loans are loans for which the lender or a third party has retained primary default risk by pledging

collateral or agreeing to accept losses on loans that default (e.g., private mortgage insurance provider). In some cases, the

lender's or the third party's risk is limited to a specific level of losses at the time the credit enhancement becomes effective.

14 The net effect of higher default rates, higher interest rates, and credit enhancements

15 Many of the computational details are shown in an appendix, in order to keep the size of this paper within proper limits.

This detailed appendix is available from the authors upon request.

16 Securities issued or guaranteed by Fannie Mae or Ginnie Mae

17 Rely primarily on subordinated tranches to provide credit loss protection and therefore limit exposure counterparty risk

18 Triangular distributions are chosen over normal distributions to ensure positive percentages and over uniform

distributions to capture estimated central tendencies

19 These are flow-through securities.

a framework for restructuring debt in developing countries through the creation of special sovereign...

Documents