Carry Trades and Funding Currencies: Why Interest Income Is Not the Total Return

A carry trade links an asset to a funding obligation

A currency carry trade is commonly described as borrowing in a lower-interest-rate currency and holding an asset in a higher-interest-rate currency without fully hedging the exchange-rate exposure. That description identifies an income differential, not a guaranteed profit. The investor owes money in one currency while the asset is valued and pays in another. The relevant outcome therefore depends on financing costs, investment performance and the conversion rate when obligations are settled. A funding currency is a role within a transaction, not a permanent label attached to a country. The same currency can fund one participant and be the investment currency of another. Some exposures are created with derivatives rather than a visible cash loan. An explanation should identify the economic exposures first and the particular legal instrument second, because similar risks can appear through different balance-sheet forms.

Carry is only one component of the result

Interest accrues through time, whereas currency prices can change continuously and asset prices can move independently. A high local yield may compensate for inflation, credit uncertainty, restricted liquidity or other risks. It should not be read as a free margin available to an international investor. In a simple cash position, the final funding-currency proceeds depend on both the foreign investment return and the exchange-rate change. When the purchased asset is a bond rather than a deposit held to maturity, its market price introduces another component. A bond can pay every coupon and still lose value over a particular holding period. Likewise, a currency can rise while the investment loses money for unrelated reasons. Separating these components avoids the common mistake of explaining every outcome through the interest-rate differential alone. The return belongs to the whole position, including its financing.

An illustrative currency move can erase a positive interest spread

Consider a fully hypothetical one-year transaction with no fees, taxes, collateral costs or defaults. An institution borrows 100 units of funding currency at 2%, converts them at an initial rate of one funding-currency unit per one investment-currency unit, and earns 5% on the investment. It finishes with 105 investment-currency units and owes 102 funding-currency units. If one investment-currency unit then buys only 0.92 funding-currency units, the investment currency has depreciated by 8% in this explicitly defined quotation. Conversion produces 96.60 funding-currency units, leaving a shortfall of 5.40. That loss is 5.4% of the initial borrowed principal, not a return on the investor's equity. The exact calculation multiplies the investment growth factor by the currency conversion factor before subtracting the repayment. Simply subtracting percentage changes gives only an approximation and can conceal cross-effects.

Low rates do not make a funding currency harmless

A currency may become attractive as a source of funding when borrowing costs are relatively low and market access is deep. But its future value is not fixed by those conditions. If it appreciates, a borrower must obtain a more valuable currency to repay the debt. The burden can increase even when the contractual loan rate does not change. Historical discussion of yen-funded positions illustrates this distinction between a low interest bill and potentially large currency exposure. Bank of Japan financial stability reports examine market developments and investor positioning, but such evidence does not imply that every yen liability is a speculative carry trade. A corporation may borrow yen against yen revenues, or an investor may hedge the currency exposure. Matching cash flows and hedging arrangements determine whether the familiar carry narrative actually describes the position under examination.

Hedging changes the economics of the interest differential

An unhedged carry exposure and a fully currency-hedged foreign investment are different propositions. A forward contract can fix a future conversion rate, but its price generally reflects interest differentials and market funding conditions. The investor cannot simply retain the full foreign interest advantage and assume that exchange-rate insurance is costless. Covered interest parity explains the benchmark relationship; cross-currency basis, transaction costs and contract terms complicate the implemented outcome. Moreover, a short-dated hedge renewed against a longer-lived asset introduces renewal and liquidity questions. This does not mean hedging is ineffective. It means the analysis must describe which risk has been reduced and which obligations remain. A comparison of two portfolios should specify their hedge ratio, horizon, financing instrument and collateral treatment. Without those details, identical references to carry can conceal fundamentally different exposures and cash-flow demands.

Leverage turns an asset loss into a capital and liquidity problem

Leverage scales the relationship between exposure and the capital supporting it. A loss measured against total assets is not the same percentage loss measured against a smaller equity base. Margin requirements add a timing dimension: a participant may need cash or eligible collateral before the position reaches its intended maturity. Even a belief that prices will later recover cannot satisfy an immediate contractual margin call. The IMF's financial stability work discusses how leverage and maturity mismatches can amplify market disturbances. For an original hypothetical balance-sheet exercise, imagine assets of 100 financed with debt of 90 and equity of 10. A five-unit asset loss, with debt unchanged and ignoring other effects, removes half of equity. This arithmetic does not describe every carry trade, but it shows why the size of the interest spread alone says little about the resilience of the institution holding it.

An unwind can reinforce the original price move

When a leveraged participant closes an unhedged foreign investment, it may sell the investment asset, convert the proceeds, and buy the funding currency needed for repayment. If many participants do this together, their transactions can push the investment asset down and the funding currency up. Those moves worsen the position of remaining participants and can generate further sales. BIS analysis of the August 2024 turbulence describes how deleveraging and margin increases amplified an initial reaction to news. That historical episode is evidence of a mechanism, not a template guaranteeing the same sequence in every future event. Positioning, market depth and institutional constraints determine the scale of feedback. It is therefore useful to distinguish the initial catalyst from the amplification process. A modest information shock can produce a large price move when the system has accumulated fragile, similarly financed exposures.

A calm return history can conceal an uneven risk distribution

A position can earn small positive amounts repeatedly and still have an unattractive distribution of losses in adverse conditions. Average historical income does not reveal the size, timing or clustering of bad outcomes. Federal Reserve research connects currency risk premia with uncertainty and tail risks, reminding readers that compensation for risk need not appear as a smooth annual payment. Such studies estimate relationships over defined samples; they do not establish an invariant rule for tomorrow's exchange rate. A useful conceptual distinction is between the probability of a loss, the size of that loss and its coincidence with other financial pressures. A currency loss is especially consequential when it arrives alongside falling collateral values or tighter financing. Looking only at normal-period volatility can miss these joint states. Economic resilience depends on the combined scenario rather than on the historical average of one variable.

Positioning data reveal fragments rather than a complete carry map

The global size of carry positions is difficult to measure because activity spans cash borrowing, forwards, futures, swaps and investments that have several possible motives. A short position in a currency future may be speculative, a hedge for another asset, or part of a relative-value package. Nonresident borrowing in a low-rate currency is similarly informative without being a clean count of unhedged carry. The IMF and BIS explicitly discuss these measurement limitations, while the Bank of Japan uses multiple indicators to examine investor activity. Analysts should resist adding overlapping datasets as if each represented an independent population. They should also distinguish gross positions from net positions and dealer intermediation from ultimate risk taking. A defensible conclusion identifies what the evidence covers and what remains unknown. Apparent numerical precision is not a substitute for understanding the perimeter of the data.

Read carry headlines as explanations to test

A claim that a currency move reflects a carry unwind should be tested against changes in financing conditions, asset prices, positioning and the timing of the move. Interest-rate expectations matter, but the starting valuation and the speed of repricing also matter. Federal Reserve analysis of emerging-market currencies demonstrates why yield differentials and realised currency outcomes can diverge. BIS work on changing global financial conditions places these exposures within broader cross-border financial links. The analytical objective is not to identify a permanently safe funding currency or recommend a trading combination. It is to explain who owes what, in which currency, when payment is due, and what could force an early exit. All numerical examples here are hypothetical educational calculations. Carry becomes intelligible only when income, total return, leverage and liquidity are kept separate and then examined together.

Explore the data and markets