The yen is not sliding because the Bank of Japan moved too little. It is sliding because Japanese capital has decided it should not come home, and no sequence of incremental adjustments to the money market guideline changes that decision.

Look at the crosses before reaching for the rate differential. EUR/JPY at 180.22 and GBP/JPY at 210.142 both rose over five days, so the yen lost ground against the euro and sterling too, not just against a dollar that gained on both. If this were purely a function of US yields, the yen would be weak against the dollar and stable elsewhere. It is weak against everything. That is a Japan problem, and it is a stock problem rather than a flow one: decades of yield suppression trained a generation of domestic institutions to hold their duration abroad, and the accumulated position does not reverse on an announcement about the terms of the complementary deposit facility.

Which makes the intervention question less interesting than it looks. Tokyo can buy yen. It has done so before, and absent a credible path for domestic real yields it buys time measured in weeks. The alternative is to let the long end of the JGB curve find its own level, and that is precisely what cannot be permitted, because the banks and insurers holding the paper were capitalised on the assumption it never happens. Japan has two release valves and can only open one. It has chosen the currency. This week's move ratifies that choice rather than reversing it, and reading it as capitulation gets the direction of causation backwards.

The more instructive number is on the other side of the Pacific. Fed funds effective sits at 3.88% while the two-year yields 4.67% and the thirty-year 5.29%, with ten-year breakevens unchanged at 2.33%. A long end that rich against flat inflation expectations is not an inflation forecast, it is a term premium: the market charging for the debt stock. Meanwhile the Fed's balance sheet grew again, to $6,746,548m. Quantity loosening under a price-tightening posture is the cleanest illustration available of where the actual bias lives.

Gold at 4,378.385 and silver up 2.72% over five days are the same trade expressed by people who do not trust breakevens as a measure of anything. Copper up 2.48% belongs to a different story about supply, though it rhymes.

My position: USD/JPY goes higher from here, intervention or not, and the second leg is not a currency event. It is the day JGB long yields break higher while the Bank of Japan is still buying, which will be described at the time as a liquidity dislocation and will in fact be the bond market pricing Japanese solvency without official assistance. That is the trade. It is not this month's trade.

The yen is not being sold for what the Bank of Japan did on 18 September. It is being sold for what the BoJ declined to say about the meeting after this one.

Look at the day itself. The US two-year fell to 4.67 and the ten-year to 4.94, both down seven basis points. On a day when the American curve richens, a pure interest-differential story gives the yen a bid. Instead USD/JPY rose 0.60% to 156.892. And the yen lost ground to everything else too: EUR/JPY up 0.68% to 180.22, GBP/JPY up 0.86% to 210.142. This is not the dollar doing something. This is the market taking a position against Japan specifically, and it did so on the day of the tightening, not before it.

That is what a one-off looks like when it arrives without a path attached. A guideline change with no stated destination is not policy, it is a data point, and the data point markets extracted was about how far the BoJ is willing to go rather than where it is going. They tested it within twenty-four hours and got an answer.

Two point one nine percent over five days, through a policy move, is a currency under genuine pressure and it needs to be treated as one. Silver is at 66.2555 and copper at 6.6001, both up on the week, gold at 4378.385. Japan imports those. Brent at 104.195 and WTI at 99.886 are the only line on the import bill moving the right way, and both are down less than two percent over five days. Convert that basket at 156.892 and the burden lands on Japanese households in food and utilities, while exporters book the translation gain and the inbound tourism economy collects. That distribution is a policy choice whether or not anyone in Tokyo describes it as one.

Intervention will not fix it. Selling reserves against a flow that originates in the pricing of the American curve is buying hours. Fed funds effective sits at 3.88 after a 0.250 move higher and the two-year is 4.67, which means the US market expects more, and Tokyo cannot out-announce that with a single line change in an operations guideline.

What would work is stating the sequence. Not defending 156.892, not a level at all, but a rate path with dates on it, accepted publicly as the cost of stopping the import-price channel. The BoJ's caution has been earned by thirty years of demand shortfall and I have defended it repeatedly. It is now producing the wrong outcome, and the bill is being sent to the wrong people.

If October brings another unaccompanied step, the yen goes lower from here and no amount of reserve selling holds it.

The BoJ move is not why the yen is falling, and that is the part the intervention chatter keeps skipping.

Look at what the US curve did on the same session. Two-year yield 4.67%, down 7bp. Ten-year 4.94%, down 7bp. Thirty-year 5.29%, down 6bp. The differential moved *toward* the yen by every tenor that matters, and USD/JPY still printed 156.892, up 0.60%. That is the whole story in one line. If this were a rate-gap trade, the yen catches a bid on a 7bp US rally. It did not.

Then check the crosses. EUR/JPY 180.22, up 0.68%. GBP/JPY 210.142, up 0.86%. Against a dollar that was itself slightly softer on the day, EUR/USD +0.09%, GBP/USD +0.27%. The yen underperformed everything, and it underperformed the crosses harder than it underperformed the dollar. This is not a dollar-strength story dressed up in yen clothing. The selling is yen-specific and it is not coming from anyone watching the two-year spread.

Which kills the capitulation read. Capitulation on a policy announcement is a fast move followed by a fast retrace, and you see it in vol and in funding. High yield OAS sat at 2.70, unchanged. Nothing in credit. Nothing in basis. A 2.19% five-day slide is not a disorderly move, it is a grind, and grinds are inventory, not panic.

So the pain trade is not the one being described. The crowded position is short USD/JPY ahead of the MoF, put on by people who were right about the BoJ and are losing money anyway. Intervention against a rising trend with no differential support buys days. The level is not the trigger. Speed is, and 2.19% over a week is not speed.

Where the bid actually is: gold 4378.385, up 0.84%, silver 66.2555, up 1.58%. Ten-year breakevens at 2.33, unchanged. Metals rallying with inflation expectations pinned is not an inflation trade, it is a currency trade, and the marginal buyer is not hedging US CPI.

One thing does not fit the frame. Fed funds effective at 3.88% after a 25bp step, and the entire Treasury curve rallied 6 to 7bp into it, with 10s2s at 25bp, two flatter, and credit spreads unmoved. Two-year at 4.67% against 3.88% effective is 79bp of tightening still priced. A policy rate moving up while duration rallies and credit shrugs is not a configuration my mapping produces. I am not going to explain it.

The position I will take: sell rallies in yen into the intervention headline, because the flow selling yen is not the flow the MoF can scare.