On August 5, 2024, the Tokyo Stock Exchange's TOPIX index fell 12% in a single session. The VIX — Wall Street's fear gauge — briefly spiked to levels not seen since the Covid crash of March 2020. US Nasdaq stocks sold off hard. The Mexican peso collapsed. The Australian dollar cratered. And the proximate cause of all of it was a 25 basis point rate hike from the Bank of Japan and a slightly cautious US payrolls number. Markets recovered within days and the consensus quickly settled on a narrative that the worst was over, that 65-75% of carry trade positions had been unwound as JP Morgan estimated, and that the episode was a contained technical event rather than a systemic warning.1 That consensus is almost certainly wrong. What August 5 demonstrated is not that the carry trade has been resolved. It demonstrated how violent the resolution can be when it starts moving — and the structural conditions that generated the trade in the first place have not been dismantled. They are being eroded, slowly and unevenly, by a BOJ that has now raised rates to a 30-year high and is not finished.
The yen carry trade is one of those market structures that everyone in finance knows about, most people underestimate, and almost nobody can measure precisely. The mechanics are straightforward. Japan has maintained near-zero or negative interest rates for the better part of three decades. Borrowing in yen was essentially free. You borrow yen, convert it to dollars, euros, or any higher-yielding currency, invest in assets offering a positive return, and pocket the spread. The trade works beautifully as long as the yen stays weak or stable and volatility remains low. It falls apart catastrophically when yen strength and volatility arrive simultaneously, because the currency loss on your funding leg can exceed the entire return you have been collecting on your investment leg — in a matter of days or even hours.
Why the size cannot be measured and why that matters
One of the most dangerous features of the yen carry trade is that nobody knows exactly how large it is. The Bank for International Settlements, in its definitive August 2024 bulletin on the episode, estimated the narrow FX carry trade at around ¥40 trillion or $250 billion going into the event, but explicitly acknowledged this was biased downward due to data gaps in off-balance sheet activity.2 A broader measure incorporating yen-denominated cross-border bank lending to non-banks reached $880 billion in Q1 2024. Add in derivatives exposure through FX swaps and forwards — the instrument most commonly used by hedge funds to express carry positions synthetically — and the BIS's broader estimate reaches $1.7 trillion.3 Deutsche Bank has cited figures as high as 505% of Japanese GDP, though that captures the entire stock of yen-related financial activity rather than specifically speculative carry positioning.
The reason the size uncertainty matters so much is that it means nobody can reliably estimate how much unwinding remains. When JP Morgan said 65-75% of carry positioning had been cut by mid-August 2024, that estimate was based on observable CFTC positioning data for speculative yen futures — a relatively small and transparent slice of a much larger iceberg.4 The bulk of the carry trade lives in OTC derivatives, bilateral bank lending, and the unhedged foreign asset positions of Japanese institutional investors — life insurers, pension funds, and regional banks that have been buying US Treasuries, European bonds, and global equities for decades because domestic returns were inadequate. That institutional carry trade does not show up in CFTC data. It does not unwind in days. It unwinds over months and years as the economics of holding foreign assets on an unhedged basis gradually deteriorate against a rising yen and rising domestic yields.
The anatomy of August 5
To understand why August 5 happened the way it did, you need to understand the specific chain of events and why it produced such an unusual pattern of asset price moves. The standard carry trade unwind narrative — sell high-yield EM currencies, buy yen — only partially explains what occurred. The more interesting piece is what happened to US technology stocks.
The Wellington Management analysis of this episode makes an observation worth sitting with. The most dramatic sell-off during the unwind was in US momentum stocks — not in emerging market currencies or credit, which is where carry trade unwinds normally show up first.5 The reason is that Japanese investors and hedge funds running yen carry had deployed the proceeds disproportionately into the most liquid, highest-returning assets available, which in 2023 and 2024 meant US large-cap technology. When the yen moved and volatility spiked, they sold what they had the most of and what they could sell fastest. The Nasdaq became, in a very direct sense, collateral for a Japanese monetary policy decision. That is the kind of cross-market linkage that standard risk models completely miss and that makes the carry trade a systemic issue rather than a bilateral FX story.
What has actually changed since August 2024
The BOJ has continued raising rates. December 2024 brought the policy rate to 0.75%, a 30-year high.6 The 10-year JGB yield has pushed toward and above 2% — a level not seen since 1999. The interest rate differential between Japan and the US has narrowed from its peak. Net short yen positions among leveraged funds declined roughly 40% from their November 2024 peak, according to CFTC data, suggesting meaningful position reduction by the most observable segment of carry traders. These are all genuine changes. None of them mean the trade is over.
The reason is arithmetic. Even at 0.75%, the BOJ's policy rate is deeply negative in real terms against inflation running at 2-3%. The US Fed funds rate, even after cuts, remains well above 3%. The interest rate differential that funds the carry trade has narrowed but is nowhere near closed. Morgan Stanley estimated, as of mid-2025, that roughly $500 billion in outstanding yen-funded carry positions remained despite August's partial unwind.7 The trade still pays. It just pays less, costs more, and carries materially more risk of a violent reversal than it did when the BOJ was at zero.
Carry trades do not fear today's rate. They fear tomorrow's path. The BOJ has made tomorrow's path unmistakably clear. It is higher. The pace is the only open question.
The repatriation risk that dwarfs the carry trade
The carry trade itself — the explicitly leveraged, hedge-fund-driven version — is the visible part of the story. The more consequential and less discussed risk is what happens to the $5 trillion or more in overseas assets held by Japanese institutional investors who have been exporting capital for thirty years because domestic returns were inadequate.8 Life insurers, pension funds, regional banks, and the Government Pension Investment Fund — the world's largest pension fund — have built enormous positions in US Treasuries, European government bonds, and global equities, largely on an unhedged or partially hedged basis, because the cost of currency hedging has historically eaten into returns that were already thin.
As Japanese domestic yields rise toward levels that offer real competition to foreign assets for the first time in a generation, the calculus on holding those foreign positions shifts. A Japanese life insurer that can earn 1.5-2% on a 10-year JGB has meaningfully less incentive to hold unhedged US Treasuries at 4-4.5% once you factor in the currency hedging cost and the yen appreciation risk. The repatriation does not have to be dramatic or rapid to matter. Even a gradual, multi-year reallocation of Japanese institutional capital back toward domestic assets would represent a persistent and substantial source of selling pressure on US Treasuries and global equities that markets are not pricing. The Federal Reserve's own analysis has acknowledged, indirectly, that the scale of Japanese foreign holdings creates a constraint on how aggressively it can ease — because dollar weakness that accompanies Fed rate cuts accelerates the economics of Japanese repatriation, creating a feedback loop between US monetary policy and Japanese capital flows that the standard macro toolkit struggles to model.
What it means for Japanese equity investors
For someone positioned in Japanese equities — as this firm is — the carry trade dynamic creates a specific and manageable set of considerations. A strengthening yen, which is the natural consequence of BOJ rate normalisation and carry unwind, is a headwind for Japan's large export-oriented companies. Toyota, Sony, Canon, and their peers earn the majority of their revenues in foreign currencies and report in yen. Every 10-yen move in USD/JPY translates directly into earnings impact. This is why Japanese equities and USD/JPY have been so tightly correlated over the past decade — the carry trade and the export earnings story are in fact two sides of the same coin.
The domestic-oriented companies — the small and mid-cap compounders with revenues derived predominantly from the Japanese market — have a very different relationship with the yen. For them, a stronger yen means cheaper imported inputs, less imported inflation, and a more stable consumer environment. It is, broadly, a positive. This is one reason why the small-cap governance story discussed in J-1 and the inflation story explored in J-5 interact constructively with a carry unwind scenario rather than against it. The companies most interesting for their governance catalysts and balance sheet quality are typically less exposed to the export sensitivity that makes large-cap Japan vulnerable to yen strength. The carry unwind is a headwind for the Nikkei as a headline index. It is not necessarily a headwind for the specific opportunity set where the real value creation in Japan is occurring.
The second act has not been written yet
The August 2024 episode closed quickly because the BOJ blinked — its Deputy Governor's dovish comments after the market rout essentially signalled that the pace of tightening would be sensitive to market stability. That communication bought time. It did not change the destination. The BOJ has continued raising rates since then, the JGB yield curve has continued steepening, and the structural incentives that built the carry trade over thirty years are being systematically dismantled by an inflation regime that has been above target for almost four years running.
The second act of the carry unwind will not look like August 5. That episode was driven by leveraged hedge fund positioning and was resolved quickly once the most acute forced selling was absorbed. The second act will be slower, more institutional, and more structural. It will look like Japanese life insurers gradually increasing domestic JGB allocations at the margin. Like regional banks finding domestic lending more attractive as loan yields recover. Like pension funds rebalancing foreign currency exposure as hedging costs and domestic yield competition shift the efficient frontier back toward home. None of these moves make headlines. All of them matter. And the cumulative effect of $5 trillion in Japanese foreign assets being redirected, even partially and gradually, toward domestic use cases is a tail risk for global markets that is nowhere near adequately priced.
For the Japan equity investor, the operating principle is straightforward. Understand the yen sensitivity of every position. Favour domestic revenue over export revenue at the company level. Recognise that the governance re-rating story in Japanese small and mid-cap is structurally decoupled from the carry trade mechanics that dominate headlines. And maintain the analytical humility to accept that the precise timing and magnitude of the carry unwind's next phase cannot be modelled from publicly available data — because the most important part of this trade was never visible in the first place.