What yield differences are (and what they are not)
Yield differences usually refer to the interest-rate gap between two currencies, used to reason about the expected “carry” from holding positions exposed to those currencies. In plain terms, it’s a comparison of which currency has the higher reference interest rate and which has the lower one.
Two common misunderstandings cause most issues:
- Yield differences are not the same as a guaranteed return.
- They do not automatically predict the future direction of the exchange rate.
Even if a higher-yield currency provides positive carry in theory, that advantage can be offset by currency moves or by frictions that change what you actually earn.
How the idea works in principle
A basic yield-difference comparison is often built from a few components, such as:
- A reference interest-rate measure for each currency
- The sign and magnitude of the gap (higher minus lower)
- An expectation of how exchange rates might change over the holding period
To treat it as an expectation rather than a certainty, you must state assumptions explicitly. For example, any example you use should specify:
- Which interest-rate benchmark you mean (and why that benchmark is appropriate)
- The holding period
- Whether you assume the exchange rate stays constant, changes predictably, or changes randomly
A key point is separation of mechanics from outcomes. The “gap” is a starting input; the realized outcome also depends on exchange-rate changes, timing, and costs.
Evidence and examples of where the relationship breaks
A frequent failure mode is the “carry pays—until it doesn’t” pattern: yield differences may describe the interest differential component, but not the total return. If the lower-yield currency appreciates (or the higher-yield currency depreciates) enough, it can erase carry.
Another failure mode is model mismatch. If your calculation uses one set of assumptions (for instance, a simplified holding period or an idealized carry), real outcomes can diverge due to:
- Financing and rollover mechanics that differ from the simplified estimate
- Transaction costs and bid/ask spreads
- Timing differences between when a reference rate is set and when cash flows occur
Finally, relationships observed in history can be regime-dependent. What worked during one environment (for example, stable volatility or steady liquidity) may behave differently during stress, when funding conditions or risk pricing change.
Limitations and risks to understand
1) Uncertainty about inputs
Reference interest rates and the conventions used for carry can be ambiguous. Different market participants can treat the “yield” input differently, and those choices can change the computed gap.
2) Costs and execution can dominate
Even without changing the interest-rate gap, realized results can change due to transaction costs, spread, and execution timing. These frictions can be small in quiet conditions and materially larger in stressed or illiquid moments.
3) Exchange-rate moves are not controlled
Yield differences do not control currency appreciation or depreciation. Since total return includes exchange-rate effects, large or sudden moves can overwhelm the carry component.
4) No guarantee from past behavior
Historical relationships between relative yields and outcomes can weaken. Markets can reprice expectations, adjust risk premia, or shift liquidity in ways that break prior patterns.
How to verify the concept without over-interpreting it
To independently verify claims about yield differences, check three things using your own inputs:
- Benchmark alignment: Confirm what interest-rate measure is being used and whether it matches your context.
- Cash-flow assumptions: Ensure your example includes realistic holding period conventions and the direction of carry.
- Sensitivity: Test how outcomes change when exchange rates move modestly in either direction and when costs are included.
If your setup cannot state assumptions clearly, the “yield difference” comparison is likely being treated as more certain than it is.
You can also ask a precise follow-up: “Under what conditions do funding conditions, volatility, or liquidity change the relationship between relative yields and realized results?”