Japan Spent a Record ¥15.4 Trillion — And the Yen Still Fell Below 160

On Friday Japan's finance ministry totted up its latest bill for defending the yen: ¥15.4 trillion, about $96bn, spent buying its own currency between 30 July and 26 August, the largest monthly intervention on record. It worked, briefly and spectacularly. The yen, which had plumbed a four-decade low of 163.99 to the dollar, shot to 157.40 on 31 July. Then it unwound. Tokyo had spent the summer treating 160 as the line it would not cross. This week the yen weakened past ¥160 to the dollar, erasing more than half of its intervention-fueled gains. The money is gone; the exchange rate is not. The question is whether that proves official firepower is a rounding error against global currency flows, or merely that Tokyo acted without policy follow-through.

The wrong lever

The arithmetic makes size look like the whole story. Dollar-yen, the world's second-most-traded currency pair, turns over roughly $1.4trn a day, and global FX about $9.6trn, the Bank for International Settlements estimated last year. Japan's reserve stock of $1.29trn is roughly what dollar-yen alone changes hands in a single day. Measure the intervention in days of flow and the record ¥15.4trn was less than one day's normal trading in the very pair it sought to steer.

But size is not the binding constraint, and the failure is better read as a lesson about levers. Intervention moves the yen's level, never its trend. The trend is set by the expected path of the interest gap between Japan and America, which is decided by central banks, not by a finance ministry. The July operation worked because it was a signal: Washington joined Tokyo for the first time since 1998, and the market briefly repriced the odds that policy would change. Signals decay when the people who set policy speak. Federal Reserve Chairman Kevin Warsh had warned that anyone expecting the Fed to go easy on inflation would be disappointed; his renewed pledge to hit the target revived rate-hike bets and pushed the dollar up, eroding in days what four weeks and ¥15.4trn had staged. The operation did not fail for lack of ammunition; it failed because its premise—that the aggressive Fed would stay quiet—did not survive contact with the chairman.

The September test

The defender's retort has force. Intervention may be the wrong instrument, but it was never meant to be the only one; the Bank of Japan holds the real lever. Fitch Ratings said in August that further yen appreciation is likely to require BOJ rate hikes. The test arrives on 18 September. The BOJ took its policy rate to 1.0% in June, its highest since 1995, and markets price a near-80% chance of another increase this month. If the hike—or the promise of more—firms the yen durably, then the ¥15.4trn was spent without policy back-up: importers get relief and the exporters' tailwind thins. If the yen rallies for a week and slides again, as it did after the interventions of 2022, 2024 and this spring, the adjustment is structural, and no size of official buying will stand against the differential.

Watch the political arithmetic behind the central bank. An 80%-priced hike is a low bar for moving a currency, and every increase raises the government's debt-service bill just as JGB yields are spiking on fears about Takaichi's budget. The same government is cutting a food tax worth an estimated ¥4.4trn a year in revenue. The voter who would welcome a stronger yen is not the one who elected the fiscal programme. That is why the BOJ marches in small steps, and why the yen carries a fiscal premium as well as a rate premium: investors are selling a currency whose government wants growth and must pay dearly for it.

Whose margins

The transmission is already visible in earnings guidance, which is where an investor meets it. A weak yen does not hit "Japan" evenly; it is a transfer within Japan from those who buy abroad to those who sell there. Toyota is the cleanest exhibit. When the carmaker raised its profit outlook—operating income of ¥3.4trn against ¥3.0trn before—management credited changes in foreign-exchange assumptions, namely a yen that averages 160 for the fiscal year. That is not a forecast so much as a verdict that intervention will not hold. Carmakers built ¥150–160 into their plans, and the median Topix company's assumption has drifted from ¥150 in May to ¥154 now. Exporters are hedging against a stronger yen and locking in exchange rates for longer periods; importers are stretching supplier contracts too. Neither is the behaviour of firms expecting the yen problem to be fixed.

On the other side of the transfer the costs concentrate. Nitori, the country's largest furniture chain, figures each one-yen rise in dollar-yen costs it about ¥2bn of profit and is weighing hedges of its own. Wholesalers, thin-margin middlemen, are the first to break: weak-yen-related bankruptcies reached 45 in the first half of 2026, the most since 2022. Households absorb the same squeeze through import prices—producer prices rose 7.1% year on year in June and the yen-based import index was up nearly 30%. This is not a competitiveness story for the economy as a whole; it is a cost story for everyone without foreign revenue.

For investors

For a dollar-based investor, the yen is two variables wearing one name. It is a business variable, because it changes who earns and who bleeds, and a return variable, because a weaker yen trims the dollar value of yen-denominated assets held unhedged. Separating them is the first move. An unhedged fund such as the iShares MSCI Japan ETF carries the currency; the hedged variant strips it out. Choosing between them is a statement about 18 September. The second move is to match exposure to where the transfer lands: global earners with margin assumptions like Toyota's ¥160, against domestic, import-reliant businesses and their suppliers, whose smallest members are already defaulting.

The intervention changes the pacing, not the direction. What would genuinely alter the picture is a BOJ move that firms the yen through the mid-150s, or a further coordinated round large enough to bend the differential—Goldman Sachs reckons Tokyo retains firepower for one or two more rounds at July's scale against a pool of about $200bn in cash. Against $1.4trn a day in the pair, that buys time, not control. Each line in the sand becomes a target, and markets have a long memory for the level where a central bank blinked first. The wiser reading of ¥15.4trn is not that Japan could not afford to win, but that it was never bidding for the thing that decides the price.