If you take a quiet Tuesday afternoon stroll through the trading floors of any major bank in Tokyo or London, you’ll see something that has returned over the past year. The discussions about yen positioning have resumed; they are cautious and subtle, the kind of conversation that is accompanied by astute glances rather than loud declarations. In August 2024, the carry trade strategy—which involved borrowing cheap Japanese yen and investing it in higher-yielding assets halfway around the world—collapsed spectacularly. It erased over $670 billion in market value in a single session, sent Japan’s Nikkei to its worst day since 1987, and left traders nursing losses they hadn’t anticipated despite the setup being obvious to anyone who was willing to look. And now the trade is being rebuilt with a patience that verges on obstinacy.
Silently. large-scale. with an institutional memory that appears strangely short from the outside. Once you get the hang of it, the yen carry trade’s mechanics are incredibly easy. The structure of international capital flows bent around Japan because it maintained interest rates at or close to zero, and for a while, below zero.
| Topic Overview: The Yen Carry Trade — Crash, Recovery & Rebuild | |
|---|---|
| Strategy | Carry trade — borrow in low-rate currency (yen), invest in higher-yielding assets abroad |
| Crash Date | August 5, 2024 — Japan’s Nikkei fell 12%, worst single day since 1987 |
| Market Value Lost | Over $670 billion in a single session on Japan’s markets alone |
| Trigger | Bank of Japan raised rates by 0.25% on July 31, 2024 — yen surged, forced unwind began |
| Estimated Peak Trade Size | At least $500 billion (dollar-yen carry trade alone, per UBS strategist James Malcolm) |
| Amount Unwound (Aug 2024) | ~$200 billion unwound in 2–3 weeks; JPMorgan estimated 75% of carry trade removed |
| Key Analyst | Vineer Bhansali, Founder & CIO, LongTail Alpha — longtime carry trade researcher |
| Assets Affected | US tech stocks, Mexican peso, Australian bonds, Brazilian real, global equities |
| BOJ Rate Context | Japan held near-zero or negative rates for years — structural foundation of the trade |
| Current Status | Traders returning to yen-funded positions through 2025, rebuilding leverage quietly |
| Key Risk | Any further BOJ rate movement or Fed policy shift could trigger another forced unwind |
Tokyo investors, institutional behemoths like the Government Pension Investment Fund and the Japan Post Bank, as well as the fabled character of Mrs. Watanabe, which stands for the army of Japanese retail investors looking for higher returns overseas, borrowed yen at almost no cost and transferred the funds into US bonds, Australian debt, Brazilian assets, and investments denominated in Mexican pesos. Staying in Yen offered nothing for a long time, but something is always preferable to nothing. One of the more insightful writers on the carry trade’s connection to volatility, Vineer Bhansali, founder and CIO of LongTail Alpha, put it succinctly: print funny money, buy other people’s bonds, and extract real dollars, euros, and pesos forever. The tactic was successful. for many years. Then, on July 31, 2024, the Bank of Japan increased interest rates by a quarter of a percentage point, and the entire system started to collapse.
When the unwind occurred, it had the appearance of a slow-motion avalanche that abruptly stopped being slow. Investors who had borrowed in yen saw their positions moving against them more quickly as the value of the currency increased. The forced selling was not limited to Japan; it spread to the Mexican peso, American tech stocks, and almost every asset class that had taken in yen-funded capital over the lengthy period of Japanese monetary accommodation. In the days after the crash, London-based UBS Japan macrostrategist James Malcolm calculated that the dollar-yen carry trade had reached a peak of at least $500 billion and that possibly $200 billion had been unwound in the two to three weeks preceding the chaos. In contrast, JPMorgan estimated that during the sell-off, 75% of the larger carry trade had been eliminated. In any case, the figures revealed a tactic that had grown massive in the shadows and remained undetectable until it wasn’t.
One year later, it’s remarkable how many of the traders who were burned in August 2024 have resumed using the same tactic. In August 2025, Bloomberg reported that investors who survived the crash were reconstructing yen-funded positions, drawn by a rate differential between the US and Japan that, although smaller than at its peak, still presents a significant opportunity to profit. This could be seen as reasonable because the BOJ has been cautious since its hike in July 2024, and the carry trade’s underlying logic hasn’t changed; the circumstances surrounding it have only momentarily changed. It could also be seen as one of those recurrent trends in the financial markets where the lesson is assimilated intellectually without affecting actual behavior. It is possible for both to be true at the same time.
Observing this rebuild gives me the impression that the opacity surrounding carry trades is primarily responsible for keeping the risk undetectable. In the wake of the 2024 crash, The Economist made this point explicitly, claiming that far too little is known about the true scope of carry trades and the risks involved. Currency carry trades don’t show up neatly on any centralized ledger, in contrast to stock or bond holdings. They are constructed through foreign exchange swaps, derivatives, and overlapping positions in a variety of instruments that, while appearing ordinary on their own, add up to much more. When the BIS looked into the August 2024 incident, analysts found that the circumstances leading up to the crash had been especially favorable for the accumulation of leveraged positions. This is a measured way of saying that the conditions for a blow-up were apparent overall, but the aggregate wasn’t being closely monitored by anyone with the authority to slow it down.
In some ways, the world in which the current reconstruction is taking place is more precarious than it was prior to August 2024. Oil prices have increased dramatically as a result of the Iranian conflict, and central banks in several nations are still adjusting for inflation. As we approach the second half of 2026, the Federal Reserve’s rate trajectory is still genuinely uncertain. Additionally, the Bank of Japan is no longer a guaranteed anchor of zero because it has shown that it will raise rates when it deems it appropriate. The final point merits more consideration than it currently receives. The fundamental tenet of the carry trade—that Japan’s interest rates remain low indefinitely—has already been put to the test. In essence, traders who rebuild the position are placing a wager that the BOJ won’t do it again or won’t do it quickly enough to be significant. That might be accurate. It’s still unclear if the BOJ has completed its adjustment, and anyone who was certain of the solution prior to August 5, 2024, most likely recalls what happened after that.
Regardless of what happens to carry trades, the financial system appears to have a true structural preference for them. As soon as the immediate pain subsides, the yield differential creates an almost gravitational pull that reemerges. The tactic has previously caused market crashes. It is currently being rebuilt. Furthermore, the circumstances that could unwind it a second time—a quicker-than-anticipated BOJ move, a Fed pivot, or an abrupt risk-off shock in international markets—have not been eliminated. They’ve just moved out of sight. This is when these things get the biggest historically.