A Critical Assessment of An Existing Debt ProposalComments

1
A Critical Assessment of An Existing Debt
ProposalComments

• Atish Ghosh
• Deputy Chief, Policy Review Division, Policy
  Development and Review Department, International
  Monetary Fund
• 202-623-6288 aghosh_at_imf.org

2
NPV of Debt for 27 Decision Point Countries
(Billions of US, 2002 NPV)
3
Poverty-Reducing Expenditures and Debt Service in
27 Decision Point Countries
4
Gross Flows of Official External Resources to 27
Decision Point Countries (in billions of US)
5
Net Flows of Official External Resources to 27
Decision Point Countries (billions of US,
percent of GDP)
6
Topical Issue

• As low-income countries reach Decision and
  eventually Completion Points under HIPC, need a
  prudent borrowing strategy going forward.
• Future lending by the international community
  should be geared to the countrys capacity to
  carry debt. If more resources are required (e.g.
  to help reach MDGs), these should be in the form
  of grantsnot lending that leads to an
  unsustainable debt burden.

7
Getting it Right is Important

• The Challenge Helping low-income countries
  achieve their development goalse.g. MDGswhile
  avoiding excessive debt burdens that stifle
  growth or require socially costly cuts of
  expenditure in the future.
• Too permissive Future debt-servicing problems
  that may be very costly in terms of development
  goals.
• Too conservative Deny resource transfer that
  could otherwise have been used productively.

8
Work in Progress

• IMF and World Banktogether with various
  stakeholdersworking on formulating a framework
  that can help inform low-income countries on how
  much they should prudently borrow, and how much
  the BWIs should prudently lend.
• Workshops (Paris, Berlin, Accra, Washington)
  report to Boards in early 2004 with an
  operational framework.

9
What is Debt Sustainability?

• Two concepts
• SolvencyThe present value of surpluses that the
  country (or government) must (eventually)
  generate equals the present value of its debt.
• LiquidityRegardless of solvency, a country (or
  government) may be illiquid if its available
  financing and liquid assets are insufficient to
  meet its current expenses and maturing
  obligations.
• In practice, concepts are blurred.

10
What is Debt Sustainability (cont.)?

• For market borrowers, liquidity is usually the
  most important concept because they face
  roll-over risk.
• For low-income countries, solvency is usually the
  more critical concept, because they borrow from
  official lenders who would be more willing to
  roll over debt (as long as solvency is
  fulfilled)though they need to guard against
  defensive lending.

11
Elements of a Framework

• Take account of country circumstances
• Current levels of debt (post-relief, as
  applicable)
• Policies and institutional capacity
• Realistic prospects for growth and exports
• Vulnerability to shocks
• Although HIPC was based on a uniform threshold
  (for reasons of equity and transparency),
  individual borrowing strategies for countries
  should be geared to country specific
  circumstances.

12
Methodology

• Project debt and debt-service indicators for a
  given path of deficits and assumed grants.
• Assess whether, given both cross-country evidence
  and the countrys specific circumstances, the
  projected debt and debt-service burdens become
  excessive
• Apply systematic stress testing to the
  projections.
• If baseline or adverse scenario debt dynamics
  problematic, need to narrow deficits or seek more
  grant financing.

13
Some Thresholds
14
Practical Problems and Issues

• Which indicators? Numerator Nominal vs. NPV
  External debt, public external debt, total public
  debt debt stock or debt service?
• Denominator GDP, exports (remittances?), fiscal
  revenues?
• Total public or only public external? HIPC
  public external (international community cannot
  forgive debt that is owed to someone else!) But
  solvency of the public sector depends on total
  public debt. Data problems different terms,
  track separately?
• Balance of payments gaps vs. budget gaps--the
  transfer problem (Keynes)either or both
  constraints may be binding.

15
Practical Problems and Issues (cont.)

• Use projections or current ratios? (Long
  maturityslow build up of debt problemsneed
  projections but over-optimism of projections)
• How to establish threshold values? Empirical
  studiesproblems of interpretation (e.g. are
  reschedulings indication of debt distress or a
  form of transfer from the international
  community)? Inherent limitationsType I and Type
  II errors.
• How to combine indicators? Factor analysis?
  Interpretation? Binding constraint?

16
Practical Problems and Issues (cont.)

• Stress testing Too extremewould not want to
  tailor borrowing strategy to them. Too benignmay
  miss important risks.

17
Specifics of Issues Paper

• Use of nominal stock vs. NPV
• Need to work with NPVs (despite shortcomings)
  because even concessional debts can have very
  different terms, e.g. PRGF (5-10 yrs repayment)
  vs. IDA (30-50 yrs repayment).
• Inclusion of domestic debt
• Definitely want to track total public debt (data
  problems!) but probably not aggregate (different
  dynamics)
• Focus on government debt/revenues and external
  debt/exports
• Look at various indicators, if one is out of
  line, it might point to a problem with the
  denominator (e.g. need to improve tax
  administration)

18
Specifics of Issues Paper (cont.)

• Aggregate indicators using principal components
• Problems of interpretation
• Thresholds
• Country- and time-specific thresholdslimitations
  of empirical work (what event is being modeled?
  What is an acceptable probability of debt
  distress?)
• Split concessional loans into grant plus
  non-concessional loan
• Provide grants instead of loans to countries with
  debt sustainability problemsbut ... it limits
  the resource transfer to low-income countries.

19
Conclusions

• Various technical/methodological issues and
  trade-offs.
• Analytical framework can help inform judgments
  about debt sustainability but cannot substitute
  for such judgments.
• But fair degree of concensus on basic approach
  and the need to avoid a recurrence of debt
  servicing problems.
• Explore other ways to share riske.g. Hausmann
  proposal or commodity price-linked debt service.

20
Conclusions

• Getting it right is important ...but not easy!

21
NPV of Debt of Countries that reached Completion
Point by mid-2003 (billions of US, 2002 NPV)