Quant Memo
Core

Emerging Market Hard vs Local Currency Debt

A Brazilian government bond denominated in dollars and one denominated in reais can carry wildly different yields for the same issuer, because one asks you to bear only credit risk while the other stacks currency and local rates risk on top.

Prerequisites: The Money Market and the Short End of the Curve, The FX Carry Trade

Brazil issues government debt in two flavors: bonds denominated in US dollars, and bonds denominated in Brazilian reais. Both come from the same borrower, backed by the same government, but they are not remotely the same trade. The dollar bond promises to pay dollars no matter what the real does — the only way you lose money (beyond interest-rate moves) is if Brazil actually fails to pay. The real bond promises to pay reais, and a US or European investor buying it is also betting on the real not collapsing against the dollar between now and maturity. Same country, two completely different risk exposures wearing the label "emerging market debt."

Hard-currency EM debt (usually USD, tracked by indices like the EMBI Global) isolates credit and sovereign risk — the same currency mismatch a lender faces on any dollar loan to a foreign borrower. Local-currency EM debt (tracked by the GBI-EM) adds currency risk and local interest-rate risk on top, which is usually the larger and more volatile component of total return.

Two different things that both get called "EM yield"

Hard-currency debt is priced as a spread over the equivalent-maturity US Treasury, because the dollar cash flows are directly comparable to a Treasury's — the entire extra yield is compensation for the chance Brazil restructures or defaults on dollar obligations, plus a liquidity premium. Local-currency debt is priced as an outright yield in the local market, driven by that country's own inflation, central bank policy, and fiscal credibility — and a foreign buyer's realized return depends enormously on where the currency goes, which has nothing to do with whether Brazil pays its bills on time.

Worked example: decomposing the extra yield

A 10-year US Treasury yields 4.20%. Brazil's 10-year dollar-denominated bond yields 5.50%. Brazil's 10-year real-denominated local bond yields 10.50%.

Credit spread (hard-currency risk).

5.50%4.20%=1.30%5.50\% - 4.20\% = 1.30\%

This 130 basis points is compensation for pure Brazilian sovereign credit risk — default or restructuring probability on dollar-denominated obligations, roughly comparable to a corporate credit spread.

Currency and local-rate premium.

10.50%5.50%=5.00%10.50\% - 5.50\% = 5.00\%

This 500 basis points is compensation for holding real-denominated cash flows instead of dollar ones — it reflects Brazil's higher domestic inflation, its own term premium, and the market's expectation (via uncovered interest parity, imperfectly) that the real will depreciate against the dollar over the holding period to offset some of that extra yield.

What actually happens to a dollar-based investor. If the real depreciates by 4% against the dollar over the year, the local bond's realized dollar return is roughly 10.50%4%=6.50%10.50\% - 4\% = 6.50\% — still ahead of the hard-currency bond's 5.50%, but with far more variance, since a 4% currency move is a routine, not extreme, annual outcome for the real. If the real instead depreciates 8% in a bad year, the realized dollar return falls to about 2.50%, underperforming the hard-currency bond despite the far higher headline yield.

UST 10y 4.20% Brazil USD 5.50% +1.30 credit Brazil BRL 10.50% +5.00 currency & rates
The same issuer's yield builds up in layers: Treasury base, then pure credit spread, then a much larger currency and local-rate premium.

Why local yields don't just get arbitraged away

If uncovered interest rate parity held exactly, the extra 5% yield on the real bond would be exactly offset by expected real depreciation, leaving no free lunch either way. In practice this parity holds only loosely and with huge variance — high-yielding EM currencies have historically drifted down against the dollar on average, but not by nearly enough to erase the yield pickup most years, which is the empirical basis of the EM local-currency carry trade, and also why it periodically produces sharp, currency-driven drawdowns when risk sentiment turns.

"Emerging market debt" as a single asset class is a category error for risk purposes. A portfolio manager who buys hard-currency Brazil to express a credit view and ends up also long local-currency Turkey for the yield has taken on a currency and local-rates bet they may not have intended — the two debt types respond to almost entirely different macro shocks (a global credit selloff hits hard-currency debt hardest; a dollar rally or local inflation surprise hits local-currency debt hardest).

Where it shows up

Global macro and EM-dedicated funds routinely trade the two markets against each other — for instance going long local-currency debt while hedging the currency exposure back to dollars with FX forwards, isolating the local-rates view from the currency view entirely (see The FX Carry Trade). Index providers track the two markets separately (EMBI Global for hard currency, GBI-EM for local currency) precisely because blending them would obscure which risk is actually driving returns.

Key terms

  • Hard-currency debt — EM sovereign or corporate debt denominated in a major reserve currency (usually USD); isolates credit risk.
  • Local-currency debt — EM debt denominated in the issuer's own currency; adds currency and local-rates risk.
  • Credit spread — the extra yield of hard-currency EM debt over the equivalent Treasury.
  • Uncovered interest rate parity — the (imperfect) theory that high-yield currencies should depreciate enough to offset their yield advantage.

Related concepts

Practice in interviews

Further reading

  • JPMorgan, EMBI Global and GBI-EM Index Methodology
  • Reinhart & Rogoff, This Time Is Different (ch. 6–7)
ShareTwitterLinkedIn