Direct answer
Carry relationship is a way to describe how interest-rate differentials between two currencies can translate into returns when you hold one currency against another. The main limitations are that (1) the exchange rate can move against the trade idea, (2) costs and execution details can reduce or change results, and (3) the “relationship” between rates and outcomes is not stable over time.
Because carry relationship is a conceptual mapping rather than a guaranteed outcome, it is most useful when you can clearly state assumptions and independently check whether those assumptions are still plausible.
Mechanism and definition
A standard carry relationship description starts with two currencies, A and B, and an interest-rate differential. If currency A’s interest rate is higher than currency B’s, then the carry mechanism suggests that, all else equal, holding exposure consistent with “A over B” can earn a benefit related to that differential.
However, “all else equal” is a critical assumption. The realized result depends on more than the differential itself. In practice, at least three components matter:
- Interest component: how the relevant rates accrue over the holding period.
- Exchange-rate component: changes in the spot exchange rate (or the equivalent effect in forwards/rolling instruments).
- Friction component: costs such as spreads, commissions, financing/roll costs, and any platform or contract-specific mechanics.
To discuss limitations responsibly, you have to separate these components: the interest part may be positive while the exchange-rate part is negative and larger.
Evidence or example (with explicit assumptions)
Consider an example with clear assumptions. Suppose you compare a scenario where the interest differential favors currency A over currency B for a short horizon, but you also assume the exchange rate moves in the opposite direction.
If currency A depreciates versus currency B during the holding period, then the exchange-rate component can offset or outweigh the interest component. Even if the interest differential remains the same on paper, the realized outcome can still be negative because carry is not purely an interest payment; it is the combined effect of accrual and currency valuation.
This is a common failure mode: a carry idea based on the differential can lose when the currency that “should” benefit from higher interest instead strengthens less than expected—or moves against the exposure.
Limitations and risks (failure modes and uncertainty)
1) Exchange-rate risk can dominate
Carry relationship implicitly treats the interest differential as the primary driver. A material limitation is that exchange-rate changes are not controlled and can be substantial. When the currency associated with the higher interest rate falls (or depreciates more than expected), the net effect can be unfavorable.
2) The differential itself can change
Interest rates are not fixed. Even without assuming dramatic events, expected differentials can shift due to new economic information, policy expectations, or revisions of rate paths. That makes carry relationship time-sensitive in a practical sense.
3) Costs and execution change the realized economics
Conceptual carry often omits detailed frictions. Real outcomes can be materially affected by:
- transaction costs (spreads and commissions),
- execution quality and slippage,
- financing or roll mechanics for the specific instrument used,
- and constraints like contract size or margin rules.
Even with the “right” rate differential, these friction components can reduce the net benefit.
4) Historical relationships may not generalize
Another limitation is methodological: showing that past interest differentials correlated with outcomes does not guarantee the same behavior in the future. Market regimes can change, and relationships that held in one period can weaken or reverse later.
Verification and next questions
To independently verify what carry relationship implies in your context, focus on these checks:
- Assumptions: What interest differential and horizon are you using, and how are you treating the exchange-rate component?
- Instrument mechanics: Are you modeling a spot exposure, a forward-like exposure, or a rolling product with specific roll/financing behavior?
- Costs: What frictions apply, and are you including them in the net result?
- Regime sensitivity: Are you relying on a historical pattern that may not persist?
A useful next question is: under what specific conditions does the exchange-rate component tend to overwhelm carry, and how would your assumptions break in those scenarios?