Carry trade meaning: interest rate vs exchange risk metrics

A 350-basis-point rate gap — the kind observed historically between US and Japanese short-term rates — can be economically neutralised once implied volatility crosses critical thresholds, or once the target currency begins to depreciate by an amount approaching the carry itself. The carry trade meaning extends well beyond the textbook motif of borrowing cheap to lend expensive. It denotes an unhedged cross-currency position whose realised return depends on the rate spread, the volatility regime, the liquidity backdrop and the leverage applied at the moment of execution.

This breakdown focuses on the comparison the headline obscures: the relative weight of the interest-rate input versus the exchange-rate input in the eventual P&L. Both inputs are observable, both are quantifiable, and neither is sufficient on its own.

Funding and target currency selection: policy mandates as the starting point

A currency enters the funding column when its central bank maintains a policy stance that keeps short-term rates below the rates prevailing in the currency it is paired against. The European Central Bank's glossary frames this as borrowing in a "funding currency" and investing in a higher-yielding "target currency," without hedging the exchange-rate exposure. The framing is unambiguous on one point: the position retains currency risk by construction. The trade is "positive carry" when the funding rate is below the target rate; reversing the direction produces a "negative carry" trade that bleeds basis points daily and survives only if the funding currency depreciates by enough to offset the spread.

Selection is not static. A currency that functioned as a funding leg in one cycle can rotate into a target leg when its domestic rate path diverges from the perceived global floor. The yen held the funding role almost continuously from the late 1990s through 2022 because the Bank of Japan maintained a yield-curve control regime anchored near zero. Once the BoJ permitted yields to drift higher and formally exited negative rates in 2024, the funding designation began to fragment — the magnitude of carry available against the yen compressed, and the pool of currencies still offering sub-1% funding rates narrowed accordingly. Selection therefore tracks the relative path of monetary policy, not the absolute level of any single rate.

The table above does not describe permanent identities; it describes state-dependent characteristics. A BIS Quarterly Review from 16 September 2024 documented that the peso depreciated sharply against the yen during the August 2024 unwind — the textbook target-currency behaviour. Whether the peso retains that designation in subsequent cycles depends on the rate differential between Banxico and the rest of the high-yielding complex, on realised volatility, and on positioning.

A carry trade is not a long interest-rate position and a long currency position layered on top of each other. It is a single unhedged cross-currency exposure whose inputs are inseparable.

Quantifying the risk: why the carry-to-risk ratio reshapes the headline

Measuring carry trade attractiveness by the size of the interest-rate differential treats the spread as if it were the only relevant variable. The carry-to-risk ratio, in the formulation used by the Bank for International Settlements, divides that differential by a measure of FX volatility. The BIS specification uses the 12-month forward-implied interest-rate differential over the option-implied volatility of the same currency pair — a market-priced input rather than a survey-based or model-derived estimate.

The ratio is a risk-adjustment, not a forecast. A larger value indicates that the implied carry per unit of implied volatility is more favourable, all else equal. The "all else equal" caveat excludes the most disruptive variables: an abrupt repricing of the volatility surface, a regime change in the funding currency's policy path, or a liquidity squeeze in the target-currency market. None of these is captured by option-implied volatilities, which reflect expectations as of the pricing date.

The ECB's March 2010 Monthly Bulletin noted that the carry-to-risk ratio's interpretation depends critically on the rate tenor chosen, the volatility measure selected, and the historical window for context. There is no universal "good" or "safe" cutoff; the metric's reading shifts with each input. Drawing an execution threshold from a single observation — say, the ratio prevailing in mid-2007 — and applying it to a 2025 or 2026 volatility regime would compare a structurally different market state.

The carry-to-risk ratio does not tell a trader whether a trade will be profitable. It tells a trader whether the current market is paying carry cheaply or expensively relative to its own pricing of risk.

Uncovered interest parity and the reality of market returns

Under uncovered interest parity (UIP), a currency with a higher interest rate should be expected to depreciate against a lower-yielding currency by an amount equal to the rate differential. The expected return from holding the higher-yielding currency is therefore zero in theory: the carry is exactly offset by the anticipated depreciation. Empirically, the relationship has been unstable for prolonged periods. The ECB has noted that UIP is a theoretical benchmark whose empirical performance varies across regimes, holding in some windows and failing sharply in others.

Where UIP has failed, the failure has been directional: high-yielding currencies have, on average, tended not to depreciate by the carry, producing positive realised carry for the unhedged position. This empirical deviation from UIP — sometimes labelled the "forward premium puzzle" in academic literature — is what makes carry trades viable as a strategy at all.

The negative skewness documented by the National Bureau of Economic Research is the qualifier. Carry-trade exchange-rate returns cluster: small positive outcomes occur frequently, while a small number of episodes generate disproportionate losses when the strategy unwinds. The strategy is therefore not "high-yielding currency pays you to wait" — it is a distribution in which the right tail is truncated and the left tail is fat. The carry-to-risk ratio is silent on which tail a given market state belongs to; volatility is priced symmetrically around the at-the-money strike, while realised FX returns are not symmetrically distributed.

The anatomy of an unwind: volatility, liquidity, and rate-path repricing

A carry trade does not unwind because the rate differential closes. It unwinds because the conditions that supported the position — low implied volatility, abundant funding liquidity, and a policy path aligned with the position — change simultaneously. The BIS Quarterly Review released on 16 September 2024 documented the August 2024 episode in which carry positions were unwound amid changing rate-path expectations and a sharp rise in realised volatility. Funding currencies appreciated; target currencies depreciated. The yen was a focal funding leg; the Mexican peso was a representative target leg.

The August 2024 episode illustrates a structural feature of the strategy: the exchange-rate move is not the cause of the unwind but the consequence of the funding/liquidity shock. As leveraged carry positions are reduced, the funding-currency leg is bought back and the target-currency leg is sold. The mechanical flow of redemptions and liquidations becomes a self-reinforcing exchange-rate move, and that move can exceed the size of the original rate-spark that triggered the deleveraging. From end-August to end-October 2008, for instance, the yen appreciated 28% against the euro and 11% against the dollar over a two-month window — a magnitude that has not been reproduced as the dominant single-currency move in subsequent episodes.

The implication for risk management is asymmetric: an unwind need not be telegraphed by a narrowing rate spread. It can originate in funding stress, in a volatility-regime shift unaccompanied by rate moves, or in a single policy surprise that re-prices a long-standing forward curve. Each of these triggers is observable in real time; none is captured by the interest-rate-differential input alone. A carry trade that looks attractive on a carry-to-risk reading can be the same carry trade that triggers the largest drawdown for the year if the underlying assumptions about policy-path stability break.

Leverage as a catalyst: why positioning scales systemic sensitivity

The carry trade is, by definition, a leveraged cross-currency position. BIS research describes the position as a leveraged structure explicitly designed to exploit interest-rate differentials and low volatility. Leverage increases sensitivity to changes in exchange rates, interest rates and volatility — but not symmetrically across inputs. Adding leverage to an existing position magnifies the sensitivity to a given basis-point move linearly; the losses generated by an adverse move are amplified by the same factor, while the carry earned in stable conditions scales sub-linearly with leverage because funding costs rise with balance-sheet demand.

A BIS Bulletin published on 6 May 2026 found that significant short positions in funding currencies held by carry traders can amplify the exchange-rate effect of a monetary-policy tightening event. The mechanism: leveraged short positions in the funding currency are unwound around the policy announcement, accelerating the appreciation of the funding currency and the depreciation of the target currency beyond what the rate-differential move alone would imply. The policy decision is the catalyst; the positioning is the accelerant. Where this feedback loop is large, the resulting FX move can dominate the rates move in the macro narrative that follows.

This finding formalises a structural concern that has applied since carry trades re-emerged as a fund strategy in the early 2000s: the exchange-rate response to a policy event is conditional on the existing positioning of carry traders, not on the absolute change in the policy rate. Forecasts of central-bank decisions that price in only the rate-differential change may understate the realised FX impact when carry positioning is large in one direction. The same announcement can produce a 50-basis-point-equivalent rate move and a 300-pip FX move, or a 50-basis-point-equivalent rate move and an 800-pip FX move, depending on where the leverage sits when the release hits.

The exchange rate does not follow the interest rate. The exchange rate follows the marginal flow from leveraged positioning around policy events, of which the interest-rate change is one input.

Structural implications for FX positioning

Comparing interest-rate metrics against exchange-rate metrics is not a question of which variable carries more weight. The two are functionally inseparable in the realised return of an unhedged carry position. The interest-rate differential determines the daily accrual and the headline economic rationale. The exchange-rate risk determines whether that accrual survives a quarter, a year, or a multi-year cycle. The carry-to-risk ratio attempts to compress both inputs into a single comparable figure, but the compression is a convenience for screening, not a substitute for disaggregated analysis.

Three structural points remain stable across regimes. First, currency designations — funding versus target — are state-dependent and rotate with relative rate paths, balance-of-payments configurations, and the prevailing volatility regime. There is no currency that is a permanent funder or a permanent target; the role is assigned by the policy differential at each decision point. Second, the carry-to-risk ratio is a market-implied metric whose reading depends on the volatility regime prevailing at the moment of calculation. A falling ratio can indicate either a compressing rate spread or a rising implied volatility, and the two carry materially different implications for forward positioning. Third, leverage is the input that converts a market position into a systemic variable — one that policy events can re-price, as the May 2026 BIS Bulletin documented.

For positioning purposes, the analytical discipline mirrors that of an institutional rates desk: identify the policy mandate driving the rate path, map that path onto the carry-to-risk ratio using market-priced inputs, layer positioning data on top, and size the exposure to the volatility regime the curve is currently pricing. The interest-rate differential is the first filter. It is not the last word, and treating it as such is the most common entry point into the kind of drawdown the carry-to-risk ratio was designed to flag.