Sovereign Risk

1

• Chapter 16
• Sovereign Risk

2
Introduction

• In 1970s
• Expansion of loans to Eastern bloc, Latin America
  and other LDCs.
• To meet these countries demand for funds beyond
  those provided by the World Bank and the IMF to
  aid their development.
• Beginning of 1980s
• Poland and Eastern bloc repayment problems.
• Debt moratoria announced by Brazil and Mexico
  government in the fall of 1982.
• At the time of the 1982 moratoria, the 10 largest
  US money center banks had overall sovereign risk
  exposure of 56 billion, 80 of which was to
  Latin America.
• Increased loan loss reserves
• In 1987, more than 20 US banks announced major
  additions to their loan loss reserves, with
  Citicorp alone setting aside 3 billion.

3
Introduction (continued)

• Late 1980s and early 1990s
• Expanding investments in emerging markets.
• Peso devaluation on December 20, 1994.
• The US provided loan guarantees over three to
  five years that would amount to up to 20 billion
  to help restructure Mexican debt. The IMF and the
  Bank for International Settlement provided loans
  of 17.8 billion and 10 billion, respectively.
  Mexican oil revenues were promised as collateral
  for the US financial guarantees.
• More recently
• Asian and Russian crises.
• In 2001, concerns have been raised about the
  ability of Argentina and Turkey to meet their
  debt obligations and the effects this will have
  on other emerging market countries.
• In December 2001, Argentina defaulted on 130
  billion in government issued debt, and in 2002,
  passed legislation that led to defaults on 30
  billion of corporate debt owed to foreign
  creditors.

4
Credit Risk versus Sovereign Risk

• When making loans to borrowers in foreign
  countries, two risks need to be considered.
• First, the credit risk of the project needs to be
  examined to determine the ability of the borrower
  to repay the money. This analysis is based
  strictly on the economic viability of the project
  and is similar in all countries.
• Second, unlike domestic loans, creditors are
  exposed to sovereign risk. Sovereign risk is
  defined as the uncertainty associated with the
  likelihood that the host government may not make
  foreign exchange available to the borrowing firm
  to fulfill its payment obligations. Thus, even
  though the borrowing firm has the resources to
  repay, it may not be able to do so because of
  actions beyond its control.
• Need to assess credit quality and sovereign risk

5
Credit Risk versus Sovereign Risk

• If a sovereign risk happened,
• The lenders legal remedies to offset a sovereign
  countrys default or moratoria decisions are very
  limited.
• Lenders can and have sought legal remedies in US
  court, but such decisions pertain only to foreign
  government ot foreign corporate assets held in
  the US itself.
• Should the credit risk or quality of the borrower
  be assessed as good but the sovereign risk be
  bad, the lender should not make the loan.

6
Sovereign Risk

• Debt repudiation
• Loan repudiation is an outright cancellation of
  all a borrowers current and future foreign debt
  and equity obligations. The borrower refuses to
  make any further payments of interest and
  principal.
• Since WW II, only China (1949), Cuba (1961) and
  North Korea (1964) have repudiated debt.
• Rescheduling
• Loan rescheduling refers to temporary
  postponement of payments during which time new
  terms and conditions are agreed upon between the
  borrower and lenders. In most cases, these new
  terms are structured to make it easier for the
  borrower to repay.
• Most common form of sovereign risk.
• South Korea, 1998
• Argentina 2001(severe ongoing economic problems
  at time of writing)

7
Debt Rescheduling

• More likely with debt financing rather than bond
  financing.
• Loans usually are made by a small group
  (syndicate) of banks as opposed to bonds that are
  held by individuals and institutions that are
  geographically dispersed. Even though bondholders
  usually appoint trustees to look after their
  interests, it has proven to be much more
  difficult to approve renegotiation agreements
  with bondholders in contrast to bank syndicates.
• The group of banks that dominate lending in
  international markets is limited and hence able
  to form a cohesive group. This enables them to
  act in a unified manner against potential
  defaults by countries.
• Many international loans, especially those made
  in the post-war period, contain cross-default
  clauses, which make the cost of default very
  expensive to borrowers. Defaulting on a loan
  would trigger default clauses on all loans with
  such clauses, preventing borrowers from
  selectively defaulting on a few loans.
• In the case of post-war loans, governments were
  reluctant to allow banks to fail. This meant that
  they would also be actively involved in the
  rescheduling process by either directly providing
  subsidies to prevent repudiations or providing
  incentives to international agencies like the IMF
  and World Bank to provide other forms of grants
  and aid.

8
Country Risk Evaluation

• Outside evaluation models
• The Euromoney Index
• The Euromoney Index was originally published as
  the spread of the Euromarket interest rate for a
  particular countrys debt over LIBOR. The index
  was adjusted for volume and maturity. The index
  recently has been replaced by a large number of
  subjectively determined economic and political
  factors.
• The Economist Intelligence Unit ratings
• The EIU rates country risk by combined economic
  and political risk on a 100 point scale. The
  higher the number, the worse the sovereign risk
  rating of the country.
• Institutional Investor Index
• The Institutional Investor Index is based on
  surveys of the loan officers of major
  multinational banks who subjectively give
  estimates of the credit quality of given
  countries. The scores range from 0 for certain
  default to 100 for no probability of default.

9
What about historical premiums for other markets?

• Historical data for markets outside the United
  States tends to be sketch and unreliable.
• Ibbotson, for instance, estimates the following
  premiums for major markets from 1970-1996
• Country Annual Return on Annual Return on
  Equity Risk Premium Equity Government bonds
• Australia 8.47 6.99 1.48
• France 11.51 9.17 2.34
• Germany 11.30 12.10 -0.80
• Italy 5.49 7.84 -2.35
• Japan 15.73 12.69 3.04
• Mexico 11.88 10.71 1.17
• Singapore 15.48 6.45 9.03
• Spain 8.22 7.91 0.31
• Switzerland 13.49 10.11 3.38
• UK 12.42 7.81 4.61

10
Assessing Country Risk Using Currency Ratings
Latin America - June 1999

• Country Rating Default Spread over US T.Bond
• Argentina Ba3 525
• Bolivia B1 600
• Brazil B2 750
• Chile Baa1 150
• Colombia Baa3 200
• Ecuador B3 850
• Paraguay B2 750
• Peru Ba3 525
• Uruguay Baa3 200
• Venezuela B2 750

11
Country risk
12
Country risk

• Countries with track records of sound economic
  policymaking such as Singapore, Hong Kong and
  Chile have traditionally featured among the
  best-rated countries. Several Gulf oil producers
  are more recent entrants to this group,
  reflecting the positive impact of the oil bonanza
  on growth, public finances and their external
  accounts.
• The worst rated countries are dragged down by
  poor payment records, institutional failings and,
  in some cases (such as Iraq and Sudan) by civil
  violence. As well as payment arrears, Zimbabwe's
  poor rating reflects mismanagement which has
  brought the economy to the brink of collapse.
  Ecuador's low rating reflects doubts about
  President Rafael Correa's willingness to service
  the country's external bonds rather than a lack
  of capacity to pay. The government is currently
  in a comfortable financial position owing to high
  oil prices.

13
JP Morgan EMBI Index
14
Country Risk Evaluation

• Internal Evaluation Models
• Statistical models
• Country risk-scoring models based on primarily
  economic ratios.
• The analyst uses past data on rescheduling and
  nonresheduling countries to see which variables
  best discriminate between those countries that
  rescheduled their debt and those that did not.
• To develop a CRA-Z acore

15
Statistical Models

• Commonly used economic ratios
• Debt service ratio (Interest amortization on
  debt)/Exports
• The debt service ratio (DSR) divides interest
  plus amortization on debt by exports. Because
  interest and debt payments normally are paid in
  hard currencies generated by exports, a larger
  ratio is interpreted as a positive signal of a
  pending debt rescheduling possibility.
• Import ratio Total imports / Total FX reserves
• The import ratio (IR) divides total imports by
  total foreign exchange reserves. A growing
  amount of imports relative to FX reserves
  indicates a greater probability of credit
  restructuring. This ratio is positively related
  to debt rescheduling.

16
Statistical Models

• Investment ratio Real investment / GNP
• The investment ratio (INVR) measures the
  investment in real or productive assets relative
  to gross national productive. A larger
  investment ratio is considered a signal that the
  country will be less likely to require
  rescheduling in the future because of increased
  productivity thus the relationship is negative.
  However, because the bargaining position of the
  country will be enhanced, some observers feel
  that the relationship is positive. That is, a
  stronger ratio gives the country more power to
  request, even demand, rescheduling to achieve
  even better terms on its debt

17
Statistical Models

• Variance of export revenue
• Export revenues are subject to both quantity and
  price risk due to demand and supply factors in
  the international markets. Increased variance is
  interpreted as a positive signal that
  rescheduling will occur because of the decreased
  certainty that debt payments will be made on
  schedule.
• Domestic money supply growth
• Rapid domestic money supply growth indicates an
  increase in inflationary pressures that typically
  means a decrease in the value of the currency in
  international markets. Thus, real output often
  is negatively impacted, and the probability of
  rescheduling increases.

18
Example

• An FI manager has calculated the following values
  and weights to assess the credit risk and
  likelihood of having to reschedule the loan.
  From the Z-score calculated from these weights
  and values, is the manager likely to approve the
  loan? Validation tests of the Z-score model
  indicated scores below 0.500 likely to be
  nonreschedulers, while scores above 0.700
  indicated a likelihood of rescheduling. Scores
  between 0.500 and 0.700 do not predict well.
• Country
• Variable Value Weight
• DSR 1.25 0.05
• IR 1.60 0.10
• INVR 0.60 0.35
• VAREX 0.15 0.35
• MG 0.02 0.15
• Z 0.05DSR 0.15IR 0.30INVR 0.35VAREX
  0.15MG
• 0.05(1.25) 0.15(1.60) 0.30(0.60)
  0.35(0.15) 0.15(0.02)
• 0.488
• This score classifies the borrower as a probable
  nonrescheduler.

19
Problems with Statistical CRA Models

• Measurements of key variables
• Measuring the variables accurately and in a
  timely manner often is difficult because of data
  accessibility.
• Population groups
• The choice of rescheduling or not rescheduling
  often is not a dichotomous situation. In effect,
  many other payment alternatives may be available
  through negotiation.
• Finer distinction than reschedulers and
  nonreschedulers may be required.
• Political risk factors are extremely difficult to
  quantify.
• Strikes, corruption, elections, revolution.
• Portfolio aspects
• The portfolio affects of lending to more than one
  country are not considered. Thus the true amount
  of systematic risk added to the portfolio may be
  less than estimated by evaluating the
  rescheduling probability of countries
  independently.

20
Problems with Statistical CRA Models (continued)

• Incentive aspects of rescheduling
• Statistical models are ill-prepared or designed
  to evaluate the incentives of both the borrowers
  and the lenders to negotiate a rescheduling of
  the debt. Borrowers benefit by lowering the
  present value of future payments at the expense
  of reducing the openness of the market to future
  borrowing as well as withstanding potentially
  adverse effects on trade. Lenders benefit by
  avoiding a possible default, collecting
  additional fees, and perhaps realizing tax
  benefits. Lenders, however, may also be subject
  to greater scrutiny by regulatory authorities and
  may have permanent changes in the maturity
  structure of their asset portfolios.
• Stability
• Many of the key variables suffer from the problem
  of stability. That is, predictive performance in
  the past may not be good indicators of predictive
  performance in the future.

21
Using Market Data to Measure Risk

• Secondary market for LDC debt
• Since the mid-1980s, a secondary market for
  trading LDC debt has developed among large
  commercial and investment banks in New York and
  London.
• Sellers
• The primary sellers of LDC debt include large FIs
  who are willing to accept write-downs of loans
  and small FIs who no longer wish to be involved
  with the LDC market.
• Buyers
• Buyers tend to be wealthy investors, hedge funds,
  FIs, and corporations who wish to use debt-equity
  swaps to further investment goals or speculative
  investments.

22
Using Market Data to Measure Risk

• Market segments
• Brady Bonds
• Brady bonds are recollateralized loans that have
  lower coupon interest rates and longer maturities
  than the original loans. The principal usually
  is collateralized with the purchase of U.S.
  treasury bonds by the issuing country. Although
  yields are lower, the Brady bonds have more
  acceptability in the secondary markets than the
  original loans.
• Sovereign Bonds
• Sovereign bonds constitute the second largest
  segment of the LDC debt market. These bonds are
  issued to repay Brady bonds, and thus they have
  higher credit risk premiums because they no
  longer have the cost of the U.S. treasury
  collateral.

23
Using Market Data to Measure Risk

• Performing LDC loans
• Performing loans are the original or restructured
  sovereign loans on which the originating country
  continues to remain current in the payment of
  interest and principal.
• Nonperforming LDC loans
• Nonperforming loans are traded in the secondary
  markets at deep discounts because of nonpayment
  situations.

24
Key Variables Affecting LDC Loan Prices

• Most significant variables
• Debt service ratios
• Import ratio
• Accumulated debt arrears
• Amount of loan loss provisions