A cash carry trade combines funding in one currency with an asset in another. A positive interest spread can be offset by exchange-rate changes, funding renewal and costs. A long USD/JPY retail position has product-specific financing terms; it does not automatically credit the difference between Fed and BoJ policy rates every night.

What is the USD/JPY carry trade, really?

In a cash carry trade, borrow a funding currency, exchange it and invest in an asset denominated in another currency. A long USD/JPY position gives positive USD exposure relative to JPY, but its interest or financing treatment depends on the instrument. It need not involve taking a bank loan or owning an interest-bearing deposit. The separate article on carry trade explains the broader structures.

JPY has historically been used as a funding currency during periods of low Japanese rates. Policy settings change: the BoJ announced the end of its negative-interest-rate policy in March 2024. Use dated decisions and actual funding quotes rather than treating negative Japanese rates or high US rates as permanent assumptions.

How to compare funding terms and maturities

A funding choice depends on the actual borrowing rate, term, availability, collateral and currency risk. A central-bank policy rate is context, not a borrowing offer available to every trader. Historical low Japanese rates do not prove that JPY is always the cheapest funding currency.

Carry depends on realised asset income, financing and exchange-rate changes, including any renewal. Covered interest parity relates comparable hedged rates and forward quotes; it is not a guaranteed spot-return rule. The articles on interest-rate parity, central-bank monitoring and BoJ policy develop those distinctions.

Carry can lose with the same final rate

Hypothetically borrow 150000 JPY and buy 1000 USD at 150 JPY/USD. The one-year deposit earns 5% simple interest, but funding is fixed for only the first three months at 1% annually. The remaining nine months renew at 8%.

Scroll sideways to see the full diagram

A yearly deposit and funding renewed after three monthsA fictional 1000 USD deposit earns 5% simple interest for one year. Funding 150000 JPY costs 1% annually for three months and 8% for nine months: 375 and 9000 JPY. At the same final rate 150 JPY/USD, the result after both costs is − 1875 JPY.The maturities do not matchUSD deposit: 12 months, 5% annually1000 USD → 1050 USDJPY loan: renew after 3 months1%8% annually0312 monthsCost: 375 + 9000 = 9375 JPY1050 × 150 = 157500 JPY157500 − 150000 − 9375= − 1875 JPY
The funding bars represent 3 and 9 months. The final exchange rate is an assumed scenario value, not a hedge.

The deposit ends at 1050 USD, or 157500 JPY at the assumed rate. Loan interest totals 375 + 9000 = 9375 JPY. The result is − 1875 JPY. If 1% applied for the whole year, the model would give + 6000 JPY, but that was not the funding agreement. Record the asset maturity separately from the date through which financing costs are known.

Assumptions: Fictional cash model, not a CFD tariff. Periods are exactly 0.25 and 0.75 year; interest is simple. The same 150000 JPY principal renews. The initial 375 JPY interest is paid from an extra reserve and included in the total result without compounding. Renewal is assumed available; final rate 150 JPY/USD and no other costs. Official material: BIS: currency and maturity mismatches.

Pips, financing entries and the complete result

Start with units. In this hypothetical USD/JPY model, one pip is 0.01 JPY per USD and the provider’s tick is 0.001. For USD 10,000, they change price P&L by JPY 100 and JPY 10 respectively. A quoted rate of 150.020 means 150.020 JPY per USD. Higher USD/JPY benefits the long position before costs; a stronger yen against USD means a lower quote.

Assume a linear CFD and a JPY account, full entry at Ask 150.020 with initial Bid 150.000, and JPY 400 total entry-and-exit commission. Three hypothetical funding entries are +120, +180 and +300 JPY: +600 total, already net of charges included in those entries. No other fee or conversion applies. The values are educational assumptions, not current swap rates or an IG offer.

Alternative full exits: hypothetical USD 10,000 position
Exit Bid 150.000Price −200 JPY; funding +600; commission −400; net 0 JPY.
Exit Bid 151.000Price +9,800 JPY; funding +600; commission −400; net +10,000 JPY.
Exit Bid 147.000Price −30,200 JPY; funding +600; commission −400; net −30,000 JPY.

Price P&L is 10,000 × (exit Bid − 150.020). Add JPY 600 and subtract JPY 400 once. The entry spread is already inside the fills: a return to initial Bid loses JPY 200 before those adjustments. Funding here covers only six pips of this exposure. With the same notional, changing leverage does not change pip value; it changes margin requirements. Larger notional changes price exposure and potentially the charges. A real account may be closed out before an assumed final exit; this model does not simulate that path.

Actual CFD financing is product-specific. IG Australia describes tom-next plus an administrative charge for its forex CFDs; it is an example of provider terms, not every broker’s tariff. Check the sign, unit, volume basis, cutoff and day multiplier. A single positive entry multiplied by 365 is not promised annual income. See changing swap rates and how swap works.

Exchange-rate losses, conversion and intervention risk

In a separate conversion-only model, buy USD 10,000 at 150 JPY per USD and assume alternative exits at 147 and 153. Use a single price, no spread, commission or funding, and convert the final JPY price result at the respective exit rate. Both moves are 300 pips: −30,000 / 147 ≈ −204.08 USD, versus +30,000 / 153 ≈ +196.08 USD. Equal JPY amounts do not become equal dollar amounts because the denominators differ.

The formula is N × (S₁ − S₀) / S₁. It is an assumed final conversion convention, not the broker’s executable conversion rate. Neither scenario has an assigned probability, so this pair alone establishes no expected return. In a real account, convert financing and fees at their actual settlement times. Leveraged carry unwinding can amplify policy reactions, as BIS Bulletin 124 explains; it does not give a guaranteed USD/JPY path.

JPY has displayed safe-haven behaviour in some periods, but the response depends on the episode and comparison currency. It is not guaranteed by a zero-rate label. Compare the Swiss franc using dated observations rather than assuming both havens always behave alike.

Japan's Minister of Finance decides on foreign-exchange intervention and the BoJ executes it as the Minister's agent. Intervention can buy or sell yen. Do not infer a standing trigger at 150–160 JPY per USD or a guaranteed number of pips from past operations; check the dated official record.

Does the carry trade make sense for a retail trader?

A retail carry position may use leverage, but leverage is not required for every carry structure. Financing often relates to the contractual exposure while margin is only part of the funds required. Compare realised interest, conversion, costs and possible losses; client size alone does not establish who suffers more from a reversal.

Use the product's financing cutoff and holding-period rules when calculating carry; even a short holding period can cross a charge or credit boundary. Monitor Fed and BoJ decisions, funding renewal and available terms. Comparing USD/MXN requires its own rates and risk assumptions.

Your next step with the carry trade

Separate the contractual financing income from price risk and funding renewal. Carry can lose even at an unchanged final exchange rate when renewed funding costs exceed asset income, as the cash model shows. Review the whole result rather than treating every credit as profit.

  1. Check financing units and dates. Record long and short rates, the calculation mode, cutoff, currency conversion and any multi-day charge. A comparison with policy rates does not isolate broker markup because market funding rates and the product structure can differ.
  2. Add both banks’ meetings to your calendar. Mark the next decision dates for the Fed and the Bank of Japan. For each, note what the market expects — document the expectation and actual release separately. Recheck the contract’s funding terms and renewal date; two policy rates do not describe every component of the position’s result.
  3. Stress-test the total result. Use dated prices, realised financing and costs to review several adverse scenarios. Calculate price loss and cumulative carry in the same currency; a chart alone does not show how many months of actual interest a move offsets.