Sovereign risk is the risk that a government is unable or unwilling to meet its own financial obligations as agreed. It is a form of credit risk: holding a government bond does not, by itself, establish that every promised payment will arrive in full and on time. For a portfolio model, the issuer's identity matters alongside the interest-rate exposure of the bond.
The term is used here for sovereign credit risk. Broader country risk can affect private businesses and investments even while the government continues servicing its debt. This distinction follows the OCC's Country Risk Management handbook, which treats the government's ability and willingness to pay separately from the wider country environment.
Probability and recovery answer different questions
A sovereign assessment needs both a view of whether agreed payments will be missed or altered and a view of what creditors might recover if that happens. A probability of default requires a defined event and horizon. A recovery rate measures the portion of a specified claim recovered under a stated valuation convention. Neither quantity determines the other.
Consider a wholly fictional sovereign borrower with an outstanding claim of USD 20 million. Stipulate a debt restructuring that delivers USD 8 million of value to its creditor. Both amounts are measured on the same valuation date, and the example excludes additional costs and any other receipts. The recovery value may represent cash or the value of replacement instruments; it need not be immediately spendable cash.
The shortfall is 60% of the specified claim. Under this example's exposure and recovery convention, that is the scenario's loss given default.
Original fictional recovery ledger, in USD million. Bar widths use a common scale and both amounts share one valuation date. Recovery value can include replacement debt; it is not assumed to be cash received immediately.
A 40% recovery is not a 40% chance of default. No event probability has been assigned, so the example supplies a conditional severity rather than an expected-loss estimate. The USD 12 million shortfall is also not automatically the investor's trading loss: that depends on the price paid, prior income and the comparison being made.
The denominator deserves attention. Comparing the value received with an outstanding claim differs from comparing the present values of the old and new payment streams. Future coupons and principal cannot simply be added and compared with today's cash. Grigorian's IMF working paper on domestic sovereign restructuring, p. 21, explains alternative haircut conventions and their sensitivity to the discount rate. The arithmetic above uses only the stipulated same-date values; it predicts no recovery for an actual country.
For the definitions behind exposure, default probability and loss severity, Credit Instruments, PD/LGD/EAD, and Rating Migration provides further teaching (public preview; full lesson requires membership). Applying those ingredients to a sovereign requires assumptions about that issuer and the particular obligation.
Separate the payment failure from neighbouring risks
| Situation | Main question |
|---|---|
| The government misses or restructures its own debt payments | What value and payment timing will the sovereign creditor receive? |
| A private borrower can pay locally, but government restrictions obstruct conversion or remittance | Can the payment currency be obtained and transferred? This is a convertibility or transfer-risk channel within broader country risk. |
| The payment arrives as agreed, but the currency loses value in the investor's reporting currency | What is the foreign-exchange risk? |
| A government bond's market value falls as yields change, before any missed payment | What market risk has the holder incurred? |
These channels can reinforce one another. A currency shock can impair a borrower's foreign-currency debt service, and a sovereign credit deterioration can change bond prices before a default occurs. The table identifies mechanisms; it does not create mutually exclusive loss buckets.
Model the obligation, not just the country label
Start with the actual debtor, claim amount, payment currency and schedule. A state-owned company is not automatically the sovereign obligor: inspect who owes the payment and any guarantee being relied on. Then state the stress event, recovery convention, valuation date and treatment of delayed receipts. Contract terms and applicable law matter to the outcome; no single restructuring rule is assumed here.
At portfolio level, identify exposures that share the same sovereign borrower or depend on the same government guarantee. Several securities need not provide independent protection against that common event. Separate the direct sovereign claim from private exposures affected through country conditions, and avoid counting the same economic loss twice when combining credit, currency and market scenarios.
Sources
- OCC, Country Risk Management, Comptroller's Handbook, version 1.0, February 2016, updated March 2025, pp. 1–5 and 37–38: sovereign credit risk, country risk, convertibility and transfer risk. This is US banking-supervision guidance; the distinctions here do not assert a universal legal classification.
- David A. Grigorian, Restructuring Domestic Sovereign Debt: An Analytical Illustration, IMF Working Paper 23/24 (2023), p. 21: alternative recovery/haircut denominators and the choice of discount rate. This is research by the author, not a binding IMF rule.
- Alexander J. McNeil, Rüdiger Frey and Paul Embrechts, Quantitative Risk Management: Concepts, Techniques and Tools, revised edition (Princeton University Press, 2015), §10.1.1–10.1.2, pp. 367–368: sovereign borrowers, credit-risky bonds, recoveries and the distinction between spread and default risk.