Currency Carry Trade Return Calculator
Estimates the expected gross return on a leveraged FX carry trade — borrowing in a low-yielding currency to invest in a high-yielding one — net of a fixed transaction-cost assumption. Useful for sizing a carry position against a target return and for comparing the carry premium across currency pairs.
Last updated: May 2026
Compare with similar
About this calculator
The formula is return = investment × leverage × ((highYield − lowYield) ÷ 100) × (days ÷ 365) − investment × leverage × 0.002. The first term is the interest-rate differential earned (carry) prorated for holding period and scaled by leverage; the second term is a flat 20bps transaction-cost estimate on the leveraged notional (covering broker spread, swap-rollover funding cost, and exchange fees, but NOT including currency moves themselves). The crucial caveat: this is the EX-ANTE expected carry, assuming the currency pair does not move. The realized return depends entirely on the spot exchange rate at unwind: if the high-yielder appreciates, you double-win (carry + FX gain); if it depreciates, the FX loss can wipe out the carry many times over. Covered Interest Parity (CIP) says the forward exchange rate is priced exactly to offset the interest differential, so the EXPECTED return on a fully covered carry trade is zero (the differential is locked in but exactly offset by the forward price). Uncovered Interest Parity (UIP) says the expected SPOT rate at unwind also offsets the differential — so even uncovered carry trades should have zero expected return in efficient markets. Empirically UIP fails: high-yield currencies have on average appreciated, not depreciated, against low-yielders over the 1990–2007 period (the "forward premium puzzle"), giving carry trades historically positive returns of ~4–6% Sharpe-adjusted. Post-2007 carry returns have been much weaker; post-2022 the rate-cycle synchronization has reduced opportunity. Carry trades are NEGATIVELY skewed: small steady gains punctuated by large losses (the carry-trade "crash risk" of 2008 saw JPY-funded carries lose 30–50% in a few weeks). Leverage amplifies both directions linearly: 5x leverage on a 4% differential gives 20% gross but a 4% adverse FX move becomes a 20% drawdown. Edge cases: lowYield > highYield gives negative carry (you pay the differential — usually a hedge, not a trade); leverage = 1 is unlevered carry; very long holding periods compound funding cost on the rollover side; intra-day trades have day-count fractions <1/365 and microscopic carry that does not justify even the 20bps assumed cost.
How to use
Example 1 — Classic JPY-funded AUD carry, moderate leverage. AUD overnight policy rate 4.35%, JPY -0.10% (2024). High = 4.35, low = -0.10, differential = 4.45%. You commit $100,000 of equity at 3x leverage and hold 180 days. Enter investmentAmount = 100000, highYieldRate = 4.35, lowYieldRate = -0.10, leverageRatio = 3, tradingDays = 180. Result: 100000 × 3 × 0.0445 × (180/365) − 100000 × 3 × 0.002 = 6583.56 − 600 = $5,983.56 expected gross carry on $300,000 of notional, $5,983.56 ÷ $100,000 equity = 5.98% over 6 months. Verify: differential earnings 13350 × (180/365) = 6583.56 ✓; cost 300000 × 0.002 = 600 ✓. The annualized expected carry is about 12% on equity (if you maintain the position and the FX rate is unchanged). But: AUDJPY can move 8–15% in a normal year and 20–40% in stressed periods (2008 saw a 40% drop in months) — so the realized return is dominated by FX direction, not carry. Example 2 — USD-funded MXN carry, higher rate gap and higher risk. MXN overnight rate ~11.25% (Banxico, 2024), USD ~5.25%. Differential = 6.00%. Commit $50,000 at 4x leverage for 90 days. Enter investmentAmount = 50000, highYieldRate = 11.25, lowYieldRate = 5.25, leverageRatio = 4, tradingDays = 90. Result: 50000 × 4 × 0.06 × (90/365) − 50000 × 4 × 0.002 = 2958.90 − 400 = $2,558.90 expected gross carry on $200k notional, $2,558.90 ÷ $50k equity = 5.12% over 90 days, ~22% annualized on equity. The MXN has historically delivered some of the highest carry premia in the EM universe but with periodic 15–30% drawdowns (2008, 2015, 2020 COVID). The 20bps cost assumption is too low for MXN — real broker spreads on USDMXN are 50–150bps for retail and 5–15bps institutional, so this calculator understates costs in EM unless you adjust the formula or interpret the 0.002 as a rough average.
Frequently asked questions
Why does the carry trade make money if Uncovered Interest Parity says it shouldn't?
UIP says the expected change in spot exactly offsets the interest differential — so a 5% rate-differential trade should see the high-yield currency depreciate by 5% on average, leaving zero expected return. Empirically UIP fails systematically: high-yield currencies tend to appreciate or stay flat against low-yielders over multi-year windows, especially in non-crisis periods. The explanation that has the most academic support is the "peso problem" plus risk premium: carry-trade returns look high in calm periods because the rare crash episodes (1998 Russia, 2008 GFC, 2011 Swiss-franc unpeg, 2015 EM rout, 2022 yen crisis) are not present in any given sample window. When you include the crashes, carry-trade Sharpe ratios drop from ~1.0 to ~0.3–0.4, comparable to other risk-bearing strategies. The forward premium itself can also be viewed as a risk premium: investors require compensation to bear the asymmetric crash risk inherent in funding-currency strength. Recent academic work (Lustig, Roussanov, Verdelhan; Brunnermeier, Nagel, Pedersen) treats carry as a systematic risk factor that earns a premium because it loads on funding liquidity and global risk aversion. So carry works on average but you are paid for taking real risk.
What is the difference between covered and uncovered carry?
Covered carry uses a forward contract to lock in the unwind exchange rate. By Covered Interest Parity, the forward rate is set so that F = S × (1 + r_high × t) ÷ (1 + r_low × t) — meaning the locked-in FX move EXACTLY offsets the interest differential, leaving net carry of zero (minus transaction cost). Covered carry is therefore essentially a zero-expected-return arbitrage in efficient markets; it works only when there is a CIP deviation, which since 2008 has occurred persistently in USD pairs due to bank balance-sheet costs (the so-called cross-currency basis), allowing tiny but real arbitrage trades that prop desks and hedge funds exploit. Retail or small-corp carry trades cannot capture this. Uncovered carry leaves the FX exposure open — you borrow in low-yield, hold in high-yield, and at unwind sell back at the prevailing spot rate. This is the standard "carry trade" of macro and EM funds, and it captures the empirical UIP failure described above, but at the cost of bearing FX volatility. Almost all carry trades referred to in industry are uncovered. This calculator models uncovered expected carry under the simplifying assumption of zero FX move; realized returns will differ by the actual FX P&L.
How does leverage actually work in carry trades and where does it come from?
FX is naturally leveraged because positions are typically funded via the broker's margin: a retail FX broker requires 3–5% margin (20-30x leverage) on majors and 5–10% on EMs (10-20x leverage). Institutional desks use prime-brokerage credit lines with even higher implicit leverage, or use FX swaps (sell USD spot, buy USD forward — effectively a USD-funded carry with no margin call risk on the FX leg). The leverageRatio field in this calculator represents your gross notional ÷ equity, so 3x means $300k notional on $100k of equity. Risk scales linearly: a 3% FX move costs 9% of equity at 3x leverage; a 7% move at 5x leverage is a 35% drawdown. Leverage also amplifies financing cost (in real markets the broker rollover swap is set by the interbank market and includes the rate differential — your funding cost is already netted into the carry, NOT a separate charge). The 0.002 cost in this formula is a rough proxy for spread + slippage + non-rate broker fees. Margin calls force liquidation at the worst possible time (during the FX move that hurts you), which is why naive high-leverage carry trades blow up disproportionately in crashes — see the 2008 Australian and New Zealand dollar moves that wiped out many JPY-funded retail carry traders.
Common mistakes with carry trades?
First, ignoring FX risk — the calculator's expected return is the carry only; the realized return is dominated by spot moves and a typical 10% adverse FX move can wipe 4 years of carry. Second, over-leveraging — 5x leverage on a 4% differential is 20% gross expected, but a 4% adverse FX move (within a single-day standard deviation for many EM pairs) becomes a 20% equity drawdown; 10x or 20x leverage is gambling, not investing. Third, ignoring crash risk and skew — carry returns are heavily left-skewed; a Sharpe ratio computed in calm periods understates the real risk. Fourth, ignoring transaction cost on EM pairs — the 20bps assumption in this formula is reasonable for major-pair institutional execution but understates EM by 5-10x for retail. Fifth, ignoring intervention risk — central banks (especially BOJ, SNB, Banxico, BOK) periodically intervene to defend or weaken their currencies, and the SNB unpeg of January 2015 was the textbook "impossible" event for short-CHF carry positions. Sixth, ignoring tax implications — carry interest is ordinary income in most jurisdictions, while FX gains may be capital gains, ordinary income, or 60/40 (US Section 1256 for futures), giving very different after-tax economics. Seventh, mistaking the high-yield rate for a risk-free yield — high-yield is high BECAUSE the currency is risky; the rate differential is fundamentally a risk premium, not a free lunch.
When should I not use this calculator?
Skip it for short-term tactical FX trades (<7 days) where the carry component is negligible and the model adds nothing beyond a directional FX call. Skip it for option-based carry strategies (selling vol to harvest premium, fenced via wings) — those have completely different risk/return profiles and require Black-Scholes-style valuation. Skip it for any carry trade where you don't have an explicit view on the FX direction — the carry math is the tip of the iceberg; the FX volatility and crash risk are 80% of the actual P&L. Skip it for emerging markets with capital controls (CNY, ARS, NGN, EGP) where the realized funding rate, the FX execution, and the unwind path are all complicated by non-market frictions. Skip it for situations where you can't survive a drawdown (e.g., retail trading with margin call risk) — carry trades historically have multi-year drawdowns of 30%+ during global risk-off episodes, and forced unwinds at the worst time destroy the expected return. For institutional sizing of carry exposures, use a multi-factor risk model that includes vol, skew, funding-cost risk, and political risk; this calculator gives the rough "if nothing moves" baseline only.