Japan may be quietly dismantling one of the world’s biggest financial trades

For many years Japan had a peculiarity that suited the rest of the world rather well. Its money cost practically nothing. The Bank of Japan kept rates at zero, when not actually below it, bought government debt, intervened on the yield curve and did everything possible so that a country obsessed with deflation for decades could manage to generate some inflation. The result reached far beyond its borders. The yen ended up becoming one of the great funding currencies of the international financial system.

Two Fridays ago the Bank of Japan raised rates again, this time to 1.25%, the highest level in 31 years. It may seem like very little. After seeing rates of 4%, 5% and even higher in other economies, 1.25% still almost looks like cheap money. The problem is that we are looking at the snapshot when we should be watching the film. Barely two years ago Japan had a negative policy rate of -0.1%. Then came 0.25%, 0.50%, 0.75%, 1% and now 1.25%. The Bank of Japan itself continues to leave the door open to further hikes if inflation and the economy evolve as it expects.

I do not think what really matters is how much it has raised rates now. The interesting thing is to ask what happens when one of the world’s last great suppliers of cheap money decides it no longer wants to be one.

For decades there was an extraordinarily simple trade. Borrow in yen paying practically nothing, convert those yen into dollars, Mexican pesos, Brazilian reais or any other currency offering a higher return, invest the money and pocket the difference. It is what we know as the carry trade. As long as the yen did not appreciate too much, it could be a magnificent business.

Japan does not need to cause a crisis to change the global market. It is enough for it to stop exporting savings at the pace it has for decades.

Suppose someone borrowed in yen at 0.5% and bought an asset yielding 5%. They earned 4.5 points before costs. With leverage, the return on their own capital could be multiplied. The problem was hidden in an apparently small condition. At the end of the trade, the yen had to be bought back to repay the loan. If during that time the yen appreciated by 8%, the return earned over months could vanish in a matter of days. And if the trade was also leveraged, a manageable loss could quickly turn into a margin call.

This is where one of the traps in this story appears. Gigantic figures have been published on the size of the Japanese carry trade, some of several trillion dollars, mixing loans, derivatives, hedges and financial operations that have nothing to do with speculative bets. The Bank for International Settlements (BIS) tried to measure it after the scare of August 2024 and put a reasonable estimate at around 40 trillion yen, some 250 billion dollars at the time, while warning that the figure was probably underestimated because a significant part of these trades does not appear transparently in the statistics.

Two hundred and fifty billion dollars is a lot of money, although not enough on its own to explain a global financial crisis. The problem, once again, lies elsewhere. That money is leveraged and connected to many other assets. In other words, that money comes largely from bank loans and is used to invest in other assets. It may sound very technical, but it is not. It is very simple.

We already saw a small demonstration in August 2024. The Bank of Japan had begun to withdraw stimulus, the market started to price in a future narrowing of the rate differential with the United States, and the yen appreciated violently. In a few days, positions funded in the Japanese currency had to be closed. To close a short yen position you have to buy yen. The more funds bought yen to get out, the more the yen rose. The more it rose, the greater the losses for those still inside. And the greater those losses, the more positions had to be liquidated.

Japan’s cheap money is running out. And the problem may not be in Japan

No bank had to fail and no new pandemic had to appear. It was enough to combine leverage, currencies and volatility. The BIS later concluded that the unwinding of carry trades had considerably amplified the markets’ initial reaction and that the increase in required collateral had accelerated the deleveraging.

That is one part of the story. The other, probably more important in the medium term, has little to do with hedge funds, which are the ones doing what I have just described.

Japan is one of the largest exporters of capital on the planet. At the end of 2025 it held foreign assets worth some 1,805 trillion yen and had a net international investment position of close to 562 trillion. Its foreign portfolio investment alone amounted to some 769 trillion yen. The figures are so large that they end up losing meaning, but the economic reason behind them is quite simple. If for years a Japanese bond paid practically nothing, a Japanese pension fund, insurer or bank had every reason to look for returns abroad.

The carry trade does not depend only on the rate differential. It also depends on the yen, on volatility and, above all, on leverage.

Now the maths is changing.

The Japanese ten-year bond has risen above 3% for the first time since 1996. That does not look spectacular compared with a US Treasury, which is already at 5%, but for a Japanese institution there is an important difference. Buying US debt means taking on currency risk against the dollar or paying a hedging premium to eliminate it. If the domestic yield rises while the cost of hedging the foreign currency stays high, there comes a point where going abroad is no longer so worthwhile.

And here a distinction is needed, because otherwise we end up mixing different things. A Japanese insurer buying US bonds does not mean it is doing a carry trade. Nor, necessarily, is the huge Japanese public pension fund, the GPIF, which as of June managed roughly 320 trillion yen and held close to half its portfolio in foreign bonds and equities. These are structural investments, not speculative bets funded in the short term.

But both stories end up meeting in the same place. For decades Japan provided the world with a very cheap currency to fund trades and, at the same time, a gigantic amount of savings that needed to find returns abroad. What is changing now is both of those things. One detail: Japan is the largest foreign holder of US sovereign debt.

There is no need to imagine a scenario in which the Japanese sell hundreds of billions of dollars of US bonds tomorrow and take the money back to Tokyo. It probably will not happen that way. Something far less spectacular is enough. That part of the new savings stops going abroad.

For years Japan offered the world two things at once: almost free funding and an enormous amount of savings looking for returns abroad. Both are starting to change.

Financial markets live off the marginal buyer. If Japan stops adding foreign capital at the pace it used to, someone will have to take its place. And if that buyer demands a higher return, bond prices will have to fall and their yields rise.

The United States should watch this closely. In July Japan still held around 1.104 trillion dollars in US Treasuries, but in February the figure was roughly 1.239 trillion. That is some 135 billion less in five months. I would not claim we are looking at a massive repatriation, because these figures are affected by maturities, valuations, portfolio moves and other factors, but the direction at least deserves watching.

And it is happening at a particularly awkward moment. The United States needs to finance enormous volumes of debt, the ten-year Treasury has once again risen above 5%, Europe is also living with growing deficits, and Japan is starting to offer its own investors an alternative that for many years practically did not exist. It does not seem the best moment to lose marginal buyers of fixed income.

There is another paradox worth noting. When the Bank of Japan raised the policy rate to 1.25%, the yen weakened instead of appreciating. Some investors expected a more aggressive message, and two board members voted against the hike. The market understood that Kazuo Ueda will keep normalising, but cautiously. The reaction shows precisely why it would be a mistake to claim that the carry trade is dead. It is not dead. The United States keeps fed funds between 3.75% and 4%, after the hike approved last Wednesday, while Japan is at 1.25%. The differential still exists.

What is disappearing is something different: the peace of mind with which one could assume that Japan would stay close to zero forever. And that is a far more serious change than it seems.

The carry trade does not depend solely on the difference between two interest rates. It also depends on the exchange rate, on volatility and on leverage. An investor may accept earning two or three points a year if they believe the yen will stay stable. But if the Japanese currency appreciates 8% in a week, the calculation stops making much sense. With five times leverage, that move can become a gigantic loss on the capital put up.

That is why I think the specific level of the yen matters less than the speed. Going from 157 yen to the dollar to 145 over a year may be digestible. Doing it in four sessions can trigger forced selling in places that apparently have nothing to do with Japan.

So let us get to what affects the reader. Because it does affect them, and a great deal. Think of a fund that borrows in yen and buys US tech stocks, Mexican bonds or any other asset with a higher return. The yen starts to rise, the position loses money and the broker demands more collateral. The fund needs liquidity. What does it sell? Usually whatever it can sell, not necessarily what caused the problem. Suddenly selling appears on the Nasdaq, in emerging-market debt or even in crypto assets, and it is hard to understand where it comes from. The origin may lie thousands of kilometres away.

That is precisely what made August 2024 interesting. Japan did not cause a global economic crisis. It acted as a mechanism for transmitting and amplifying a move that had started elsewhere.

And it could happen again. The trigger does not even have to occur in Tokyo. It could be a US slowdown that forces the Federal Reserve to cut rates while Japan keeps raising them, narrowing the differential from both sides. It could be a yen intervention. An episode of stock market volatility could force leveraged positions to be reduced. Or there could simply come a time when Japanese investors conclude that a domestic bond at 3% is attractive enough again.

The Bank of Japan does not have much freedom either. If it keeps rates too low, the yen may keep depreciating and make the energy imports Japan needs even more expensive. If it raises them too quickly, it increases the funding cost of a state whose debt is well above 200% of GDP and may accelerate the unwinding of international positions funded in its currency. The government, meanwhile, keeps proposing expansionary fiscal measures and tax cuts to offset the rising cost of living. Monetary policy and fiscal policy are pulling in directions that are not always easy to reconcile.

If Japanese investors find enough return at home again, the problem may show up in Treasuries, the Nasdaq or emerging markets well before it does in Tokyo.

I do not think we should necessarily expect a major accident. It is perfectly possible that the Bank of Japan keeps raising rates slowly, the yen stays relatively weak and carry positions are reduced in an orderly way. In fact, the reaction after the last hike suggests that this remains a perfectly possible scenario.

But nor do I think that is the right question. The interesting question is what price some global assets would have if Japan gradually stopped financing them the way it has for thirty years.

For a long time we have grown used to money having an artificially low price somewhere in the system. First Japan, then the Federal Reserve, later the ECB and finally all of them at the same time after the great financial crisis and the pandemic. The financial world learned to build ever larger structures on that abundance of liquidity.

Japan was the survivor of that model. While other central banks raised rates, Tokyo stayed anchored near zero. While Japanese bonds offered barely any yield, its savers kept buying foreign assets. And while borrowing in yen remained cheap, the carry trade kept finding new lives every time volatility disappeared.

Now the country is doing something very different. It is not shutting the tap all at once, but it is turning it. Perhaps the market will have enough time to adapt. Perhaps in a few years we will remember this process simply as Japan’s return to monetary normality. But it is worth remembering that for Japan normality is something we have not seen for three decades. And when a country that holds more than 1,800 trillion yen in foreign assets starts to change the incentives that took that money abroad, I would not look only at Tokyo. I would also look at New York, at the bond market, at the Nasdaq, at emerging-market currencies and at all those places where for years someone was buying assets with money that, somewhere in the trade, was extraordinarily cheap.

Japan’s cheap money is starting to run out. The unknown is who will be the first to discover how much they depended on it.