A carry trade is one of the oldest ideas in currency markets: borrow a currency with low interest rates and use the proceeds to hold one with higher rates, keeping the difference. It sounds simple, and for long stretches of calm markets it works quietly — small daily gains that add up. The catch is equally old: when the trade unwinds, the exchange-rate move can erase years of accumulated carry in days. Understanding both sides matters before you ever let positive swap quietly become part of your strategy.

What a carry trade actually is

The Wikipedia entry on carry describes it plainly: a trade "in which profit is incurred from the spread between borrowing a low carry asset and lending a high carry one." In foreign exchange, that means borrowing (or selling) a low-yielding currency to buy a higher-yielding one.

The textbook example for decades has been the yen. The Bank of Japan has kept Japanese rates at very low levels since the mid-1990s, which made the yen the classic funding currency — the currency you borrow or sell — for positions in higher-yielding currencies. The US dollar and the Japanese yen have been the most heavily used currencies in carry transactions since the 1990s, and the euro took on a similar funding role after the ECB's accommodative policy from 2015 onward.

Two terms are worth fixing in your head:

• Funding currency: the low-rate currency on the borrowed side of the trade (historically yen, franc, and later euro).

• Investment (target) currency: the higher-rate currency you hold — often, in recent years, emerging-market currencies like the Mexican peso.

If the interest-rate gap between the two is wide and the exchange rate stays roughly stable, you collect the differential. That daily accrual is the "carry."

Where retail traders meet carry: the swap line

For a retail forex trader, the carry trade usually shows up not as a deliberate macro strategy but as the swap (rollover) charge or credit applied to positions held overnight. Hold a long position in a currency pair where the base currency has the higher rate, and you may receive a small credit each night; hold the other side and you pay it.

We explain that mechanic in detail in our rollover and swap rates explainer and in what a forex swap is. The important connection for this article: swap rates are the retail reflection of the same interest-rate differential that institutional carry traders exploit — and they are set by your broker per instrument and entity, so identical trades can carry different amounts at different firms.

An illustrative worked example

The following numbers are an illustrative example, not a broker quote or a forecast. Assume a hypothetical 4.5% annualized interest-rate differential in favor of the US dollar, and a 1-standard-lot long USD/JPY position (100,000 USD notional).

• Daily carry ≈ 100,000 × 4.5% ÷ 360 ≈ $12.50 per day, or about $4,500 per year before costs.

• Now suppose the yen strengthens 5% against the dollar. On the same 100,000 USD notional, the exchange-rate loss is roughly $5,000.

• That single adverse move costs about 400 days — more than a year — of accumulated carry to offset, and it can happen in a single session.

• At 30:1 leverage, the margin behind that position is about $3,333; a 5% adverse move equals roughly 150% of the required margin, meaning the position's equity can be wiped out long before the trade "earns back" its carry.

The asymmetry is the lesson: carry accrues slowly and linearly, while unwind losses arrive fast and non-linearly. (This simplified arithmetic ignores spread, swap adjustments, slippage, and liquidation behavior, and is not a prediction of any actual trade.)

Why it is not arbitrage

If carry were risk-free profit, markets would eliminate it instantly. The Wikipedia carry entry stresses that carry trades "turn profit only if nothing changes against the carry's favor, and thus they are not necessarily arbitrages, which need to be risk-free."

What you are being paid for is risk. Researchers have long noted the pattern that high-yield currencies tend to depreciate in bad times — the exchange-rate equivalent of an equity crash premium, sometimes called the "peso problem." Two well-documented episode types make the point:

• 2008: the yen carry trade largely collapsed in 2008, contributing to a rapid yen appreciation as leveraged positions unwound in a self-reinforcing loop. The 2008–2011 Icelandic financial crisis also had among its origins the undisciplined use of carry trades, including euro-denominated loans that became unaffordable when the Icelandic currency depreciated dramatically.

• Early August 2024: the BIS Quarterly Review of September 2024 describes how, after perceived hawkish signals from the Federal Reserve and the Bank of Japan and a disappointing US labour release, "leveraged positions, above all carry trades, came under pressure, with the yen appreciating sharply." Carry trades unwound, causing "a sharp, if short-lived, appreciation of the funding currencies, predominantly the yen, and a depreciation of the investment currencies, such as the Mexican peso and other emerging market economy (EME) currencies."

Crowding makes this worse. The same BIS review found hedge fund strategies "had become crowded and more exposed to carry trades for at least a year" before the August 2024 turbulence, with returns to global macro and managed futures funds showing increasing sensitivity to carry-trade returns from 2022 onward. When many participants hold the same leveraged bet, an unwind forces everyone to exit through the same narrow door — which accelerates the move that triggered the exit.

What actually goes wrong, mechanically

An unwind is a feedback loop:

1. A shock — a weak jobs report, a hawkish central bank, a volatility spike — raises the perceived risk of the high-yield currency.

2. Leveraged carry holders sell the investment currency and buy back the funding currency to cut exposure.

3. That buying pushes the funding currency up and the investment currency down, worsening every remaining position's losses.

4. Margin calls and forced liquidations add more of the same selling.

The BIS noted that the August 2024 moves, while sharp, were "large but not outsize compared with past carry trade crashes, and foreign exchange (FX) volatility did not rise nearly as much as that of equity markets," and markets stabilized within days. That nuance matters: carry unwinds range from a nasty multi-day squeeze to a systemic event. You cannot know in advance which kind you will meet.

For a retail trader, the practical exposure points are:

• Leverage. Because carry trades feel slow and steady, they tempt holders into size. Our explainer on ESMA/FCA-style retail leverage caps covers why major regulators limit retail CFD leverage to levels where a normal adverse move does not automatically liquidate an account.

• Gap and slippage risk. Unwinds are exactly when stop-loss orders can execute far from their level. See how forex slippage works and why weekend gaps happen.

• Negative swap on the other side. If you hedge carry exposure by holding both directions of a pair, both legs can cost you: swap schedules are typically asymmetric, so the "safe" hedge still bleeds financing.

Practical checks before carry matters to your account

You do not need to run a macro carry portfolio for these to be relevant. Overnight positions in trending pairs already embed carry.

• Read the actual swap/rollover values for your instrument on your broker's trading conditions page — per pair, per direction, and per contracting entity, because they differ.

• Treat a positive swap as a minor offset, not a reason to hold a position you would otherwise close. A few dollars per lot per night will not survive an ordinary adverse move on a leveraged position, as the example above shows.

• Size positions so that a funding-currency-style squeeze of the kind seen in 2008 or August 2024 is survivable — meaning your account can tolerate a fast move several times larger than the daily carry you are collecting.

• Remember that small retail traders have access to limited currency pairs, and yields shrink further after spreads and costs are factored in, as the Wikipedia carry entry notes. Whatever headline differential a strategy shows on paper, your realized carry is the broker's swap schedule minus trading costs.

Carry is a real component of currency returns — but it is compensation for taking crash risk, not a coupon. The traders who keep their accounts are usually the ones who sized for the unwind, not the ones who counted the nightly credits.

Frequently Asked Questions

Is the carry trade the same as collecting positive swap?

For a retail trader, effectively yes at the position level: a long overnight position that earns positive swap is a small carry position. Institutional carry trades are the same idea run deliberately, with leverage and across currencies. The swap line on your statement is the retail expression of the interest-rate differential.

Why did the yen carry trade unwind in August 2024?

Per the BIS's September 2024 review, a combination of a disappointing US labour-market release and perceived hawkish communication from the Federal Reserve and Bank of Japan triggered deleveraging. Carry trades unwound, the yen appreciated sharply against investment currencies such as the Mexican peso, and the squeeze amplified global volatility — though it faded within days.

Can you lose more on carry trades than the interest you earn?

Yes, easily. The exchange-rate leg is many times larger than the interest leg for any position held over a realistic horizon. A 5% adverse move on a 1-lot position costs roughly 400 nights of carry at an illustrative 4.5% differential — and at 30:1 leverage, that move alone exceeds the position's margin.

Is carry trading banned or restricted?

No regulator bans carry trading as such. What major regulators restrict is retail leverage and risk features of CFDs, which indirectly caps how much carry a leveraged retail trader can accumulate. Swap amounts themselves are commercial terms set per broker and entity.