fell to around 153.50 during the early European session on Tuesday and extended lower to 152.89 as the day progressed, putting the Japanese yen at a seven-month high against the dollar. The pair has now declined more than 4% in September alone.
That is a violent move by the standards of a major currency pair, and it has happened in six trading sessions.
The driver is not subtle. Money-market pricing now assigns 97% odds to a Bank of Japan rate increase of 25 basis points to 1.25% at its September 17–18 meeting, up from 52% a month ago. A 45-point swing in probability inside four weeks is what a repricing looks like, and the currency has responded accordingly.
The technical damage is already done. On the daily chart, USD/JPY holds well below both the Bollinger Band midline and the 100-day simple moving average, with a clearly bearish near-term bias. The pair has fallen through the floor of the 155-to-165 range that contained it for months, and 155 had been identified as more than a conventional support level.
The context makes the move more remarkable rather than less. The yen hit 40-year lows in July, weak enough to trigger a joint US-Japan intervention. Seven weeks later it is at a seven-month high. That is a complete round trip in a currency whose defining characteristic for the past four years has been one-directional weakness.
The broader dollar has not moved nearly as far. closed Monday down just 0.25% at 98.91 and has been oscillating near 99.0 to 99.2. EUR/USD sits at 1.1611 and GBP/USD at 1.3544, both essentially unchanged on the week.
So this is a yen story, not a dollar story. Something specific to Japan is repricing, and the rest of the G10 complex is watching rather than participating.
The two central bank meetings that decide the next leg land within 48 hours of each other next week.
Four Percent in Six Sessions and the Floor That Failed
The speed of the decline is the detail that matters most for risk management.
USD/JPY extended its September fall beyond 4%, and the speed of the move suggests stop-loss orders and the unwinding of leveraged yen shorts have reinforced the direction. That is a mechanical accelerant rather than a fundamental one — each level broken triggers the next tranche of forced buying in yen.
The 155 level was the structural line. It sat at the bottom of the central trading range and functioned as more than ordinary horizontal support, marking the boundary between a market that trusted the carry trade and one that did not.
Losing it changed the character of the pair. Before the break, dips were bought on the logic that Japan’s rate advantage was structurally negative and that carry flows would reassert. After the break, the same dips became opportunities to add to short dollar exposure.
The performance math on positioning illustrates how fast this has moved. A short recommendation entered at 159.70 targeting 149.0 with a stop at 164.0 was published less than a week before the pair printed 152.89. That entry-to-current move represents roughly 4.3% before carry and trading costs.
Carry is the important qualifier. Holding a short USD/JPY position means paying the interest differential every day — currently around 262.5 basis points annualised, or roughly 0.7% per quarter. That is the cost of being early, and it is what has punished yen bulls repeatedly since 2022.
For the first time in years, the price move is compensating for that cost within days rather than months.
Volatility has expanded accordingly. A pair that spent much of 2026 in tight ranges has produced multiple sessions of one-percent-plus moves in a single week, which changes position sizing for anyone trading it.
The immediate technical question is whether 152.89 holds as a short-term low or whether the momentum carries through toward the next structural reference. Nothing in the current momentum picture suggests exhaustion.
From a Forty-Year Low in July to a Seven-Month High
The round trip deserves examination because it explains why positioning is so lopsided.
The yen hit 40-year lows in July. That weakness was severe enough to prompt a joint US-Japan intervention — coordinated official action, which is rare and signals that both governments considered the level disorderly rather than merely uncomfortable.
USD/JPY reached 161 in July 2024, a level not seen since 1986. The 2026 lows in the yen exceeded that.
Intervention on its own rarely reverses a trend. What it does is buy time for fundamentals to change, and in this case they have. Between July and September, three things shifted: BOJ hike expectations moved from a coin flip to near-certainty, early signs of capital repatriation appeared, and US pressure on Japan over the currency added a political dimension.
That combination is what has taken the yen from a four-decade low to a seven-month high in under two months.
The structural backdrop supports the move on Japan’s side. Wage growth above 3% and core CPI holding above 2% give the Bank of Japan room to hike, which is the precondition the market has waited on for two years. Inflation running above the 2% target has been cited by the Bank itself as the reason further hikes are likely.
The policy path has been slow but persistent. The Bank raised the policy rate to 1.00% on 16 June 2026, effective 17 June, after holding at 0.75% since December 2025. Yield curve control ended in March 2024, with rates moving from −0.1% to 0.25% by July of that year, then 0.50% in January 2025.
Roughly one hike per year from 2024, accelerating to two moves in the twelve months to December 2025, and a third in June 2026. Each move has carried outsized impact because the starting point is so low.
The Bank’s calendar and statements are published at boj.or.jp.
September 18: Ninety-Seven Percent Odds, Up From Fifty-Two
The single most important number in this forecast is the probability shift.
Money-market pricing puts the odds at 97% that the Bank of Japan raises its key rate by 25 basis points to 1.25% at the meeting concluding September 18. A month ago that figure stood at 52%.
Beyond September, the same pricing shows a 27% chance of an increase in October and 61% odds in December. Those follow-on probabilities matter more than the September number, because a fully-priced hike produces no move when it lands — the reaction comes from what the Bank signals about the path afterward.
The meeting runs 17–18 September. The policy statement typically arrives around midday Tokyo time, with the Governor’s press conference at 3:30 p.m. JST. The Bank does not publish its statement at a fixed time, which makes its decision days among the most volatile events on the global calendar.
The sequencing next week is tight and consequential. The Federal Reserve decides on 16 September with updated projections. The Bank of Japan follows on 18 September. Two days apart, both with live hike probabilities, both directly relevant to the same currency pair.
At 97%, the September move is done as far as the market is concerned. The trade is entirely in Governor Kazuo Ueda’s communication and in whether the statement leaves October and December genuinely open.
The Bank’s historical communication style complicates that. It is known for vagueness, and if the Governor stays vague the market reaction is typically small. If he sounds surprisingly direct about future increases, it confirms the aggressive expectations already building and produces a large, sudden move.
That asymmetry favours the yen at the margin, because 97% pricing on September means the surprise capacity sits almost entirely on the hawkish side of the guidance rather than on the decision itself.
Fed detail is published at federalreserve.gov.
The Vote Math: Eight-to-One in July, Seven-to-One in June
Board composition tells you how close the next move is, and the recent record is informative.
The June 2026 hike passed by a 7–1 vote, with Asada Toichiro preferring to hold. At the July meeting, the Board held at 1.00% by an 8–1 vote that rejected a proposal to go further, to 1.25%.
Read those two tallies together. In June, one member wanted to stay put and lost. In July, one member wanted to go to 1.25% and lost. The centre of gravity sat firmly at 1.00% for two consecutive meetings.
Earlier in the year the dissents ran the other way, with Takata Hajime and later two colleagues wanting to raise sooner. The growing number of dissents in favour of higher rates through early 2026 indicates the Board’s centre has been shifting toward tightening for some time.
That shift is what makes the 97% September probability credible rather than speculative. A Board that already had a member voting for 1.25% in July needs only a modest change in the inflation or wage picture to deliver it in September.
The vote split at the September meeting will therefore carry information beyond the decision. A 25-basis-point hike passing 8–1 with a dissent arguing for a hold would signal that 1.25% is a terminal level for the near term. The same hike passing 7–2 with two members preferring 1.50% would signal an accelerating path and would take the yen materially higher.
Statements name every dissenter and give the vote tally, so the split is available immediately rather than in the minutes weeks later.
The Board meets eight times a year, normally in January, March, April, June, July, September, October and December, with the Outlook Report — the Bank’s quarterly projections — released at four of them. Those Outlook meetings typically produce the largest market reactions because they carry updated inflation and growth forecasts alongside the decision.
The political layer has shifted as well. Prime Minister Sanae Takaichi has been characterised as dovish on monetary policy, but reporting has indicated government support for an earlier rate increase — a reversal that removes one of the constraints the Bank has historically operated under.
Takata’s Comment and What “Back-to-Back” Would Mean
One board member’s remarks last week are doing more work in this move than the probability numbers suggest.
Board member Hajime Takata said the central bank could take a more aggressive approach than expected. He stated that a 25-basis-point hike “is not necessarily set in stone,” and that back-to-back rate hikes would generally be a possibility.
Two phrases in that. The first leaves open a larger single move — 50 basis points rather than 25, which is not currently priced anywhere. The second opens the door to consecutive meetings, which would mean September and October rather than September and then a pause.
Neither scenario is in the 97% number, which prices a single 25-basis-point step. The October probability sits at 27% and December at 61%, so a back-to-back path is roughly a quarter priced.
If the September statement or the press conference validates the back-to-back framing, the repricing in October probability alone would be worth several yen. Moving October from 27% toward 70% would carry USD/JPY through the next support cluster without requiring any change in the Fed’s path.
That is the specific mechanism through which this could extend, and it is why the pair has continued falling even as the September hike moved to near-certainty. The market is no longer trading the September decision. It is trading the path implied by it.
Markets price Bank of Japan moves differently from Fed moves for structural reasons. Decades of dovishness mean traders need convincing that each hike is not the last, which historically has capped the yen’s reaction to individual increases. Takata’s comment attacks exactly that scepticism.
The counterweight is that individual board members speak for themselves. A single hawkish voice on a nine-member board is not policy, and the July vote showed the Board rejecting a move to 1.25% by 8–1 only weeks earlier.
The gap between what one member says and what the Board does is the risk in the current positioning.
The Differential Trap: Both Hike and Nothing Changes
Here is the arithmetic the momentum may be under-weighting.
The Federal Reserve’s target range is 3.50% to 3.75%, a midpoint of 3.625%. The Bank of Japan’s policy rate sits at 1.00%. The differential is 262.5 basis points.
If the Bank of Japan raises to 1.25% on 18 September and the Federal Reserve holds on 16 September, the gap narrows to 237.5 basis points — a 25-basis-point improvement for the yen.
If both hike, the differential returns to exactly 262.5 basis points. Nothing changes.
Fed funds futures price a September increase at roughly 58% to 60% following August payrolls of 162,000 against a consensus near 56,000, with unemployment steady at 4.1%. Employment detail is at bls.gov.
Multiply those probabilities and the modal outcome next week — both central banks hiking — leaves the rate differential precisely where it started. That is not a fundamentals-driven case for a lower USD/JPY.
The broader compression trend is real but slower than the price action implies. The differential has narrowed from roughly 325 basis points in early 2026 toward 250 to 275 basis points by the fourth quarter. Every 100 basis points of compression has historically correlated with a five-to-eight-yen move.
Apply that relationship. Compression of 62.5 basis points from early 2026 to now should have produced roughly three to five yen of USD/JPY decline. The pair has fallen from the July highs by considerably more than that.
The excess is positioning, not arithmetic. That is precisely why the stated path lower is framed around the size of the outstanding short-yen position rather than around a rate-differential target — a further unwind could push USD/JPY toward the mid-140s given the sizeable outstanding short position.
Positioning-driven moves are faster and less durable than differential-driven moves. They also reverse harder when they exhaust.
Original Post