USD/JPY Breaks 155 for the Third Time: The Twilight of Dollar Carry Trades?

Original article from Stephen Innes

Compiled by | Speed Delivery Guy@Odaily

The third time is often the charm, and in FX, it happens more often than people think.

• A third test is not necessarily more likely to break simply because it is the third. What makes the level matter is that it was successfully defended twice before. What changes is what accumulates around it: more traders recognize the line, more positions are built against it, more stops cluster behind it, and more breakout traders wait on the other side.
• This is not a statistical law, but it has stayed with me since my early days trading USD/JPY at a Japanese bank. The chief trader there was so superstitious about round numbers and repeated tests that I nicknamed him the “Tokyo Round Number Oracle.” He believed that the third decent test was often the one that mattered, and that once a major USD/JPY level finally gave way, the market rarely looked back until the underlying mechanism itself began to lose momentum. Dark Side of the Boom.

USD/JPY finally broke below 155 on September 7, after twice holding successfully near the same area following the Golden Week intervention and again after the late-July intervention. By early trading on September 8, the pair had fallen below 154.50, and once the market cleared around 155.50—the level that marked the lows after the two intervention episodes—another yen level disappeared quickly.

That is exactly why this break matters more than the surface move suggests.

Markets remember levels, especially levels that have been defended repeatedly. The first test can be dismissed as noise, the second starts to build conviction, and by the third, the market has often accumulated enough positioning around the idea that the floor will hold again. When it finally breaks, the move can accelerate because traders are not just reacting to new information; they are also unwinding the confidence built around that level.

That appears to be what happened here.

It is still hard to pinpoint the immediate catalyst for the latest sharp decline, but the macro narrative behind the yen has clearly shifted since the middle of last week, and several forces are now pulling in the same direction.

The first is Washington.

US Treasury Secretary Scott Bessent has become increasingly assertive in his comments on Japan’s fiscal and monetary policy. Around the G20, he argued that Japan should move away from its reflation stance and said he believed the Japanese government and the BOJ would take steps that ultimately push the yen stronger.

The timing is striking, because Japanese ministries had just submitted total FY27 budget requests of about 143 trillion yen, far above the roughly 122 trillion yen initial budget for this fiscal year, reinforcing the impression that Japan’s fiscal backdrop remains highly expansionary. Bessent’s comments may simply have coincided with those data releases, but in markets, timing often matters as much as intent.

The message overseas investors heard was fairly direct: Washington wants a stronger yen, and Tokyo may have less room to ignore that preference than it did in the past.

That perception was reinforced by reports that Bessent had already expressed dissatisfaction with Japan’s economic policy during a May visit, and by the widespread view that the late-July coordinated intervention was carried out with US cooperation. Whether every detail of that story is accurate is almost secondary; what matters is that it gives global investors a political framework for expecting Japan to shift away from the reflation policy mix that has long helped keep the yen weak.

The second factor is the BOJ itself.

Bessent met Governor Kazuo Ueda on the sidelines of the G20, and the US Treasury subsequently stressed the importance of monetary policy communication, inflation expectations, and avoiding excessive exchange-rate volatility. Ueda then said rate hikes would be fully discussed at every meeting, including the next one, keeping the September 17-18 meeting firmly in play for a possible hike.

BOJ board member Hajime Takata pushed that shift further, arguing the central bank should be ready to raise rates flexibly rather than be bound by the pace the market has already priced in. He later downplayed the possibility of a large move at the next meeting, but by then the market had already absorbed the most important part of the message: the BOJ may be willing to act faster than investors had assumed.

That matters because the bullish USD/JPY trade rested for much of the summer on a very comfortable foundation: US rates high, Japanese rates low, carry trades profitable, and yen rebounds hard to sustain.

Now that policy gap may be narrowing from both ends.

The third leg of the story is less certain, but potentially far more powerful.

Speculation has resurfaced around possible changes to the Government Pension Investment Fund’s asset allocation. GPIF manages about 300 trillion yen, which means even a modest shift toward domestic financial assets could have a significant impact on Japanese markets and the yen.

The issue first surfaced in July, when Finance Minister Satsuki Katayama said the government wanted to explore ways to encourage GPIF and other pension funds to increase investment in Japanese financial assets. Market interest rose again after the GPIF board met on August 21, and a later agenda showed discussion of the Basic Portfolio Review Project Team.

Reports that this was the first board meeting held in August in about seven years only give the market more room for speculation.

For now, no one knows whether a meaningful allocation change will actually happen. Details of the discussion may not be released for months. But markets do not always wait for certainty, especially when the institution in question manages 300 trillion yen.

The mere possibility of capital repatriation to Japan is enough to have an impact.

And it comes as the dollar side of USD/JPY is starting to look less convincing.

The August jobs report was strong, with nonfarm payrolls rising by 162,000, enough to restore some probability expectations for another Fed hike. Yet the dollar’s reaction was surprisingly muted, which is itself a useful signal. Such a strong NFP print would normally be expected to drive a stronger dollar move, especially with the market already discussing a September hike.

Instead, the dollar struggled to gain momentum.

Part of the reason is that Fed officials including Christopher Waller have made clear they want to see the September 11 CPI before making a final judgment. Wage growth also slowed to 3.1% year over year, extending a gradual downward trend and reducing the urgency of the argument that the labor market is creating a new round of inflation pressure.

So the NFP strengthened the case for a hike, but it did not settle the matter.

CPI still holds the decisive vote.

President Trump has also been pressing hard in the opposite direction, calling for lower rates and threatening illogical Trump-style moves if the Fed refuses to cut. With current data, a cut at next week’s meeting would be very hard to justify, but the political message is clear enough: the White House does not want another round of tightening.

That helped cap the dollar, while Japan-specific factors began to favor the yen, which is why this move feels different from previous intervention-driven rebounds.

Pressure is now coming from both ends of the pair: Japan is turning more hawkish, or at least being perceived that way, while the dollar is getting less support from one of its strongest prior arguments.

Technically, the break below 155 is significant because USD/JPY also fell through the 38.2% retracement of the rally from the April 2025 low above 139.50 to the July 2026 high just below 164. That puts the January low below 152.50 and the 50% retracement area above 151.50 into view.

If USD/JPY long positions built around the old carry mechanism continue to unwind, the pair has room to fall further.

Unless USD/JPY can quickly recover back above 155, the market may begin to treat old support as new resistance. That would be a meaningful shift, but it still would not automatically amount to a full trend reversal.

Much of what is driving the recent yen move rests on expectations that have not yet been fully tested: a faster BOJ, less reflation bias in Japanese policy, possible GPIF repatriation, a Washington preference for a stronger yen, and no substantive shift toward more hawkishness from the Fed.

So the third attempt finally broke 155, and that deserves attention.

But the bigger question is whether the market merely kicked open a stubborn technical door, or whether Japan is truly starting to change the policy architecture on the other side. In FX, those are two very different trades.