This case traces a sovereign debt crisis. A government that borrows in foreign currency sees its debt ratio jump when its own currency falls: a 30% fall lifts dollar debt worth 30% of GDP to 42.86% overnight. The shock then spreads through auctions, ratings, forced selling and wider spreads. A company that owes dollars on a known date can use a forward priced by interest-rate parity to turn an open-ended loss into a known cost.
Why it matters: A hedge turns an unlimited risk into a price you can plan for.
Summary: A government that borrows in foreign currency sees its debt ratio jump when its currency falls — a 30% fall lifts dollar debt worth 30% of GDP to 42.86% overnight — and the shock spreads through auctions, ratings, forced selling and wider spreads. For a company owing dollars on a known date, a forward priced by interest rate parity turns an open-ended loss into a known cost.
- Foreign-currency debt cannot be repaid by creating money, which is why currency mismatches sit at the center of most emerging-market crises.
- The forward rate reflects the interest differential: 50 × 1.12 ÷ 1.05 = 53.33, a price, not a forecast.
- Hedging costs 8.9% of equity if nothing happens; not hedging costs 57.1% of equity if the currency falls 30%.
- Hedge the exposure, never more; matching dollar revenue on the same dates is a natural hedge.
- Currencies outside CLS, such as the rupee, settle leg by leg and carry settlement risk.
The situation: Country X (illustrative), an emerging-market economy, has government debt of 60% of GDP, half of it in US dollars. Its currency trades at 50 local currency units (LCU) per dollar. Investor confidence breaks and the currency loses 30% of its value. The decision belongs to the treasurer of a Country X manufacturer that imports components, earns in LCU, and owes $40 million due in one year. The question, asked before the crisis: hedge or not?
Why the debt ratio jumps overnight. The dollar debt is 30% of GDP at 50 LCU per dollar. A 30% fall in the LCU means a dollar now costs 50 ÷ 0.7 = 71.43 LCU, so the same dollar debt is worth 30% × (71.43 ÷ 50) = 42.86% of GDP in local terms. Total debt goes from 60% to 30% + 42.86% = 72.86% of GDP — up 12.86 points with no new borrowing. Local-currency debt does not jump this way, and a government can in the last resort create its own currency to pay it, which is why foreign-currency debt sits at the center of most emerging-market crises — most, not all, as Part 3.8: Local vs Foreign Currency Debt — Revisiting the Core Distinction explains. This is the currency mismatch behind Sri Lanka’s 2022 default (Volume I §9.4) and the 1997 Asian crisis (Volume II §10.3).
How the crisis spreads. At the next auction (Part 3.2: The Sovereign Debt Issuance Process), bids thin and the auction clears at a higher yield. If the government must roll over debt equal to 15% of GDP and its yield rises from 9% to 14%, its annual interest bill rises by 15% × 5 points = 0.75% of GDP, and r climbs while g falls — the debt arithmetic of Case One with worse numbers (Part 3.7: Debt Sustainability Analysis — r versus g). A downgrade from BBB− to BB+ (Part 3.6: Sovereign Credit Ratings) crosses the investment-grade line, forcing sales by funds that may hold only investment-grade debt (Volume II’s Part 3). Foreign holders sell first (Part 3.5: Who Holds Sovereign Debt, and Why It Matters). Dealers widen bid-ask spreads on the bonds because the next seller may know more than they do (Part 5.3: Market Makers and the Bid-Ask Spread) — the adverse selection that widened stock spreads in Case One. The usual way out is the one Sri Lanka took: after defaulting in 2022, it secured a $3 billion IMF Extended Fund Facility in March 2023 (Volume I §9.4). The other figures in this case are illustrative.
Pricing the hedge. The one-year forward exchange rate is set by interest rate parity: F = S × (1 + iLCU) ÷ (1 + iUSD) = 50 × 1.12 ÷ 1.05 = 53.33 LCU per dollar, with LCU deposits paying 12% and dollar deposits 5%. The forward is 6.67% above spot because of the interest differential, not because anyone forecasts a 6.67% fall; a forward is a price, not a prediction (Volume II §4.1). A bank that sold the forward at any other price could be arbitraged by borrowing in one currency and depositing in the other.
Suppose a bank quoted the one-year forward at 52.00 instead of 53.33. A trader borrows $1 million at 5%, owing $1.05 million in a year; converts it at spot into LCU 50 million and deposits it at 12%, which grows to LCU 56 million; and today buys dollars forward at 52.00, so the LCU 56 million becomes 56 ÷ 52 = $1,076,923 at maturity. After repaying $1.05 million, $26,923 remains, with no exchange-rate risk taken. If the bank quoted 55.00, the trade runs the other way: borrow LCU 50 million at 12% (owing LCU 56 million), convert it to $1 million, deposit at 5% to receive $1.05 million, and sell that forward at 55.00 for LCU 57.75 million, leaving LCU 1.75 million. Only at 53.33 do both trades earn nothing, which is why banks quote that price and why the forward premium measures the interest differential rather than a forecast. Small gaps do persist: Du, Tepper and Verdelhan (“Deviations from Covered Interest Rate Parity,” Journal of Finance, 2018) document lasting deviations among major currencies since 2008, which they tie to the balance-sheet costs regulation places on banks. For the treasurer those gaps are far smaller than the 6.67% premium the interest differential creates.
Cost in LCU of paying $40 million in one year, unhedged or hedged with a forward at 53.33. The company’s equity is LCU 1,500 million.
| Scenario in one year | LCU per dollar | Cost unhedged | Cost hedged | Hedge gain or (cost) |
|---|---|---|---|---|
| LCU stable | 50.00 | LCU 2,000.0m | LCU 2,133.3m | (LCU 133.3m) |
| LCU falls 6.25% | 53.33 | LCU 2,133.3m | LCU 2,133.3m | LCU 0 |
| LCU falls 30% | 71.43 | LCU 2,857.1m | LCU 2,133.3m | LCU 723.8m |
The flip point is a fall in the LCU of 1 − 50 ÷ 53.33 = 6.25% (the dollar rising 6.67%). The asymmetry decides it: if nothing happens, the hedge costs LCU 133.3 million, 133.3 ÷ 1,500 = 8.9% of equity; unhedged in the crash, the payment costs LCU 857.1 million more than at today’s spot, 857.1 ÷ 1,500 = 57.1% of equity. Decision and outcome: the treasurer hedges exactly $40 million for the payment date — the exposure, not more (Volume II §4.8). When the LCU falls 30%, the company pays LCU 2,133.3 million instead of 2,857.1 million and keeps its equity largely intact.
Settlement and operations (Part 8: Operational Risk Governance & Reconciliation). The LCU, like the Indian rupee, is not one of the 18 currencies CLS settles payment-versus-payment, so at maturity the company pays LCU and receives dollars through correspondent accounts, leg by leg (Part 8.4: STP Rate, Tolerance, and Matching). Between the two payments one side has paid and the other has not, which is settlement (Herstatt) risk; in a crisis, when banks’ credit is in question, that gap is when it bites. If the forward is collateralized, its daily value also drives margin: as the LCU falls, the company is in the money and the bank owes it collateral, and the treasury must reconcile each call against its own valuation the same day. An unreconciled call is a credit exposure the company does not know it has.
Ideas (Part 9.8: Modern Monetary Theory and Today’s Debates). Modern Monetary Theory’s central claim — that a government issuing and borrowing in its own currency faces real-resource and inflation limits rather than purely financial ones — does not reach Country X’s dollar debt, which it cannot issue. Whatever one makes of the argument inside its own premise, a currency mismatch is outside it.
What the treasurer would have faced unhedged. The LCU 857.1 million loss arrives in the same months that local banks, holding more of the government’s repriced bonds, cut credit lines. That double hit — a larger bill and less funding to pay it — is why foreign-currency mismatches turn currency falls into corporate defaults, and corporate defaults back into sovereign stress.
If you owe foreign currency on a known date and earn in local currency, hedge the known amount for that date when the hedge’s cost if nothing happens (the interest differential, here 8.9% of equity) is smaller than the loss you could not absorb in a severe depreciation (here 57.1% of equity at −30%). Size the hedge to the exposure, never above it; the excess is a speculative position. Subtract any matching dollar revenue on similar dates, which is a natural hedge, and hedge only the net exposure, skipping the hedge only when that revenue covers the whole payment; check before trading whether the currency settles through CLS or leg by leg.
Leaving a known dollar payment unhedged because the forward “looks expensive.” The forward’s premium over spot reflects the interest differential, so it looks like a cost every year until the year it is not. In this case, saving LCU 133.3 million (8.9% of equity) in a calm year meant risking LCU 857.1 million (57.1% of equity) in a crash — and the crash comes at the moment refinancing is hardest, because the same confidence shock that sinks the currency closes local credit markets. Avoid it by deciding the hedge policy in advance, in writing, by exposure and date, not by view.
Why is foreign-currency debt riskier for governments than local-currency debt?
Because a fall in the currency raises its burden instantly and the government cannot create the currency to repay it. In this case a 30% fall lifted dollar debt from 30% to 42.86% of GDP with no new borrowing. Local-currency debt can still default or be inflated away, but it does not jump with the exchange rate.
How is a currency forward rate calculated?
By interest rate parity: forward = spot × (1 + local interest rate) ÷ (1 + foreign interest rate) for one year. At 50 LCU per dollar, with 12% local and 5% dollar rates, the one-year forward is 50 × 1.12 ÷ 1.05 = 53.33. The difference reflects interest rates, not a market forecast of the currency.
Does a downgrade below investment grade force investors to sell?
Often, for some of them. Funds whose mandates or index rules allow only investment-grade bonds must sell when a bond falls from BBB− to BB+, whatever their own view. That mechanical selling widens spreads further, which is why the line matters so much for sovereigns close to it.
Can a government default on debt in its own currency?
Yes, although it is rarer. A government can always create its own currency, but it may choose default over the inflation that would cause, as Russia did on its ruble debt in 1998, and members of a currency union such as the eurozone cannot create the currency at all (Part 3.8: Local vs Foreign Currency Debt — Revisiting the Core Distinction).
A currency fall raises the burden of foreign-currency debt at once — from 60% to 72.86% of GDP in Country X with no new borrowing — and the stress spreads through weak auctions, downgrades across the investment-grade line, foreign selling and wider spreads. A treasurer who owes dollars on a known date can hedge with a forward priced by interest rate parity, paying a known cost (8.9% of equity) to avoid a loss of 57.1% of equity. Settlement outside CLS and collateral calls on the hedge add operational risks that only same-day reconciliation catches.
Six questions on this chapter. Decide on your answer first, then click “Reveal Answer.”
1. Spot is 80 local units per dollar, one-year local deposits pay 9% and dollar deposits 4%. What is the one-year forward rate?
- 84.00
- 80.00
- 76.33
- 83.85
Reveal Answer
Answer: D. F = 80 × 1.09 ÷ 1.04 = 83.85. 76.33 inverts the rates; 84.00 simply adds the 5-point differential; 80.00 ignores interest rate parity.
2. A government’s debt is 50% of GDP: 30% in local currency and 20% in dollars. Its currency falls 20% against the dollar. What is the new debt ratio, with no new borrowing?
- 54% of GDP
- 50% of GDP
- 55% of GDP
- 60% of GDP
Reveal Answer
Answer: C. A 20% fall means a dollar costs 1 ÷ 0.8 = 1.25 times as much local currency, so dollar debt becomes 20 × 1.25 = 25% of GDP; 30 + 25 = 55%. Multiplying by 1.2 (54%) understates it.
3. A company owes $10 million in one year and expects $4 million of dollar revenue on the same date. Under the chapter’s rule, how much should it hedge?
- The $6 million net exposure
- The $14 million gross flows
- The full $10 million payment
- Nothing, since revenue offsets it
Reveal Answer
Answer: A. The $4 million of matching revenue is a natural hedge, so the exposure is 10 − 4 = $6 million. Hedging more than the exposure creates a speculative position.
4. A treasurer settles a rupee–dollar forward at maturity. Why does settlement risk remain even if the trade was priced correctly?
- Payment-versus-payment settlement delays the dollar leg a day
- The rupee is outside CLS, so the two legs settle separately
- Correspondent banks net the legs and can reverse the payment
- The forward is cleared by a central counterparty that may fail
Reveal Answer
Answer: B. CLS settles 18 currencies payment-versus-payment and the rupee is not one of them, so one side can pay while the other has not (Herstatt risk). CLS is not a central counterparty.
5. Worked problem: Foreign-currency debt is 40% of GDP and the currency falls 25%. What is the debt ratio?
Reveal Answer
Answer: 40% ÷ 0.75 = 53.3% of GDP.
6. Worked problem: Spot is 40 local units per dollar, local rates are 9% and dollar rates 4%. What is the one-year forward, and is it a forecast?
Reveal Answer
Answer: F = 40 × 1.09 ÷ 1.04 = 41.92. It is a price set by the interest differential, not a forecast.
- IMF press release 23/79, Sri Lanka Extended Fund Facility, March 20, 2023 — $3 billion program after the 2022 default (Part 10.4)