Emerging-market bonds: hard-currency versus local-currency sovereign debt

What you will learn

• How hard-currency and local-currency emerging-market (EM) sovereign bonds differ in who carries the currency risk and the default risk.
• How to read a credit spread and compare it with a rough expected default loss.
• What the default history of EM sovereigns teaches about recoveries and restructurings.
• How the main EM bond index families and ETFs are built and where their risks hide.

Explanation

Bond basics (yield to maturity, duration, credit spreads) are covered in the course "Bonds & Fixed Income in Depth"; currency risk of foreign bonds for euro investors is covered there too. This lesson applies those tools to sovereign bonds issued by emerging-market governments.

Hard currency versus local currency

A hard-currency bond is issued by an EM government in a foreign currency, usually US dollars, sometimes euros. The government must earn or buy dollars to pay coupons. The investor carries the default risk (the country cannot or will not pay) but no local-currency risk.

A local-currency bond is issued in the country's own currency, for example Brazilian real or Indian rupee. The government can always print its own currency, so outright default is rarer, but the investor carries the currency risk: inflation and depreciation can erase a high coupon.

Spreads and expected loss

The credit spread is the extra yield of a hard-currency bond over a US Treasury of similar maturity, measured in basis points (1 bp = 0.01 percentage points). A rough test: compare the spread with an expected annual default loss, defined as the probability of default per year multiplied by the loss given default (one minus the recovery rate). A spread well above the expected loss pays for the risk and for illiquidity; a spread below it means the bond is expensive.

Default history

EM sovereign defaults are not rare. Examples include Russia (1998, domestic debt), Argentina (2001, 2014 and 2020), Ecuador (2008 and 2020), Lebanon (2020), Zambia (2020), Sri Lanka (2022) and Ghana (2022). Russia's foreign bonds also went into default in 2022 when sanctions blocked payments. Recoveries vary widely: some restructurings returned investors most of the face value after a delay, others only a fraction. Three lessons: defaults cluster in strong-dollar years; restructurings take years; and prices fall far below the final recovery during the crisis, so forced sellers lose most.

Index families and ETFs

Most EM bond ETFs sold in Germany track one of two families. Hard-currency ETFs usually follow a J.P. Morgan EMBI-type index (Emerging Markets Bond Index); local-currency ETFs follow a GBI-EM-type index (Government Bond Index – Emerging Markets). "Diversified" versions cap large issuers so that one country cannot dominate. Points to check in a factsheet: average credit rating (hard-currency indices mix investment grade and high yield), effective duration (often 6 to 8 years), yield to maturity, the share of the top five countries, and whether a euro-hedged share class exists. Dollar hedging costs roughly the euro–dollar short-rate difference, as explained in the currency lesson of this course.

Worked example

Hypothetical example: a euro investor compares two bonds of the same fictitious country, "Country X". All numbers are invented and rounded; fees, bid-ask spreads and taxes are excluded.

• Bond A: 7-year US-dollar bond, yield to maturity 7.5 %. The 7-year US Treasury yields 4.0 %.
• Bond B: 7-year local-currency bond, yield 11.0 %. The investor expects the local currency to lose 5 % per year against the dollar.

Step 1 – spread of Bond A: 7.5 % − 4.0 % = 3.5 percentage points = 350 bp.

Step 2 – expected default loss: assume a 3 % probability of default per year and a 50 % recovery. Expected loss = 3 % × (1 − 0.50) = 1.5 % per year. The spread of 3.5 % exceeds this by 2.0 percentage points. That 2 % pays for volatility, illiquidity and optimistic assumptions.

Step 3 – Bond B in dollar terms: (1 + 0.11) × (1 − 0.05) − 1 = 1.11 × 0.95 − 1 = 0.0545, about 5.5 % per year. Roughly the same return as Bond A, with currency risk instead of default risk.

Step 4 – a bad year for Bond A: if the spread widens by 200 bp and the bond's duration is 6, the price falls by about 6 × 2.0 % = 12 %. With 7.5 % of carry, the year ends at roughly −4.5 % in dollars before the exchange rate.

Step 5 – the euro view: +7.5 % in dollars with a 4 % dollar fall gives 1.075 × 0.96 − 1 = 3.2 % in euros.

Common mistake

Practice activity

On paper: a hard-currency bond yields 8.2 % against a US Treasury at 4.2 %. You estimate a 4 % annual default probability and a 40 % recovery. The same country's local-currency bond yields 13 %, and you expect 7 % annual depreciation against the dollar. (a) Calculate the spread and the expected default loss. Is the investor paid for the risk? (b) Calculate the approximate dollar return of the local bond. (c) The hard-currency bond has a duration of 7; the spread widens by 150 bp during the year. Estimate the dollar total return for that year.

Key takeaways

• Hard-currency bonds carry default risk; local-currency bonds carry currency and inflation risk. Neither is "safe".
• Compare the spread with probability of default × loss given default before calling a yield attractive.
• Defaults cluster in strong-dollar years and restructurings take years; prices overshoot the final recovery.
• Read duration, average rating and top-country shares in an EM bond ETF factsheet, and decide consciously about dollar hedging.

Next step

Next, we look at the companies behind EM equity indices: state-owned enterprises, dual-class shares and what minority shareholders can really expect.