Why the Japanese Yen Is Surging: Intervention, BOJ Rate Hikes and the Carry Trade Explained

Something worth understanding is happening in the Japanese yen.

USD/JPY has fallen from roughly 160 in early September to around 153, and the yen has strengthened against a much wider group of currencies than the dollar alone. EUR/JPY has moved lower, Japanese bond yields are climbing, and markets are pricing another Bank of Japan rate increase. Traders who spent years being paid to short the yen are now asking a very different question: what happens when the funding currency fights back?

That is the real story, and it did not begin with a chart pattern. It has been building through several separate mechanisms that have started reinforcing each other.

First, what Japanese FX intervention actually is

When the yen becomes excessively weak, Japan can step into the currency market and buy it. Somebody has to buy that yen with something, and in a yen support operation the Ministry of Finance uses foreign currency reserves, usually dollars, to do it. That creates a large and deliberately artificial source of yen demand.

The division of labour matters more than most coverage admits. The Ministry of Finance decides whether to intervene and the Bank of Japan executes the transaction as the government agent, funded from the Foreign Exchange Fund Special Account. The Bank of Japan does not wake up one morning and decide to intervene on its own, which is why headlines saying “the BOJ intervened” describe the mechanic rather than the decision.

Why Japan intervenes at all

Japan does not say it wants USD/JPY at a particular number. The stated objective is to address excessive volatility and disorderly moves rather than to defend a published level, which is consistent with the United States and Japan framework that intervention is for disorderly conditions and not for competitive advantage.

A weak yen becomes a genuine problem because Japan imports enormous quantities of energy, food, raw materials and industrial inputs. When the currency falls, all of that becomes more expensive in yen terms, which feeds domestic inflation and squeezes households. At that point the exchange rate stops being a market question and becomes a political one.

The 2026 intervention campaign was enormous

Japan spent serious money defending the currency well before this month. Between 28 April and 27 May the Ministry of Finance reported ¥11.7349 trillion of intervention, a record at the time.

Then came the summer operation. For the period from 30 July to 26 August Japan reported another ¥15.3993 trillion, the largest monthly yen buying total ever recorded, alongside the largest monthly fall in foreign reserves. Cumulative intervention in 2026 has now passed ¥27 trillion, against a previous annual record of roughly ¥15 trillion set in 2024.

That changes behaviour even on days when Tokyo does nothing. Markets now know the authorities will commit size, so every fresh yen short carries a risk that did not exist before: what if they come back? The threat itself has value, and it is considerably cheaper than spending.

Then the United States joined the trade

This was the genuinely unusual part. On 31 July Japan said it had bought yen in coordination with the United States Treasury, warned it would not hesitate to act again, and the two governments confirmed the operation jointly on 3 August. Central bank data put the joint operation at roughly 36.6 billion dollars.

Reuters then reported a detail that tells you a lot about Washington motives. Rather than selling dollars to buy yen, the United States side sold euros. Dumping dollars would have weakened the dollar and imported an inflation problem into the United States, so Treasury found another route into the same trade.

There is a second motive most coverage missed, and the desk has written about it before. The Federal Reserve FIMA repo facility lets foreign central banks pledge Treasuries for dollars instead of selling them, so Japan can fund intervention without dumping United States paper into a long end already under strain. Supporting the yen was also, conveniently, a way of protecting the Treasury market.

That is why EUR/JPY belongs in this analysis. Official yen demand can arrive through more than one cross, so this stopped being a USD/JPY story on 31 July.

Intervention started this. The Bank of Japan is what makes it stick

This is the most important distinction in the whole piece. Intervention changes flows very quickly, it can flush a crowded position and it can rewrite psychology in an afternoon. What it does not do is repeal economics.

If Japan buys yen while United States rates stay far above Japanese rates, the Bank of Japan stays dovish and the carry trade stays profitable, traders eventually have every reason to rebuild the short. That is why interventions so often produce a violent move and then fade. For a stronger yen to last, the incentive to be short has to change too.

That is what is different now. Japan built the perfect funding currency over two decades: borrow yen at almost nothing, convert it, buy something yielding far more, keep the difference. When the Federal Reserve was raising aggressively while Japan sat below zero, that gap became enormous, which is a large part of why the currency got so weak.

The direction has changed. The policy rate is at 1.00 per cent and the September meeting is expected to take it to 1.25 per cent, the highest in roughly 31 years. Kyodo reported on 8 September that the Bank has settled on the plan ahead of the meeting.

You might reasonably say 1.25 per cent is still low, and it is. But markets do not trade levels, they trade change. The question is not whether Japan is a high yielder. It is whether Japan is becoming a less attractive place to borrow than it was last quarter, and the answer is clearly yes.

Washington has said out loud that it wants a stronger yen

United States Treasury Secretary Scott Bessent has publicly backed Japan recent measures and described the yen as substantially undervalued. In his 30 August meeting with Governor Kazuo Ueda around the G20 gatherings in Asheville, he emphasised sound monetary policy and communication to anchor inflation expectations and avoid excessive exchange rate volatility, while explicitly supporting steps to address that undervaluation.

He has since gone further in both directions. On 31 August he argued that Japanese policy rather than intervention has to do the work of strengthening the yen, which is the same argument this article is making. And on 8 September at Southern Methodist University he told traders directly: “I am the house now, so when we intervene with the Japanese yen, I have pretty good insight into what the Japanese, what the Bank of Japan is going to do.”

Whatever you make of the language, the asymmetry has changed. A large yen short is no longer a bet against the Japanese government alone.

Japanese bond yields are rewriting the capital flow equation

Japan 10 year government bond yield reached 3 per cent on 1 September, the highest since 1996 and the first time this century. That matters more than a single yield print usually would, because Japanese institutions own an enormous stock of overseas assets bought for one reason: there was nothing to earn at home.

If domestic bonds start paying properly, a Japanese insurer, pension fund or bank can ask why it is taking foreign currency risk at all. That produces two yen positive effects at once. Less money leaves, and some of what already left can come home, and bringing it home means selling foreign currency and buying yen.

This is not theoretical. Bloomberg and the Japan Times both reported this week that yields near three decade highs are reviving the repatriation debate in earnest. Japanese investors sold 29.6 billion dollars of United States debt in the first quarter of 2026, and Japan remained the largest foreign holder of Treasuries in June at roughly 1.116 trillion dollars, down from about 1.143 trillion in May.

The carry trade is where this can turn violent

The yen has spent years as one of the world favourite funding currencies. Borrow it cheaply, then own higher yielding currencies, emerging market debt, credit, equities and whatever else pays more.

Reuters, citing a Jefferies analysis of Bank for International Settlements data, put cross border yen borrowing at a record ¥360 trillion, about 2.34 trillion dollars, as of March 2026. That is the largest build up in three decades. Not all of it is speculative, but it shows how much yen funding is embedded in the international system.

Here is the dangerous part. The trade works while the yen stays weak and Japanese rates stay low. When the yen strengthens sharply the borrowed yen becomes more expensive to repay at the same time as the rate advantage shrinks, so positions get closed. Closing them means selling the asset, converting back and buying yen, which strengthens the yen further, which pressures the next position.

Why 2024 still matters, and why this is not a repeat

Markets have seen this before. The Bank of Japan raised from 0.1 to 0.25 per cent on 31 July 2024, the yen appreciated roughly 13 per cent over the following days, and on 5 August the Nikkei fell 12.4 per cent in a single session, its worst day since 1987. A modest hike produced a global risk event because the positioning behind it was enormous.

The important difference now is preparation. In 2024 the move was a surprise. This time a hike is heavily anticipated, and something fully priced usually produces less shock than something nobody saw coming. The carry complex is still very large though, which is why the size of the hike is not really the question.

The question is what the Bank of Japan says about the next one.

The BOJ decision tree

Branch one, a hawkish hike

The Bank raises and signals that inflation risk remains live, keeping the door open. Japanese yields rise, the market pulls the next hike forward, and you get yen strength, lower USD/JPY and EUR/JPY, and more pressure on carry positions.

Branch two, a cautious hike

The Bank raises exactly as expected but leans on growth risks, global uncertainty or patience. The market says it already owned that, some yen longs take profit, and USD/JPY and EUR/JPY can bounce even though Japan just tightened. Hawkish action does not guarantee a hawkish reaction when the action was already priced.

Branch three, no hike

This is the real surprise, and it is less remote than consensus implies given vendors range from about 63 to 97 per cent. Without extremely hawkish guidance alongside it, recent yen longs unwind quickly, USD/JPY and EUR/JPY rise, and the carry trade starts rebuilding. The complication is that a sharp yen selloff simply invites Tokyo back into the conversation.

Why USD/JPY can fall even if the Fed stays hawkish

Traders regularly get this wrong. A hawkish Fed does not mechanically mean higher USD/JPY, because a currency pair is a relative instrument with two sides.

Suppose the Federal Reserve stays hawkish, which normally supports the dollar. If Bank of Japan tightening expectations rise faster, Japanese yields surge, carry positions unwind, United States authorities support a stronger yen and Japanese money comes home, then the Japan side can simply outweigh the United States side. USD/JPY falls anyway.

Today is a live example. The dollar index made a new low for the move while September Fed hike pricing rose to 60.4 per cent and Brent broke 100 dollars. Decomposed by index weight, dollar yen alone accounted for about 85 per cent of the dollar index move, out of a 13.6 per cent weight. This is exactly why you analyse both sides of a pair, and the same logic runs through the dollar index desk view and the USD/JPY pillar.

What EUR/JPY is actually telling you

The European Central Bank is itself tightening, with the deposit rate going to 2.50 per cent this week. So if EUR/JPY still falls while European rates rise, the yen is strong enough to overpower a relatively hawkish euro, which is a much higher bar than beating a soft dollar.

The practical desk check is to line up USD/JPY, EUR/JPY and GBP/JPY, then add MXN/JPY and TRY/JPY when you want the carry currencies to confirm. The yen gaining close to 5 per cent against the peso and the lira this month is the strongest single piece of evidence that this is a genuine yen flow rather than a dollar story in disguise. Three month implied volatility on dollar yen at a six month high says the market agrees something changed.

How to read an intervention when it happens

Official intervention often looks different from a normal fundamental move: a sudden very large move with no obvious headline, rapid acceleration through thin liquidity, several yen crosses moving at once, and stop driven continuation.

But do not label every fast yen move an intervention, because that is lazy and it is usually wrong. Ask whether there was an official statement, whether the timing was unusual, whether several yen pairs moved together, whether liquidity was thin, and whether officials had been escalating their language. Then wait for confirmation.

On the language there is a reasonably consistent ladder. It runs from watching FX moves, to watching with urgency, to moves are excessive, to all options are available, to ready to take appropriate action, then rate checks and dealer contact, and finally the operation itself. The wording varies but the escalation is recognisable, and it carries far more credibility once an authority has recently proved it will actually spend.

Has Japan intervened again this week?

As of 9 September there is no officially confirmed intervention after 26 August. That is not the same as saying none occurred, because Japan reports with a deliberate lag: monthly aggregates at month end and the daily breakdown quarterly. The next monthly disclosure, covering 27 August to 28 September, is due at the end of September.

So if somebody tells you Japan definitely intervened yesterday without official confirmation or overwhelming market evidence, that is speculation. The current rally has instead been attributed to tightening expectations, short covering, carry unwinds, repatriation, higher Japanese yields and the standing threat of intervention.

Which raises the most interesting question in this story. Does Japan still need to intervene if the market has started doing the job for it?

What would make the desk more bullish yen

• The Bank of Japan stays hawkish and further hikes remain priced
• Japanese government bond yields hold near current levels
• USD/JPY cannot reclaim its previous highs on a close
• EUR/JPY and the carry crosses keep confirming rather than diverging
• Repatriation shows up in the flow data rather than the commentary
• Positioning keeps unwinding and intervention credibility stays alive

At that point the market is no longer reacting to an operation. It is repricing Japan, which is a different and far more durable thing.

What would make the desk more cautious

A hike is heavily anticipated, which leaves obvious room for buy the rumour and sell the fact. Watch for the Bank delivering 25 basis points and then turning cautious, Japanese yields falling back, the carry crosses failing to confirm, USD/JPY refusing to make new lows, yen longs becoming crowded, or global risk sentiment improving enough that carry demand simply returns.

You can be fundamentally bullish the yen and still be run over by a countertrend move. A macro view is not a marriage, it is a decision tree.

Why this matters well beyond FX

The yen is not an isolated currency story, it sits underneath global liquidity. If a very large amount of global capital was funded in cheap yen, a strong yen can reach United States equities, Japanese equities, emerging markets, credit, global bonds, crypto and volatility.

So the better question is not what a stronger yen means for USD/JPY. It is what was bought with all the yen that was borrowed, because if the funding trade unwinds those assets become the source of liquidity. That is the moment an FX move becomes a global risk event.

There is a bond market version of the same point. If Japanese institutions reduce overseas exposure, let foreign bonds mature without reinvesting, or actively repatriate, that removes a large and reliable buyer from United States and European bond markets at a time when neither can afford to lose one. The same mechanism runs the other way in why the dollar can fall when US bond yields rise.

The KenMacro framework for the yen

Strip the headlines out and the process is the same one the desk runs on everything else.

Policy, meaning intervention and the Bank of Japan. Then expectations, meaning how much tightening and how much further intervention is already priced. Then yields, meaning whether Japanese yields are rising relative to everyone else. Then flows, meaning whether money is staying in Japan or coming home. Then positioning, meaning whether yen shorts and carry trades are unwinding. Then FX, across USD/JPY, EUR/JPY, GBP/JPY and the carry crosses. And finally global risk, meaning what was funded with all that yen.

The bottom line

This rally did not come from one thing. It began with authorities proving they would intervene at scale, then Washington joined and the market learned that both Tokyo and Washington want a stronger yen.

Then the economics started catching up. The Bank of Japan moved toward tighter policy, Japanese yields climbed to levels not seen since 1996, the advantage of funding everything in yen narrowed, Japanese investors gained a reason to keep money at home, shorts began closing and the unwind generated more yen buying on its own.

That is how intervention evolves into fundamental repricing, and how repricing evolves into flow driven momentum. So the question from here is not whether Japan intervenes tomorrow.

Watch Bank of Japan pricing, Japanese government bond yields, USD/JPY, EUR/JPY, the carry crosses and the repatriation data. And above all, watch whether the yen stays strong when nobody appears to be holding it up. That is the point at which intervention stops being the story and a genuine regime change begins.

Frequently asked questions