Skip to main content

Rate differential · position economics · unwind risk

6M Carry Trade: Economics, Flows, and Unwind Risk

Suppose Mexican short-rate expectations still exceed comparable U.S. expectations, yet the peso sells off and 6M falls. That is not a contradiction. Carry is one component of total return; exchange-rate loss, funding stress, hedging cost, crowding, or a change in the expected policy path can overwhelm it.

Input
Comparable expected rates
Payoff
Total return, not yield alone
Stress lens
Volatility + funding
Output
Conditional state
Positive carryMXN still weakens
carryFX moveall costs

A favorable rate gap can coexist with a negative position return. The horizon, instrument, leverage, hedge, and path of the currency decide whether the economics survive.

Yield is not a floorDirection is not guaranteed

Read the instrument first

A Peso Carry Thesis Must Be Translated Into the 6M Quote

Established fact: CME's FX Product Guide lists 6M as MXN/USD, quoted in U.S. dollars per Mexican peso. Therefore, a higher 6M price means one peso buys more dollars—a stronger MXN against USD. The commonly displayed spot convention, USD/MXN, is reciprocal, so it ordinarily moves the other way after aligning timestamps.

Spot convention

USD/MXN

Mexican pesos per U.S. dollar. A rise means the dollar strengthened against the peso.

6M convention

MXN/USD

U.S. dollars per Mexican peso. A rise means the peso strengthened against the dollar.

Do not treat futures as spot with a different label

A dated 6M contract also contains time, interest-rate expectations, and basis. Use the canonical 6M contract guide for current specifications, settlement, delivery, and roll mechanics. This page owns the carry mechanism, not contract arithmetic.

Position economics

Carry Is the Residual After Funding, Currency, Hedge, and Trading Costs

A textbook trade borrows or sells a lower-yielding funding currency and holds a higher-yielding target-currency asset. Real positions can instead use forwards, swaps, futures, options, bonds, or combinations. The labels “funding currency” and “target currency” describe the economic exposure, not a single required implementation.

A short-horizon approximation

position return ≈ target-asset return − funding cost + FX return − hedge, financing, and transaction costs

The approximation is intentionally incomplete: mark-to-market timing, convexity, collateral, credit, taxes, and basis can matter. A 6M futures position does not simply receive a posted overnight-policy-rate difference as cash carry.

Compare
Same horizon
Model
Expected path
Subtract
All implementation costs
Stress
FX path + leverage
ComponentWhat to measureWhy the shortcut fails
Rate differentialMarket-implied or investable yields at matched horizons, with credit and liquidity differences identifiedTwo policy targets are not the realized return on two tradable assets
Currency returnQuote-correct MXN/USD change over the holding periodA modest adverse FX move can outweigh accrued yield
Forward/futures basisDated price relative to aligned spot and financing assumptionsExpected rate differences can already be embedded in the contract
HedgePremium, forward points, option skew, rollover, and residual basis riskHedging can remove much of the currency return that motivated the trade
ImplementationSpread, slippage, fees, margin, collateral return, and financing termsHeadline carry is gross, not net

Before calling a regime carry-friendly

Five Conditions Must Be Examined Together

Mechanism: a positive expected Mexican-versus-U.S. return gap can attract demand for MXN exposure, all else equal. It becomes a usable hypothesis only after the return gap, volatility, funding, crowding, and macro path are examined.

Investable differential

Use comparable instruments and horizons. Distinguish the current target rate from the path priced over the intended holding period.

Contained currency variance

Carry accrues gradually while exchange rates reprice continuously. Rising realized or implied volatility changes the loss distribution.

Stable funding

Repo, swap, collateral, and margin terms can tighten even if official policy rates do not move.

Room in positioning

Crowding can make new demand less important than the risk of a common exit. Public positioning data are partial and delayed.

Credible policy path

The relevant differential is forward-looking. Inflation, growth, or central-bank communication can change expected paths before a target-rate action.

Executable market

Spread, depth, gaps, and event exposure can turn a plausible macro state into a rejected trade.

Established source boundary: Banco de México publishes its policy statements and minutes; the Federal Reserve publishes FOMC statements, minutes, and projections by meeting. Those documents establish official communication. Any inferred path or carry attractiveness remains an analyst inference.

From relative return to price

Trace the Flow Channel Without Assuming It Dominates

The same rate-gap change can reach 6M through several paths. Some reinforce one another; others offset. Label the hypothesized channel before looking at the price outcome.

1

Expected return changes

New policy or macro evidence alters the expected Mexico–U.S. return difference at a relevant horizon.

2

Portfolio choice changes

Investors may add, hedge, reduce, or leave MXN exposure unchanged depending on risk limits and existing positions.

3

FX demand changes

Conversion, derivative hedges, or dealer inventory can create peso buying or selling, but the size and timing are not directly observable from a policy rate.

4

6M reprices

Futures respond to the currency pair, dated basis, liquidity, and expectations. Confirmation must be observed, not presumed.

Potential reinforcement

Rate gap widens with calm funding

If Mexican expected returns rise relative to U.S. returns, volatility stays contained, and MXN strengthens across aligned spot and futures observations, the evidence is consistent with a supportive carry channel. It still does not prove carry caused the move.

Potential offset

Higher yield reflects higher risk

A Mexican yield rise can reflect inflation, fiscal risk, liquidity stress, or a higher required risk premium. In that state, the larger nominal gap can coexist with a weaker peso and lower 6M.

Observable, but incomplete

Build Confirmation From Independent Evidence

No public series identifies “the MXN carry trade” in real time. Confirmation is a triangulation exercise. The strongest record aligns independent rate, currency, volatility, positioning, and market-quality observations without turning any proxy into ground truth.

EvidenceConsistent with carry supportWhat would weaken that reading
Relative curveMatched-horizon Mexican pricing firms versus U.S. pricingMove is isolated, credit-driven, or reverses quickly
Currency complex6M rises while reciprocal USD/MXN falls after timestamp alignmentOnly one stale or rolled series moves
Volatility and fundingImplied/realized volatility and funding conditions remain compatible with the risk budgetVolatility, margin, or basis stress rises faster than expected carry
Positioning contextPublic reports do not show an extreme by the analyst's predeclared measureProxy indicates crowding or categories shift without economic clarity
Price acceptance6M holds the predeclared confirmation area in executable conditionsPrice rejects the area or market quality fails

Established limitation: the CFTC Commitments of Traders report breaks down reportable open interest as of Tuesday and is generally published Friday. Its categories are regulatory classifications, not a live inventory of all carry trades. Use it as delayed context, never as same-session order flow.

Where gradual economics can become nonlinear

An Unwind Is a Feedback Loop, Not a VIX Number

Established institutional evidence: a BIS review of the August 2024 turbulence described leveraged trade unwinds, procyclical deleveraging, and margin increases as amplification channels while also noting that overall carry size is difficult to measure. That episode demonstrates a mechanism; it does not set a universal threshold for MXN or 6M.

A

Catalyst changes the prior

A policy, growth, geopolitical, or market shock changes expected returns or risk tolerance.

B

Volatility and losses rise

Adverse FX movement consumes accumulated carry and increases portfolio variance.

C

Constraints bind

Stop rules, margin calls, financing terms, or risk limits force some positions smaller.

D

Common exits amplify

Further MXN selling can pressure 6M, creating more losses. The loop can also stop or reverse if liquidity and buyers absorb it.

Failure cases for the unwind story

A falling 6M price is not automatically a carry unwind. It may reflect a broad-dollar shock, Mexico-specific fiscal or trade news, contract roll, poor data alignment, or ordinary repricing by unleveraged investors. Conversely, a volatility spike need not weaken MXN if the shock is concentrated elsewhere or peso demand has a different source.

Carry-state close

Use a Carry-State Matrix That Permits “Unclear”

This matrix organizes evidence; it does not forecast 6M. Record the state at a fixed timestamp, specify the horizon, and require confirmation before turning a macro hypothesis into a possible application.

Supportive, uncrowded

Required: favorable investable differential, stable funding and volatility, independent MXN confirmation, and no declared crowding breach.

Hypothesis: carry demand may support higher 6M.

Fail: relative curve, price confirmation, or execution quality reverses.

Supportive, crowded

Required: favorable differential but a positioning or volatility proxy reaches a predeclared extreme.

Hypothesis: support may persist while unwind asymmetry grows.

Fail: risk cannot be bounded; crowding alone does not time an exit.

Differential compressing

Required: expected Mexican return advantage narrows relative to the U.S.

Hypothesis: one source of MXN demand may weaken.

Fail: other flows dominate or 6M refuses to confirm.

Funding stress

Required: adverse FX, rising volatility, tighter financing, and cross-market deleveraging align.

Hypothesis: forced reduction may amplify lower 6M.

Fail: liquidity, curve, and price recover through the invalidation area.

Unclear

Required: rates, price, volatility, positioning, or data quality conflict.

Conclusion: no directional inference.

Action: remain flat until the evidence set becomes testable.

Research status: this page explains mechanisms and a monitoring framework. It reports no original estimate of carry returns, crowding, predictive power, or profitability. Testing a carry rule requires point-in-time rate curves, quote-correct dated futures, rolls, funding and transaction costs, declared decision times, and untouched validation data.

Sources, methods and editorial disclosure — reviewed August 13, 2026

Sources and methods were reviewed August 13, 2026. Dynamic official series must be timestamped when used. Facts, mechanisms, hypotheses, inferences, and possible applications are separated in the text; no proprietary result is reported.