11 min read On 5 August 2024 the TOPIX fell 12% in a session and the VIX spiked above 60, with nothing whatever wrong with Japanese earnings. What broke was the yen carry trade: borrowing cheaply in Japanese yen and putting the money somewhere that pays more, then pocketing the difference for as long as the exchange rate behaves.
Traders call the yen the funding currency and the gap between the two rates the carry. The catch sits in the repayment. The moment the yen strengthens you owe more than you borrowed, and in that first week of August the Bank of Japan had raised rates five days earlier, a position built quietly over years came apart in 72 hours, and the S&P 500 went down 3% with it.
Why the yen became the world’s funding currency
Japan spent two decades with policy rates at or below zero. Nowhere else in the developed world offered borrowing that cheap for that long, and the yen is liquid enough to borrow in size without moving the market against yourself. Those two facts together made Japan the funding source of choice for leveraged investors worldwide.
The arithmetic as of late September 2026 still works, though less generously than it did. The Bank of Japan’s policy rate sits at 1.25% and the Federal Reserve’s target range is 3.75% to 4%. Borrow in yen, hold dollars, and about 275 basis points a year accrues to you before the exchange rate does anything at all.
On its own that is a modest return. Applied at five times leverage it becomes 13.75%, and that is where the strategy gets both its reputation and its fragility.
How a yen carry trade actually works
Follow a single position through, with round numbers chosen for clarity rather than realism.
- Borrow. Take a ¥157 million loan at 1.25%. Annual interest cost: about ¥1.96 million.
- Convert. At 157.00, that buys $1 million.
- Invest. Put the dollars into an instrument yielding 4%. Annual income: $40,000.
- Net the carry. The interest cost converts to roughly $12,500. The position earns about $27,500 a year on $1 million of exposure, or 2.75%.
- Repay. At the end you convert back to yen and repay the loan.
Step five is the whole story. If USDJPY is still 157, you keep the carry. If the yen has strengthened to 145, your $1 million now buys ¥145 million against a ¥157 million debt. The currency loss is ¥12 million, roughly $83,000, against $27,500 of carry earned. Four years of income gone in a move the pair has made in a fortnight.
That asymmetry is the defining feature. Carry accrues slowly and in small amounts. The currency risk arrives quickly and in large ones.
What the carry trade is not
It is not arbitrage. Arbitrage is riskless by definition, and this position carries unhedged exchange rate risk as its entire economic content. Hedging the currency exposure removes the risk and, through covered interest parity, removes the profit along with it. The return exists precisely because the risk is not hedged.
It is also not a Japanese phenomenon in its effects. Yen borrowed in Tokyo ends up in Mexican bonds, Australian deposits, US equities and emerging market credit. The trade exports Japanese monetary policy to every market that receives the money, which is why a Bank of Japan meeting can matter for a technology stock in California.
How large is the yen carry trade?
Nobody knows exactly, and anyone quoting a single confident number is guessing. The Bank for International Settlements, which studied the 2024 unwind in detail, measured several different slices and stated plainly that positions are difficult to estimate.
| Measure | Size | What it captures |
|---|---|---|
| Yen currency futures short positions | ¥2tn, about $14bn | Speculative shorts at their peak, the visible tip |
| Hedge fund yen forwards | About $160bn | Estimated off-exchange speculative exposure |
| Foreign bank yen funding to non-banks outside Japan | ¥14tn, about $90bn | Cross-border lending into the trade |
| Yen-denominated loans to non-banks outside Japan | ¥40tn, about $250bn by March 2024 | Total offshore yen borrowing |
| Cross-border yen claims on offshore financial centres | ¥80tn, about $500bn | The widest plausible boundary |
The range from $14 billion to $500 billion is not sloppiness. The narrow measures capture speculators, the wide ones capture everything yen-funded including positions that are not carry trades at all. The honest summary: large enough that an unwind moves global markets, and impossible to monitor in real time, which is exactly what makes it dangerous.
The August 2024 unwind, step by step
The best documented carry trade unwind in history took less than a week and is worth walking through, because the sequence repeats.
24 July 2024. A technology sell-off wiped roughly $1 trillion from valuations. Leveraged investors holding yen-funded equity positions took the first losses and began trimming.
31 July 2024. The Bank of Japan raised rates, and the market read the statement as hawkish. The rate gap that made the trade profitable narrowed, and the yen, which everyone was short, started to appreciate.
2 August 2024. Weak US labour data landed. The market moved to price faster Federal Reserve cuts, narrowing the gap from the other side simultaneously.
5 August 2024. Everything arrived at once. The TOPIX lost 12% in a day. The S&P 500 fell 3%. The VIX spiked above 60. The Japan Securities Clearing Corporation raised margin on equity index longs by 60% to 80%, forcing more selling from anyone already under pressure. Retail traders faced margin calls across several asset classes at the same time.
9 August 2024. The S&P 500 had recovered everything it lost that week.
Read that last line carefully. The fundamentals of US companies did not change between Monday and Friday. What happened was a forced deleveraging: a stronger yen made every yen-funded position more expensive to hold, holders sold assets to repay yen loans, the selling pushed the yen up further, and the loop ran until the leverage was gone. Then it stopped, and prices went back.
Why an unwind is so violent
Three structural features turn an ordinary loss into a cascade.
Everyone is on the same side. The carry trade is a consensus position by construction, because it is driven by a rate gap everyone can see. When it reverses, there is nobody positioned to take the other side of the exit.
The exit tightens the loop. Closing the position means buying yen. Buying yen strengthens the yen. A stronger yen increases the loss on every position still open, which forces more closing. The mechanism feeds itself in a way that an equity sell-off does not.
Margin moves against you at the worst moment. Clearing houses raise requirements when volatility spikes, which is precisely when holders have the least capacity to post more. The 2024 episode shows the effect clearly: the margin increase came after the losses and made them worse.
The BIS conclusion from the episode is worth stating directly. Positions accumulate during calm periods because low volatility makes the strategy look safe, and that accumulation is what makes the eventual unwind rapid. The quiet stretch is not the absence of risk. It is the risk being built.
Japan’s rate cycle and the reverse carry trade
The structural question now is what happens to a trade whose entire premise was that Japanese rates never move.
They are moving. Japan has gone from negative rates to 1.25% and the direction is upward. Each increase does two things: it raises the cost of borrowing yen, which shrinks the carry, and it makes yen assets more attractive to Japanese investors, who hold enormous foreign portfolios and may repatriate.
The reverse carry trade is the mirror image: borrowing in a currency whose rates are falling and holding yen while Japanese rates rise. That position is still unattractive on the arithmetic, because 275 basis points is still a lot of carry to pay for the privilege. It becomes attractive on expectations long before it becomes attractive on carry, which is exactly what makes the turn hard to time.
What would actually trigger the next unwind is a narrowing of the gap faster than the market expects, from either side. A Federal Reserve cutting quickly, a Bank of Japan hiking faster than the September 2026 vote suggested, or a risk shock that sends money into the yen for reasons unconnected to rates. The 2024 episode had two of those three at once.
What this means for trading USDJPY
You do not need to run a carry trade to be affected by one. If you trade the pair at all, the position sits underneath your chart.
Yen strength arrives faster than yen weakness. The pair grinds higher and falls in steps. That asymmetry is the carry trade building slowly and exiting quickly, and it argues for wider stops on short yen positions than a symmetric view of volatility would suggest.
Watch positioning, not just rates. When speculative short yen positions are near record levels, the fuel for a violent reversal is already in the tank. The trigger is almost never predictable; the vulnerability is visible in advance.
Equity drawdowns and yen strength travel together. If you hold both a short yen position and long equity exposure, those are not diversified holdings. August 2024 hit both simultaneously, for the same reason.
Overnight financing is the retail version of the same trade. Holding a long USDJPY position past the daily rollover earns a credit derived from the rate differential. That is carry, at a smaller scale, with the same exposure to a currency reversal. As Japanese rates climb, the credit shrinks. Anyone holding the pair for the swap alone is running a strategy whose economics are quietly changing.
For the fuller picture of the drivers behind the pair, see what moves USDJPY. For the events that interrupt an unwind, the guide to Japanese yen intervention covers what happens when Tokyo steps into the market. And for the execution side, including sizing around a pair that falls faster than it rises, see how to trade USDJPY.
The trade that broke global markets in August 2024 had been working, quietly and profitably, for years beforehand. That is the part worth remembering. A strategy that pays reliably for a decade and then loses a decade of gains in three days is not a strategy with an occasional bad week. It is a strategy whose losses are simply scheduled differently from its profits.
What is the yen carry trade?
Borrowing money in Japanese yen at a low interest rate and investing it in a higher-yielding asset elsewhere, keeping the difference between the two rates. The yen is the funding currency. The profit is the rate gap, and the risk is that the yen strengthens before the position is closed, which makes repaying the loan more expensive than the carry earned.
Why does the yen carry trade cause market crashes?
Because closing the position means buying yen, which strengthens the yen, which increases the loss on every position still open and forces more closing. The loop feeds itself. In August 2024 that mechanism took the TOPIX down 12% in a day and the S&P 500 down 3%, with no change in company fundamentals.
How big is the yen carry trade?
Estimates range from about $14bn in visible currency futures shorts to roughly $500bn in cross-border yen claims on offshore financial centres, depending on what is counted. The Bank for International Settlements measured yen-denominated loans to non-banks outside Japan at ¥40tn, about $250bn, as of March 2024, and states that positions cannot be estimated precisely.
Is the yen carry trade still profitable in 2026?
The gap is narrower than it was. The Bank of Japan's rate is 1.25% and the Federal Reserve's target range is 3.75% to 4%, leaving about 275 basis points before leverage and before any exchange rate move. That is a smaller cushion against yen appreciation than the trade carried through most of the past decade.
What is a reverse carry trade?
The mirror position: funding in a currency whose rates are falling and holding one whose rates are rising. In Japan's case that means holding yen while the Bank of Japan tightens. It turns attractive on expectations well before it turns attractive on the rate arithmetic, which is what makes the timing difficult.
What triggers a carry trade unwind?
A narrowing of the interest rate gap faster than the market expected, from either side, or a volatility shock that sends money into the funding currency for unrelated reasons. August 2024 had a hawkish Bank of Japan, weak US labour data pointing to faster Federal Reserve cuts, and an equity sell-off inside 12 days.
Author
Risk Disclaimer
The content provided here is for informational purposes only. It is not intended as personal investment advice and does not constitute a solicitation or invitation to engage in any financial transactions, investments, or related activities. Past performance is not a reliable indicator of future results.
The financial products offered by the Company are complex and come with a high risk of losing money rapidly due to leverage. These products may not be suitable for all investors. Before engaging, you should consider whether you understand how these leveraged products work and whether you can afford the high risk of losing your money.
The Company does not accept clients from the Restricted Jurisdictions as indicated in our website/ T&C. Some services or products may not be available in your jurisdiction.
The applicable legal entity and its respective products and services depend on the client’s country of residence and the entity with which the client has established a contractual relationship during registration.
