Shortly before noon in Tokyo on Friday, the Bank of Japan raised its policy rate by 25 basis points to 1.25 percent, the highest level since 1995 and the sixth increase under Governor Kazuo Ueda, more than any of his predecessors managed in half a century. The move itself was no surprise. Futures had priced it for weeks, Ueda had all but announced it in early September, and the only real question going in was how hawkish the language around it would be. The answer sat in the footnote of the decision: seven to two. Toichiro Asada and Ayano Sato, the two board members Prime Minister Sanae Takaichi has sent to the nine-seat panel since taking office last October, voted against the hike. In June it had been one dissent, Asada’s. Now it is two.
The reaction came within minutes and ran the wrong way. The yen, which is supposed to strengthen when its central bank tightens, fell. It dropped 0.8 percent on the decision, kept sliding through Ueda’s press conference, and by the close the dollar was up 1.2 percent at 157.84 yen, a two-week high, the biggest one-day gain since December and the largest weekly rally since September 2024. The Nikkei rose 1.51 percent to 65,102. The 10-year Japanese government bond did not move at all, closing at 2.99 percent, one basis point lower than the day before. A central bank raises rates and its currency gets cheaper, its stocks get dearer and its bonds stay put. That is not a contradiction. It is information: on Friday the market did not trade the hike. It traded the dissent.
Three hikes in eight days, three different verdicts
Tokyo’s decision was the third rate increase by a major central bank in eight days. On September 10 the European Central Bank lifted its deposit rate to 2.50 percent. On Wednesday the Federal Reserve raised its target range to 3.75 to 4.00 percent, the first hike since July 2023, unanimously, twelve to zero. In between, the Bank of England held at 3.75 percent on Thursday, but only six to three, with three members wanting to go to 4 percent immediately. It is the first month in which the Fed, the ECB and the Bank of Japan have all raised rates together, as the Seoul Economic Daily noted on Friday. And it is a case study in what markets actually price when a central bank moves.
The Fed hiked unanimously and hit its stock market: the Dow lost 631 points on Wednesday because twelve to zero means the next hike will not fail for lack of a majority. The Bank of Japan hiked seven to two and hit its currency, because seven to two means the next hike might. At the Bank of England, markets read the three hawks in the minority as the advance party for a November move, and sterling held. In all three cases the rate step itself, long priced in, was beside the point. What moved prices was the vote count, because the vote count is where the step after next is written. Central banks do not move markets with what they do. They move them with what they reveal about what comes after. On Friday, Tokyo revealed that the government is building a veto inside the board.
What the two dissenters actually said
The Bank of Japan publishes its dissenters’ reasoning in the decision text, and Friday’s two are worth reading in full. Asada dissented because the rate of increase in the CPI excluding fresh food had been “below 2 percent recently” and “it could not necessarily be said that the economic situation was strong”; he wanted the rate left at 1.00 percent. Sato argued that current economic and price developments “did not appear to have substantially accelerated compared to before,” and that “it was not appropriate for the Bank to raise the policy interest rate at this time.” Both arguments are defensible on the numbers. Japan’s core inflation came in at 1.7 percent for August, down from 1.8 in July, the eighth straight month below the 2 percent target and a tenth below what economists expected. The figure landed on Friday morning, hours before the decision.
But that is not the number the majority is arguing from. The statement points to producer prices, up 7.6 percent from a year earlier in August on oil, AI demand and the weak yen, and notes that this pressure “has started to spill over into consumer prices.” The bank’s preferred gauge of underlying inflation, which strips out fresh food and energy, ran at 1.9 percent in August. Medium- to long-term inflation expectations are rising. And then comes the sentence that carries the hike: there is a risk that underlying inflation “will deviate upward to a level above the price stability target of 2 percent.” The majority, in other words, is not raising against inflation that has arrived but against inflation it can see coming. The minority wants to wait until it is here. That is exactly the argument the Fed settled twelve to zero on Wednesday. Tokyo settled it seven to two. The market saw the difference.
Why two votes weigh more than 25 basis points
The arithmetic is simple. The Bank of Japan’s board has nine seats. Two are now held by reflationists chosen by Takaichi, whose government prefers easy money and loose fiscal policy: a 370 trillion yen public-private investment plan, a proposed food tax cut that costs 4.4 trillion yen in revenue, defense spending headed toward 3.5 percent of GDP, all on a debt ratio of roughly 260 percent. Every board member whose term expires will be replaced by this government. The market does the math: at two of nine, the majority for hikes is comfortable. At four of nine it is not. The question is not whether the Bank of Japan hikes in October. It is how many more hikes this board can carry before the government has rebuilt it.
Naka Matsuzawa at Nomura called the yen’s slide “a knee-jerk reaction to the two dissent votes.” Ray Attrill at National Australia Bank said the bank had “just clearly underwhelmed versus expectations.” Hirofumi Suzuki at SMBC: the outcome “has somewhat tempered expectations for further rate hikes.” Daiwa called the decision “likely to be seen as dovish.” Four houses, one message: the rate went up and the path got flatter. For a currency, only the path matters. Anyone borrowing yen to buy dollars pays 1.25 percent today and collects 4 percent; that 2.75-point gap was 3.5 points at the start of the year. It is shrinking, but slowly, and on Friday the market decided it will shrink more slowly still. So the carry trade stays on. So the yen falls.
What Ueda said, and what he did not
Ueda could have turned it around. He did not. At the press conference he said underlying inflation was “now quite close to 2 percent,” that “our policy phase has changed,” that the bank was in a phase where “we need to look at various data carefully, but that doesn’t mean we can move slowly.” On larger or back-to-back hikes: “That depends on how price conditions develop. There could be various possibilities. We shouldn’t rule anything out.” On the neutral rate: “It is hard to pinpoint where the neutral rate is, and therefore the terminal rate.” And the line that finished the yen off: “We don’t guide policy to control currency moves or stabilise currency rates at a certain range.”
That is a textbook central bank press conference, and that was the problem. After two dissents the market wanted to hear that the majority had already put the next hike on the calendar. It heard that the majority was studying the data. Bloomberg’s summary: the bank “delivered both hawkish and dovish signals, but the currency market is not taking his remarks as sufficiently hawkish at this point.” The statement itself says the bank “will continue to raise the policy interest rate,” which is more forward commitment than the Fed offered on Wednesday. But it appears in a document with two names missing from the bottom. And for the first time the policy rate sits inside the range the bank itself estimates as neutral, 1.1 to 2.5 percent. From here, every further increase is no longer “normalisation” but tightening, and a board argues differently about tightening than about normalisation.
The American factor
There is a reason this hike reads differently from the five before it, and it sits in Washington. On July 30, Japan and the United States intervened jointly in currency markets for the first time since 1998 to prop up a yen that had slid to 163 per dollar, a 40-year low. Japan sold roughly 59 billion dollars; the U.S. Treasury added 5 to 10 billion, funded by selling euros rather than dollars. The yen rallied to 157. Two weeks later it was back at 159, half the gain gone. On Friday it closed at 157.84. Seven weeks and some 65 billion dollars later, the exchange rate is back where the intervention left it and moving the wrong way again.
Treasury Secretary Scott Bessent has never hidden his role. He pushed for early Bank of Japan hikes at the G20, pressed Finance Minister Satsuki Katayama on the point in May, and told Fortune on September 10: “I am the house now. I have pretty good insight into what the Bank of Japan is going to do.” Katayama remarked that the phrase sounded “a little scary” in Japanese. She is right, and for a different reason than she meant. When the U.S. Treasury Secretary publicly demands rate hikes in Tokyo and the central bank delivers, the market stops reading the hike as a judgment on Japanese inflation and starts reading it as compliance with a request. And a hike made under external pressure has a limit: the point at which domestic pressure becomes stronger. The two dissents mark that point. The yen fell on Friday because the market saw it.
Why Tokyo’s stock market rose anyway
The Nikkei gained 966 points, and anyone reading that as applause for the central bank has cause and effect backwards. Two things drove the tape, and neither was the rate. The first was the yen: a currency that loses 1.2 percent is a profit uplift for Toyota, Sony and the machinery makers, and a cost problem for Fast Retailing that the market chose to ignore for a day. The second was Nvidia. Jensen Huang had said on Thursday in Scotland, at a conference convened by King Charles, that his company would sell “twice as many chips” next year as this year. Nvidia rose 2 percent and the Tokyo supply chain followed: Lasertec up 9 percent, Advantest up 7, Mitsui Kinzoku up 6, Kioxia up 3, SoftBank up 4. That is a chip rally with a rate decision in the background, not the other way round.
For the banks, the natural winners from higher rates, the day was more mixed. Mitsubishi UFJ, Sumitomo Mitsui and Mizuho earn on every basis point the central bank adds, and they have ridden the entire road from minus 0.1 percent in March 2024 to 1.25 percent now. But their share price hangs on the path, not the level, for the same reason the yen does. Two dissents tell a Japanese bank that the terminal rate is closer than it thought, and a terminal rate of 1.5 percent instead of 2 is half a net interest margin gone. Bonds, finally, went nowhere: 2.99 percent on the 10-year, a basis point lower. The bond market had priced this hike in early September, when the yield crossed 3 percent for the first time in three decades, and on Friday it traded the same reading as the currency market. What was priced was the hike. What was not priced was the doubt about the next one.
What it means for U.S. investors
For American investors the Friday decision has three addresses, and the first is the Treasury market that was the subject here on Wednesday. Japanese investors hold on the order of 1.1 trillion dollars of Treasuries, the largest foreign position in the world, and for decades they bought because they earned nothing at home. That has changed. A 10-year JGB pays 2.99 percent. A 10-year Treasury pays 5 percent, but a Japanese life insurer hedging the currency gives up roughly the short-rate differential, about 2.75 points plus basis, and keeps something near 2.2 percent. Home now pays more than America, hedged. Every Tokyo hike tightens that comparison and pulls a marginal buyer out of the Treasury market at exactly the moment the 10-year is trying to hold 5 percent. Friday’s two dissents slow that process. They do not reverse it.
The second address is the carry trade, and the memory that goes with it. On August 5, 2024, the Bank of Japan had just raised to 0.25 percent while Fed cut expectations surged, the yen-funded carry trade unwound in days, the Nikkei lost 12.4 percent in a session and the VIX touched 65. The gap then was about 5 points. Today it is 2.75 and the Fed is hiking, not cutting, which is why the trade survived Friday intact. But the trade’s lifespan is now measured in Tokyo votes, not Washington ones: the day a third dissenter appears, the market will price the end of Japanese tightening, and a carry trade that stops shrinking is a carry trade that gets bigger. The third address is the equity exposure itself. By the time New York opens, the unhedged Japan ETF, EWJ, will have given back most of Friday’s Nikkei gain in dollar terms; the hedged version, DXJ, keeps it. Year to date, the hedge has been the single largest driver of return on Japanese equities, bigger than sector selection. Toyota, Sony and the three megabanks all trade as ADRs in New York, and all of them are, in the end, a yen position with an operating business attached.
The counter-argument
There is a reading in which Friday went exactly right, and it deserves the space. First: a central bank whose core inflation has spent eight months below target hikes for the sixth time. That is not weakness but resolve, and two dissents on a hike that runs against the headline data are evidence of a functioning board, not a captured one. Second: the yen fell partly because the dollar rose against everything on Friday; the Korean won lost for a seventh straight session. Tokyo hiked into a global dollar wave, not into yen-specific weakness. Third: Ueda’s line that policy is not run for the exchange rate is the opposite of surrender. A central bank that hikes for the currency market has lost before it starts. Fourth: the statement says explicitly that the bank “will continue to raise the policy interest rate.” The market read the text through the footnote on Friday, but the footnote does not change the text.
There is also a historical objection to the captured-board narrative. The Bank of Japan under Ueda has absorbed more open disagreement than under any recent predecessor and has still hiked six times; Hajime Takata and Naoki Tamura, the two hawks, remain on the board and have repeatedly argued for faster moves. A board with two doves and two hawks around a governor who reads the data is not a board in danger. The yen may see it differently, and on Friday it did. But the currency market has already mispriced Japan twice this year: in July, when it did not see the intervention coming, and in August, when it assumed the intervention would hold.
What to watch
Three markers. The first is 160 yen per dollar. That is where the road to intervention began in late July, and Bessent has promised to do “whatever it takes” again, with Katayama confirming. If the rate is above 160 in two weeks, Friday’s hike has failed as a currency prop, and the next intervention will cost more than the last because this time the market will see it coming. The second marker is late October, when the Bank of Japan meets with its next quarterly outlook report. If the vote is still seven to two and the bank hikes, Friday was a currency-market stumble. If the bank does not hike, Friday was a preview. The third marker is core inflation: it stands at 1.7 percent and, by the majority’s own forecast, must climb above 2 percent in the second half of the fiscal year. If it does not, Asada was right, and the government has an argument for the next board appointment.
The real lesson of the day is about central banks in general. In eight days three of them raised rates, and only one paid for it with a falling currency. It was the one whose hike was demanded from outside and disputed from inside. The Fed hiked against market expectations and against the president, unanimously, and the market believed it about the next one. The Bank of Japan hiked at the request of the U.S. Treasury Secretary and over the votes of the government’s own appointees, and the market did not believe it about the next one. The rate rose 25 basis points on Friday. The credibility that it keeps rising fell by two votes. The currency market decided which of those two numbers matters more.
Try TradingView Free for 30 Days
Plus get a $15 discount on your first subscription through this link.
