The yen rallied on the possibility of a policy change, not a policy change, and the gap between those two things is the trade. A speech in Sapporo and a notice of a meeting on market operations moved USD/JPY 1.84% in a session, with EUR/JPY and GBP/JPY down 1.52% and 1.55%. That is not information arriving. That is a crowded funding position finding out how thin the door is.

The obvious read is that Japan is finally normalizing. I do not think the price action supports it, and I think the energy tape argues against it.

Brent is up 7.61% over five days to 97.939 and WTI up 9.49% to 92.27, with US crude stocks excluding the SPR drawing 4,450 MBBL in the week to 28 August. For a country that imports essentially all of its hydrocarbons, a firmer yen against that backdrop is a gift. The currency delivers the disinflation the policy rate would otherwise have to. A central bank that has spent three decades avoiding the moment when it has to mark its own JGB book will take the free tightening and thank the market for it. Yen strength reduces the urgency to hike rather than confirming it.

The constraint has not changed. The Bank cannot let the long end price freely without imposing a capital loss on the domestic institutions holding the paper, and it cannot hold the long end down without periodically surrendering the currency. It has been choosing the currency. Nothing in a Takata speech or an operations meeting notice suggests that choice has been reversed, and the Bond Market Survey published on 1 September is a temperature reading, not a commitment.

Meanwhile the dollar leg of the carry is intact. The two-year Treasury sits at 4.39% against a fed funds effective rate of 3.63%. A front end trading that far above the policy rate is not the shape of an easing cycle being priced. High yield OAS at 2.66pp says no one is being forced out of anything. The carry rebuilds from here.

Where a genuine BoJ shift gets paid is not USD/JPY. It is the global long end. The US 30-year is at 5.27% with a 10s2s spread of 0.4pp, a curve steepening from the back rather than the front, which is a term premium story and a fiscal one. Japanese capital is the marginal price-insensitive buyer of duration everywhere. If Tokyo ever makes domestic paper competitive, that bid goes home, and the mark-to-market happens in Treasuries and gilts before it happens in the currency.

I would fade the yen rally and buy the 30-year selloff that a real move would cause. Only one of those two things is being priced.

155.807 is still a weak yen, and the reason yesterday's move matters is not the Bank of Japan's credibility. It is the price of crude.

Brent is at 97.939, up 7.61% in five days. WTI is at 92.27, up 9.49%. Japan buys all of that in dollars. A yen in the mid-150s against energy at those levels is a transfer, and it runs from households paying utility bills and freight-inflated grocery prices to exporters booking translation gains and to anyone short yen for carry. That transfer has been running for years and the case for tolerating it was that Japan needed the demand. Thirty years of a shortfall earns a lot of patience. It does not earn infinite patience against a nine percent move in crude in a week.

So the repricing is correct and it is too small. A 1.84% day in USD/JPY, with EUR/JPY at 181.136 and GBP/JPY at 210.718 after moves of similar size, is what happens when a market that had written off the BoJ reads a Takata speech and a notice about market operations and adjusts by one notch. The adjustment should be larger.

The tell that it hasn't happened yet is on the receiving side. If a real carry unwind had started, the funded legs would show it. They don't. Dollar-rand is unchanged over five days. Dollar-CNH is flat. Peso is down 0.28%, lira up 0.37%. Nothing in EM has moved, which means positioning has not turned, which means the move so far is repricing by people who read the BoJ calendar, not liquidation by people who are long the trade. The second group is much larger.

Nothing on the US side is going to rescue them. The two-year sits at 4.39% against an effective funds rate of 3.63%, a front end priced for the Fed to be going somewhere it is not going. July core CPI rose more than headline, and that is why the pricing exists. But payrolls fell 23,000 in the same month that unemployment dropped a tenth to 4.1%. Unemployment falling while payrolls contract is the labor force shrinking. That is people leaving, not an economy running hot, and it is the low end of the distribution doing the leaving. A Fed reading that as tightness would be making the error I expect it to make, and even then it is not the leg that determines dollar-yen from here.

The trade that gets hurt is the one that assumes Japanese patience is structural because it has been long. It was a policy choice, subsidized by Japanese savers, and the energy bill is what ends it. Sell the carry.

This is not a carry unwind. The giveaway is silver at 66.9785, up 2.52% on the day the yen ripped.

In a genuine funding shock the sequence is fixed: the most liquid profitable position goes first, and that means metals. Gold rose 1.93% to 4472.965. Copper rose 1.01%. WTI rose 1.14%. High yield OAS sat at 2.66pp, wider by a single basis point. Nothing about that tape says forced. When leverage is being called, the yen and gold do not go up together, because the person buying yen is selling gold to do it.

What the tape says instead is that a Japan-specific policy leg got repriced against an American front end that did not move at all. Two-year Treasury yield 4.39%, unchanged. Fed funds effective 3.63%. The entire compression in the differential came from one side. Takata speaks, the Bond Market Survey lands, and then the BoJ schedules a Meeting on Market Operations. Read the order rather than the content.

The obvious framing is August-2024 redux, and it is wrong on positioning. Look at where the damage concentrated. USD/JPY off 2.25% over five days, EUR/JPY off 2.48%, GBP/JPY off 2.74%. Sterling-yen is the worst of the three, and gilt futures rallied 0.93% on the day against 0.09% in Bunds and 0.08% in Treasuries. Long sterling against yen, financed cheap, was the higher carry-to-vol expression and it is the crowded one. The dollar leg is the one everyone quotes and the least interesting. EUR/USD +0.32%, GBP/USD +0.28%: the dollar barely participated.

Meanwhile the risk currencies rallied. Peso stronger, rand stronger, forint stronger by 1.87%, lira flat at 48.3138. That is not what a funding squeeze does to EM. Carry got bought while its funding currency got bought. Both cannot be a liquidation.

The piece that does not fit is oil. Brent 97.939, up 7.61% in five days, WTI 92.27 and up 9.49%, with US crude stocks excluding SPR drawing 4,450 MBBL. Japan imports every barrel. A move of that size in crude is a straight terms-of-trade hit to the yen, and the yen strengthened 2.25% into it. Either the policy repricing is large enough to swamp a ten percent oil move, which would be a bigger BoJ signal than anything in the speech, or the crude rally is being priced as supply and geopolitics with no demand content and therefore no currency transmission. Those imply different things about what happens next and I cannot separate them from this data.

Position accordingly. The trade that hurts from here is not being short yen. It is being long cross-yen carry in sterling, where the exit is narrowest and the gilt bid is doing its own thing. Watch metals: the day gold falls with the yen rising is the day this becomes the other story.