Beyond Carry: Turning EM Local Debt Macro Views into Portfolio Decisions

Illustrative table of contents:

1. Executive Summary

2. Portfolio View and Return Framework

   - Investment Thesis

   - Return Decomposition

   - Practical Brazil Example

3. Screening and Market Snapshot

   - Quantitative Screening Model

   - Market Snapshot

   - Interpreting EM Yield Curves

   - Central-Bank Reaction-Function Matrix

4. Fiscal, Investability and Country Views

   - Fiscal Policy Watchlist

   - Investability and Market Technicals

5. Country-by-Country PM Analysis

   - Brazil, Mexico, South Africa, Colombia

   - Factor Exposure Map

6. Portfolio Construction and Risk

   - Portfolio Construction Example

   - Risk Budgeting and DV01

   - Scenario Analysis

   - Performance Attribution Framework

7. Catalysts and Implementation

   - Catalyst Calendar and Monitoring Checklist

   - Implementation and Automation

   - Suggested Production Data Stack

8. Conclusion & PM Takeaways

   
   

1. Executive summary

This is not just a macro note - it is a portfolio-construction note.

Emerging-market local-currency debt still offers one of the most interesting combinations of carry, real yield and selective duration upside in global fixed income.

The current environment favours active country selection, curve positioning, risk budgeting and FX management.

Across the four markets in scope here (Brazil, Mexico, South Africa and Colombia) the common thread is that nominal policy rates remain high relative to inflation, but the source of that return differs materially:

·       Brazil offers the richest carry and the strongest interest rate easing optionality due to elevated rates;

·       Mexico offers more policy credibility and still-solid real rates, but a steeper curve shaped by long-term fiscal uncertainty;

·       South Africa offers attractive term premium with improving fiscal signalling, but higher FX beta;

·       Colombia offers high carry and potentially large upside from fiscal normalisation, but also the highest policy and fiscal risk.

The portfolio implication is a selective overweight to Brazil and Mexico, a more tactical allocation to South Africa, and a smaller, catalyst-driven Colombia exposure sized off risk rather than headline yield. In practice, that means harvesting carry where central banks still retain credibility, owning the part of the curve where policy repricing can work in the investor’s favour, and avoiding the temptation to treat every high-yielding local market as equivalent.

In EM local debt, the spread between a good macro view and a good portfolio is driven by: curve choice, DV01 allocation, currency hedge ratio, liquidity management and event-risk control.

This report includes real yields, curve dynamics, fiscal anchors and country views, while adding the portfolio-construction modules that matter for an institutional PM process: return decomposition, screening, reaction functions, investability filters, risk budgeting, scenario analysis, attribution and implementation.

The objective is not to predict every local market move, but to show a repeatable framework for turning macro views into portfolio decisions.

2. Portfolio view and return framework

Investment thesis. EM local debt remains attractive, but the correct expression is selective rather than index-like. High real rates across Brazil, Mexico, South Africa and Colombia provide income and a buffer against volatility, yet the dispersion in central-bank reaction functions, fiscal credibility and FX sensitivity means portfolio returns should be harvested through active allocation to carry, roll-down, duration and selected currency exposure rather than through EM country beta. Today that argues for a carry-and-belly bias in Brazil, a selective duration allocation in Mexico, cautious term-premium capture in South Africa, and only tactical, smaller-sized Colombia exposure until fiscal credibility and policy independence are clearer under the incoming administration.

Return decomposition. For EM local debt, total return can be decomposed into five components: carry, roll-down, duration, FX and changes in term premium. Carry is the income earned from holding the bond, roll-down is the price gain that accrues if a bond “slides” down a positively sloped curve, duration is the mark-to-market sensitivity to an outright move in yields, FX is the gain or loss from the currency, and term-premium repricing is the residual change in the long end driven by fiscal risk, issuance expectations, liquidity and risk appetite.

In developed markets, duration and monetary policy expectations usually dominate.

In EM local debt, FX and fiscal term premium often matter just as much.

Reuters’ coverage of Brazil after the June COPOM (Monetary Policy Committee of Brazil’s central bank) decision is a good example: short rates fell as the market priced continued easing, while longer rates rose because investors demanded more compensation for inflation and fiscal uncertainty.

Practical Brazil example. Suppose a PM owns a five-year Brazil local bond in the belly of the curve. The position earns high carry because Selic remains 14.25% and inflation is 4.8% on the latest mid-June reading. If the bond rolls from a five-year point toward the four-year point on a still-upward-sloping segment of the curve, that adds roll-down return. If front-end and intermediate yields fall as the easing cycle progresses, duration contributes positively. If the BRL remains stable or appreciates on high real carry and commodity support, FX adds further return. The offsetting risk is that long-end term premium widens if election-related fiscal stimulus or debt concerns intensify, which is exactly why a PM may prefer the 2–5 year sector over the unhedged long end.

3. Screening and market snapshot

Quantitative screening model. A PM-style process benefits from a disciplined screen before any discretionary country call. The scoring model below is illustrative rather than mechanical; its purpose is to organise judgement.

Factor	Weight	What it captures	Practical interpretation
Real yield	25%	Nominal rate less inflation	Carry buffer and foreign-demand support
Disinflation and reaction function	15%	Inflation path and central-bank credibility	Probability that carry turns into capital gains
Fiscal anchor	20%	Debt trajectory, budget credibility, issuance risk	Term-premium compression or widening
FX valuation and external buffer	15%	Currency support, reserves, current-account resilience	Likelihood of keeping local gains in hard currency terms
Curve shape	10%	Roll-down and relative-value opportunity	Best maturity bucket to express the view
Investability and liquidity	10%	Market depth, access, benchmark relevance, auction cadence	Capacity to implement and exit
Politics and catalysts	5%	Election risk, reform momentum, event risk	Timing discipline and hedge needs

Scoring methodology. Each factor is scored on a 1–5 scale, where 5 is most attractive for a long local-bond position. The weighted score is then converted to a 0–100 scale.

Country	Real yield	Disinflation/ reaction function	Fiscal anchor	FX/ external	Curveshape	Investability	Politics/ catalysts	Illustrative Score /100
Brazil	5	4	2	4	4	4	3	78
Mexico	3	4	3	4	4	5	3	74
South Africa	3	3	3	2	3	4	3	69
Colombia	4	2	1	3	3	3	3	66

This ranking broadly matches the public macro picture. Brazil still screens best because very high real rates compensate for fiscal uncertainty. Mexico scores well because policy credibility and investability are strong in the medium term, even if carry is no longer as exceptional as in Brazil. South Africa sits in the middle: term premium is attractive, but FX beta is high. Colombia remains investable and potentially high-upside, but the score is dragged down by fiscal credibility and central-bank/government tension.

Market snapshot. The table below mixes official policy and inflation data with rounded, indicative public-market yields to provide a PM dashboard rather than a pricing screen.

Country	Policy rate	2Y yield	10Y yield	Latest inflation	Real policy	Real 10Y	2s10s	PM interpretation
Brazil	14.25%	13.9%	14.3%	4.8%	9.5%	9.5%	+40bp	Highest carry; front/mid curve benefits from easing, long end still carries fiscal premium
Mexico	6.50%	8.0%	9.2%	3.55%	2.95%	5.65%	+120bp	Long end demands fiscal and term-premium compensation
South Africa	7.00%	7.3%	7.8%	4.0%	3.0%	3.8%	+50bp	Attractive term premium but high FX sensitivity
Colombia	11.25%	11.8%	12.3%	5.3%	6.0%	7.0%	+50bp	High carry and political upside, but fiscal beta remains highest

Interpreting EM yield curves. Short-dated yields mostly reflect the expected path of policy rates over the next two years. Long-dated yields reflect expected future short rates plus expected inflation and a term premium for duration, fiscal uncertainty, supply, liquidity and growth. In EM, that final term-premium component is usually larger and more unstable than in developed markets, which is why a steep curve cannot be read purely as “growth optimism” and a flat curve cannot be read purely as “recession risk.”

Policy easing compress the front end, but ambiguous guidance and fiscal concern push longer yields up.

A useful PM shorthand is therefore: the 2Y is the central-bank view, the 10Y is the sovereign-risk view, and the 2s10s slope is the argument between them. On that basis, Mexico’s steeper curve is less about imminent overheating than about long-end compensation for term premium and fiscal questions; South Africa’s curve embeds both a high real-rate environment and structural risk premium; Colombia’s flatter curve relative to its policy rate reflects very restrictive front-end pricing but also market hope that a post-election fiscal reset could help the long end.

Central-bank reaction-function matrix

Country	Inflation credibility	Current stance	PM implication
Brazil	Medium	Cautious easing	Prefer 2–5Y or belly duration; long end needs fiscal risk control
Mexico	Medium-high	On hold after easing cycle	Belly / intermediate duration attractive; avoid treating front end as a one-way cut trade
South Africa	Medium-high	Mildly hawkish after surprise hike	Selective duration; hedge part of FX risk
Colombia	Uncertain	Restrictive, independence under scrutiny	Keep exposure tactical and catalyst-driven (smaller size)

The matrix matters because the policy rate alone misses what a PM really needs to know: how likely the central bank is to cut, pause, re-hike or lose credibility when growth slows or politics intrudes. Brazil and Mexico still benefit from credible inflation-targeting frameworks, even if fiscal pressure complicates the terminal rate. South Africa has shown willingness to tighten pre-emptively when shocks hit inflation. Colombia’s policy signal is strongest on paper (rates are high), but the political uncertainty makes the reaction function harder at the moment.

4. Fiscal, investability and country views

Fiscal policy watchlist.

Country	Current fiscal issue	What to monitor
Brazil	Election-year stimulus, rising debt burden, questions over medium-term sustainability of the fiscal framework	Long-end term premium can widen even if policy rates fall
Mexico	Pemex contingent liabilities, slower growth, need to preserve sovereign balance-sheet credibility	Long-end yields remain sensitive despite relatively strong macro buffers
South Africa	Debt stabilisation plan, move toward a principles-led fiscal anchor, reform delivery at Eskom and Transnet	Scope for term-premium compression if credibility is maintained
Colombia	Post-fiscal-rule slippage, ratings pressure, need for a credible reset under the incoming government	Highest sensitivity of local yields and FX to fiscal headlines

Brazil’s gross debt rose to 81% of GDP in May on the national measure, while the broader IMF-style measure rose above 94% of GDP, underlining why long-end rates remain sensitive to fiscal policy.

South Africa’s Treasury is trying to anchor debt by promising that 2026 is the peak year and by developing a principles-led fiscal anchor, a message markets initially received positively.

Mexico retains stronger sovereign credibility than many peers and recently secured a smaller IMF precautionary flexible credit line, but Pemex remains the key contingent liability, as its financial weakness may require further sovereign support, keeping a fiscal-risk premium also embedded in the long end of the Mexican curve.

Colombia remains the most fiscal-beta-sensitive case: the legacy of fiscal-rule slippage and ratings pressure means the local curve and the peso still trade on credibility as much as on inflation.

Investability and market technicals. All four countries are investable for global local-debt managers, but the implementation profile is not identical.

Brazil and Mexico offer the deepest Latin American local rates complexes, with regular sovereign auction calendars and broad domestic institutional demand.

South Africa is highly accessible and liquid for global investors, but its currency is materially more sensitive to global risk sentiment than its domestic bond market alone would suggest.

Colombia is investable and benchmark-relevant, but trading depth is thinner and foreign participation tends to be more cyclical.

Across all four, the right way to think about investability is not “can I buy the bond?” but “can I put on size, finance it, hedge it, and exit it around a stress event?” Reuters’ reporting on renewed inflows into EM local debt as the dollar weakened is consistent with that distinction: large, benchmarked markets such as Brazil, Mexico and South Africa are the first beneficiaries when the foreign flow cycle turns.

5. Country-by-country PM analysis

Brazil

Macro. Brazil still offers the most attractive blend of carry and easing optionality in the group. Growth has stayed firmer than expected despite very restrictive rates, helped by agriculture and fiscal support.Inflation. Mid-June inflation rose 4.8% year on year, above target but below consensus for the latest print. The central bank’s near-term inflation outlook has worsened, especially because of food and energy shocks.Central bank. COPOM has cut three times to 14.25% but emphasises data dependence and has had to clarify that it has not loosened its policy horizon.Fiscal. Fiscal risk is the key offset: rising gross debt and election-year stimulus threaten to keep the long end heavy.Curve and valuation. The curve recently steepened because short rates fell on easing expectations while long rates rose on risk premium - that is precisely why the belly screens better than the long end.FX. The BRL is supported by high real carry and commodity linkage, but fiscal policy needs to be monitored not to overwhelm that support.Risks. El Niño-related food inflation, oil shocks and fiscal loosening ahead of the election.Investment conclusion. Brazil remains the top local-rate overweight, but the cleanest expression is not “buy everything”: it is to own carry-rich duration where easing still matters and long-end fiscal premium hurts less.Suggested trade expression. Illustratively: overweight 2–5Y Brazil local duration and consider FX hedging if the portfolio already carries significant commodity-currency beta.

Mexico

Macro. Mexico is not the most obvious carry market anymore, but it still offers a strong institutional framework, relatively anchoring monetary policy and a structural link to US manufacturing and “nearshoring”.Inflation. First-half June inflation slowed to 3.55%, with core at 4.12%, reducing near-term pressure on Banxico to reverse course.Central bank. Banxico has ended its easing cycle and paused at 6.50%, signalling an extended hold rather than renewed cuts.Fiscal. Mexico’s sovereign credibility remains stronger than the market sometimes concedes, reinforced by the IMF’s smaller replacement flexible credit line; nevertheless, Pemex’s losses and debt burden remain a live macro-fiscal issue.Curve and valuation. Mexico’s steep curve offers roll-down and duration opportunity, while the long end still demands a risk premium.FX. The MXN continues to benefit from policy credibility and external integration, though it remains sensitive to US growth and any renewed stress around trade or energy policy.Risks. Weaker US activity, tariffs, a re-widening of core inflation, or greater-than-expected sovereign support for Pemex.Investment conclusion. Mexico belongs in the core portfolio, but as a quality carry-and-duration market rather than a pure “high real-yield” trade.Suggested trade expression. Illustratively: selective overweight in 5–7Y local duration, with room to leave some MXN exposure unhedged if the portfolio needs a lower-volatility LatAm currency compared with BRL or COP.

South Africa

Macro. South Africa remains a slow-growth market, but reform progress in energy and logistics has helped the domestic macro narrative modestly improve, while the bond market still embeds meaningful term premium.Inflation. SARB tightened after inflation accelerated sharply in April to the top of its target band; the bank also raised its inflation forecasts because of energy-price risk.Central bank. The latest move to 7.00% signalled that SARB is willing to act early when upside inflation risks threaten to de-anchor expectations.Fiscal. Treasury’s commitment to debt peaking and the prospective principles-led fiscal anchor are constructive, although sustained credibility still depends on delivery and on quasi-sovereign reform.Curve and valuation. Benchmark long yields remain below Colombia but still offer attractive real compensation if inflation stays near target.FX. ZAR is the main challenge: it is one of the highest-beta currencies to global risk sentiment and to US-dollar moves.Risks. Global risk-off shocks, electricity and logistics setbacks, or political noise that disrupts reform credibility.Investment conclusion. South Africa is a selective rather than core overweight: attractive local duration exists, but it should be paired with tighter FX risk control than in Brazil or Mexico.Suggested trade expression. Illustratively: moderate 5–10Y local duration with some or all currency exposure hedged, depending on the portfolio’s existing USD beta.

Colombia

Macro. Colombia remains the highest-beta market in the group, where local rates and currency can move sharply on politics because fiscal risk is still the central macro variable.Inflation. Inflation was 5.29% in February, still well above target when the central bank raised rates in March.Central bank. BanRep raised rates to 11.25% and defended its independence after direct pushback from the government.Fiscal. The market had already begun to price a post-Petro fiscal reset after the first round of the presidential election. Fiscal policy implementation risk remains material until the new government takes office and sets out a credible framework.Curve and valuation. Colombia’s local curve still offers very high carry, and the 10Y bond rallied sharply on hopes of fiscal tightening.FX. COP has obvious upside if fiscal credibility improves and oil holds up, but it remains more vulnerable than MXN or BRL to confidence shocks.Risks. Delay or dilution of fiscal adjustment, renewed institutional conflict, or a negative oil shock.Investment conclusion. Colombia offers the largest upside if the incoming administration delivers fiscal normalisation, but it should be run as a catalyst-driven tactical position, not as a passive carry bucket.Suggested trade expression. Illustratively: small tactical local-duration position with a higher hedge ratio on COP until the new government’s fiscal programme is clearer.

Factor exposure map.

Country	Carry	Duration upside	FX volatility	Commodity sensitivity	Fiscal risk	US-cycle sensitivity
Brazil	High	High	Medium	High	High	Medium
Mexico	Medium	Medium	Medium	Low-Medium	Medium	High
South Africa	Medium	Medium	High	Medium-High	Medium	Medium
Colombia	High	Medium-High	High	High	Very high	Medium

This map matters because optimisation is about combining exposures, not just ranking them.

Brazil and Colombia both load on LatAm and commodity beta, but Brazil offers stronger policy credibility between the two. Mexico diversifies part of that trade through the US industrial cycle and lower fiscal stress. South Africa adds a different commodity and regional exposure set, but with high global-risk FX beta.

6. Portfolio construction and risk

Portfolio construction example. The table below is expressed as risk allocation, not capital allocation. That distinction is important: in local rates, the right question is not “how much money do I put in each country?” but “how much DV01, FX VaR and liquidity budget does each theme deserve?”

Country	Illustrative risk allocation	Rationale
Brazil	35%	Best carry plus easing optionality - strongest starter position
Mexico	30%	High-quality core holding with cleaner policy credibility
South Africa	25%	Attractive term premium, but FX hedge required more often
Colombia	10%	Highest upside-to-risk dispersion and tactical, but not anchor exposure

A portfolio built this way avoids the common EM local mistake of confusing yield with risk-adjusted attractiveness. Brazil gets the largest risk budget because real carry and local policy dynamics are favourable, not merely because yields are high. Mexico earns a large but slightly smaller share because its return profile is steadier. South Africa and Colombia remain meaningful, but both consume risk faster - South Africa through FX beta, Colombia through fiscal and political dispersion.

Correlation also matters: Brazil and Colombia should not both be run at full size if the portfolio already has large commodity and LatAm exposure. 

Risk budgeting and DV01. DV01 is the change in portfolio value for a one-basis-point move in yield. It is the simplest and most important bridge between macro conviction and position size. A PM who observes “I prefer Brazil over Mexico” still has to answer how much rate risk that view should consume.

A small illustrative example makes the point. Assume a Brazil five-year position has a modified duration of 4.2 years and a Mexico 10-year position has a modified duration of 7.5 years. On a £100 million equivalent notional, the Brazil position has a DV01 of roughly £42,000 per basis point and the Mexico position roughly £75,000 per basis point.

If the PM wants each position to consume about £30,000 of interest-rate risk, that implies roughly £71 million equivalent notional in Brazil 5Y but only £40 million in Mexico 10Y.

This is why EM portfolio construction should be done in risk units first, and notional units second.

Scenario analysis. The table below shows illustrative relative local-bond total-return impacts, not GDP effects.

Scenario	Brazil	Mexico	South Africa	Colombia	PM read-through
Faster Fed cuts / softer USD	++	+	+	++	EM FX and local duration both improve; Colombia benefits most on beta
Oil shock higher	+	-	-	++	Brazil and Colombia benefit through commodities; Mexico and South Africa face inflation / FX drag
US recession	+/-	--	-	-	Mexico is most exposed through growth and FX; Brazil is partly cushioned by domestic carry
China stimulus	++	0	+	0/+	Best for Brazil trade activities; helpful for South Africa; less direct for Mexico
Domestic fiscal slippage	-	-	-	--	Long-end and FX hit first; Colombia is most exposed

The usefulness of scenario analysis is not prediction but preparation. It tells the PM which return drivers are doing the work, whether a portfolio is concentrated in a single global factor such as the USD or oil and to revisit the weights when probabilities in markets’ dynamics are evolving. In the current setup, the dominant cross-country common factor is not inflation alone but the interaction between the Fed, the dollar and domestic fiscal credibility.

Performance attribution framework. Once positions are live, realised returns should be decomposed the same way they were underwritten: carry, roll-down, duration, FX and curve or term-premium effects.

A clean attribution answer for a Brazil position, for example, might say:

·       carry delivered most of the return

·       duration added because intermediate yields fell

·       FX was neutral because the BRL was hedged

·       curve effects detracted because the long end cheapened.

This matters in real PM work because it confirms whether the thesis worked for the intended reason.

7. Catalysts and implementation

Catalyst calendar and monitoring checklist.

Catalyst	Frequency / horizon	Why it matters	What to monitor
CPI releases	Monthly	Validates or challenges easing / holding bias	Headline, core, food-energy split
Central-bank meetings and minutes	Rolling	Reprices front end and belly	Vote split, guidance, inflation horizon
Fiscal updates and budget execution	Monthly / quarterly	Drives term premium and FX confidence	Primary balance, debt path, revenue mix
Debt auctions and syndications	Weekly / monthly	Supply changes curve shape and liquidity	Bid-cover, tails, foreign participation
Rating-agency reviews	Event-specific	Can move foreign demand and hedging costs	Outlook language and debt trajectory
Political transitions and elections	Event-specific	Changes fiscal and regulatory expectations	Cabinet signals, reform priorities, spending promises

A practical PM checklist should ask similar questions every cycle: Did inflation surprise? Did the reaction function change? Did fiscal news compress or widen term premium? Did the latest auction confirm local demand? Has the FX hedge ratio become too low or too high relative to realised volatility? Those questions matter more than any static “house view” because EM local debt is ultimately a sequencing asset class: the same country can be attractive or unattractive depending on where in the policy–fiscal–FX cycle the entry point occurs.

Implementation and automation. A robust workflow would include: a daily curve monitor for key maturities; real-yield screens; CPI surprise trackers; fiscal-headline parsing; DV01 and VaR by country and factor; auction-calendar monitoring; FX hedge-ratio monitoring; and attribution reports that reconcile PM views with realised P&L. LLM tools are useful here not for “telling the PM what to buy,” but for summarising central-bank minutes, budget documents and debt-management office statements into structured fields that can be audited against the underlying source.

Suggested production data stack. Official policy and macro releases should come from the central banks and statistical agencies - BCB and IBGE for Brazil, Banxico and INEGI for Mexico, SARB and Stats SA for South Africa, BanRep and DANE for Colombia - supplemented by finance ministries, debt-management offices, IMF, World Bank and BIS publications. Tradable curves, FX forwards and hedging costs should come from executable sources such as Bloomberg, Refinitiv or Tradeweb. Public web data can support a research draft, but it should never be the production pricing source for position sizing or performance attribution.

8. Conclusion & PM takeaways

The preferred positioning is a selective overweight to Brazil and Mexico, a measured and often partially hedged allocation to South Africa, and a smaller, catalyst-driven Colombia exposure that can be scaled if fiscal normalisation becomes credible under the incoming administration. 

The case for EM local debt today is not that all local markets are cheap. It is that several large, investable EMs still offer positive real carry and differentiated curve opportunities at a time when developed-market duration remains more tightly priced and lower yielding.

The most important PM takeaway is therefore methodological. In EM local debt, investability comes before valuation; valuation comes before sizing; and sizing comes before headline conviction.

That is why the right process starts with return decomposition, filters markets through reaction functions and fiscal anchors, converts views into factor-aware risk allocations, and closes the loop with attribution and automation. This is not just a macro note - it is a portfolio-construction note.

PM takeaways

·       High real yields remain the foundation of the opportunity set, but realised returns will be driven by country selection, curve choice and FX management rather than passive EM beta.

·       Brazil is the cleanest carry-plus-easing trade, but fiscal risk argues for a belly bias rather than unconditional long-end duration.

·       Mexico is a core quality holding in EM local debt: less spectacular than Brazil on carry, but structurally strong on policy credibility and implementability.

·       South Africa and Colombia both offer upside, but they should consume risk according to their FX and fiscal beta, not according to their nominal yields.

·       A credible PM process must connect macro views to DV01, correlation, attribution and automation; otherwise the analysis remains interesting but not professionally investable.

AIncrementum - Andrea Bonini - June 2026