Key takeaways
- Japan's Ministry of Finance reported ¥11,734.9 billion of intervention in the 28 April to 27 May 2026 period — the first yen-buying since July 2024.
- USD/JPY fell from 160.23 on 29 April to 156.66 on 30 April — a move of 3.57 yen, or 2.2%.
- By 5 June the pair was back at 160.26. The entire intervention move was retraced in five weeks.
- The Bank of Japan raised the policy rate to 1.0% on 16 June by a 7-1 vote. By 25 June USD/JPY was 161.67 — weaker than before either action.
- Rengo's final 2026 shunto tally was 5.01%, below 2025's 5.25% — a 0.24-point deceleration.
¥11.7 trillion over a month bought three and a half yen, for five weeks
The MoF's monthly foreign exchange intervention release records ¥0 of operations in the 30 March–27 April 2026 window and ¥11,734.9 billion in the 28 April–27 May window. The operation began on 30 April, after the yen weakened through 160 per dollar. It was Japan's first yen-buying intervention since July 2024, and it is ordered by the Ministry of Finance; the Bank of Japan executes it as agent.
The daily record is unambiguous about what it bought. USD/JPY closed at 160.23 on 29 April and 156.66 on 30 April — 3.57 yen, 2.2%. On 6 May it was 156.44. On 22 May, 159.20. On 5 June, 160.26. The pair was weaker than its pre-intervention level within five weeks of a campaign that spent roughly ¥11.7 trillion. That figure is the Ministry of Finance’s total for 28 April to 27 May, and it spans more than one operation: Reuters’ estimates from Bank of Japan current-account data put roughly ¥5.5 trillion of it on 30 April alone, with a second operation following in May.
Then the Bank of Japan raised rates, and the yen fell further
On 16 June 2026 the BoJ's Policy Board voted 7-1 to raise the uncollateralised overnight call rate to around 1.0%, effective 17 June — the highest since 1995. The statement cites higher crude oil prices and the risk that "underlying CPI inflation" overshoots the 2% target.
A 25 basis point hike, delivered into a currency the authorities had just spent ¥11.7 trillion defending, produced this: USD/JPY was 160.39 on 16 June, 161.37 on 18 June, and 161.67 on 25 June. The chart plots the pair on five dates — 29 April (160.23), 30 April (156.66), 5 June (160.26), 16 June (160.39), 25 June (161.67), all in yen per dollar. Intervention plus tightening moved it the wrong way.
The carry trade is not priced off Japan's policy rate. It is priced off the gap.
This is the structural point the sequence exposes. A carry position funded in yen does not close because the funding rate rises from 0.75% to 1.0%. It closes when the differential against the funding currency compresses enough to make the carry not worth the volatility.
The Fed's target range has been 3.50–3.75% since December 2025. Against a 1.0% policy rate in Japan, the differential is roughly 2.6 percentage points. A 25 basis point BoJ hike narrows that by less than a tenth. The intervention, meanwhile, changes the differential by exactly nothing — it changes the spot rate, temporarily, using reserves that are finite.
Which means the yen-funded carry trade has not been meaningfully de-risked by either action. It has been left intact, with a slightly worse funding cost and a demonstrated official price floor that the market has now tested and broken twice.
Wages decelerated, which removes the strongest argument for a faster BoJ
Rengo's final tally for the 2026 shunto, released 3 July, was an average wage increase of 5.01% — a third consecutive year above 5%, but 0.24 points below 2025's 5.25%, and equivalent to ¥16,400 a month. Unions at firms with fewer than 300 employees settled at 4.69%.
The case for a fast BoJ normalisation rested on a wage-price cycle that was accelerating. It is not accelerating. It is running above 5% and decelerating — which supports gradual normalisation and argues against the aggressive path that would actually close the rate differential.
What this implies for position sizing
For a portfolio with unhedged Japanese equity exposure, the relevant number is not the BoJ's next move but the retracement speed. A 2.2% intervention move that fully reversed in five weeks establishes a rough half-life for official action of under a month. Sizing an unhedged JPY exposure on the assumption that 160 is a defended level is sizing against a floor that has already failed once at a cost of ¥11.7 trillion.
The threshold that would change this: a compression of the US-Japan policy differential below about 2 percentage points — which requires either two more BoJ hikes or a Fed cutting cycle, and the Fed is currently holding against a supply-side inflation shock.
What would falsify this
The MoF's monthly figure covers a period, not a day. The ¥11,734.9 billion total for 28 April–27 May does not isolate the 30 April operation, and the ministry publishes daily detail only quarterly. If a material share of that total was spent in May rather than on 30 April, the cost-per-yen calculation above is wrong in magnitude, though not in direction.
The retracement reading also assumes the counterfactual — that without intervention the yen would have gone further than 160. That cannot be tested. What the data can say is narrow and firm: the pair ended June weaker than it started May, after ¥11.7 trillion and a rate hike.
The number to watch is not the exchange rate
It is the MoF's quarterly daily-detail release, which reveals whether Tokyo is still spending. Intervention that stops being reported is intervention that has been abandoned, and an abandoned floor is a very different risk object from a defended one. and .
This content is for informational purposes only and does not constitute financial advice. Always do your own research or consult a qualified financial adviser before making investment decisions.