There is no weak-dollar move to lag. The broad trade-weighted index sits at 118.2548 and moved 0.078 on the day. What happened over five sessions was a euro-bloc repricing that got mistaken for dollar softness, and once you see it that way the EM non-response stops being a puzzle.

Look at which currencies rallied. EUR_USD +1.29%, NOK +1.75%, PLN +1.08%, HUF +0.90%. Poland and Hungary are not carry trades in this move, they are euro beta with a wider spread. They travelled with the single currency because they always travel with the single currency. The currencies that did not move, MXN, ZAR, CNH, and TRY, which went the other way, are the ones whose funding actually depends on dollar liquidity. And dollar liquidity has not loosened. Effective fed funds is 3.63%. The balance sheet is still draining, WALCL down 14,256 million on the week to 6,745,699 million. The tide is not running in. Nothing was supposed to fire.

I will give the Fed its due here, because I have spent enough time predicting the opposite. They have not cut into this and they have not stopped the runoff. Unemployment at 4.1% is a tenth lower, not higher. That is a line being held, and holding it is the harder thing.

The problem is what the rest of the board is saying. Ten-year breakevens are 2.34%. Gold is 4518.955 and up 3.85% in five sessions, silver up 5.61%, WTI up 7.05% to 87.157. Those two sets of prices cannot both be right about the next decade. One of them is a market where inflation compensation is bolted to a CPI whose July core index rose 0.724 against a headline gain of 0.245. Core running hot while headline cools is not a supply shock washing out. It is the residue the 2021 episode left behind and never cleaned up. I take the metals over the breakeven.

And the long end agrees. Thirty-year at 5.23%, ten-year at 4.69%, ten-minus-two at 0.5 and steepening. The curve is doing the tightening the front end has declined to do, which is the usual sequence when a bond market starts pricing the debt stock rather than the mandate.

So the call. EM carry does not catch up. The gap closes with the euro complex giving back rather than MXN and ZAR rallying into it, because the euro's strength here is fiscal premium leaking out of Bunds and Gilts, both of which fell over the week, not a global risk bid. High yield at 2.75pp tells you risk appetite is already spent, not building.

The trade in the FACTS is distillate: stocks down 1,530 MBBL while crude built 4,405. Crude is comfortable, the barrel is not.

There was no weak dollar last week. The broad trade-weighted index sits at 118.2548 and moved 0.078 on the day. What happened was a euro-bloc repricing that dragged its satellites along: EUR_USD +1.29%, and then PLN, HUF and NOK following at -1.08%, -0.90% and -1.75%. Norway had oil doing the work as well, with WTI up 7.05% over five sessions. Poland and Hungary trade as euro derivatives and always have. Calling that a dollar move and then asking why Mexico and South Africa did not join is asking why the passengers in a different car did not arrive.

So the framing is wrong before the puzzle starts. EM FX is not lagging and it is not being discounted. It is being priced off the thing that actually governs it, which is the US long end, and the US long end is not cooperating. Thirty-year at 5.23, ten-year at 4.69, funds effective at 3.63. A curve that steep with a policy rate that low is not telling you the Fed is about to ease into a carry-friendly environment. It is telling you term premium is rebuilding. Carry trades need a stable long bond, not a soft dollar, and nobody funding a peso position cares much about the euro cross.

The gold and silver move fits that reading rather than contradicting it. Bullion up 3.85% and silver up 5.61% while high yield OAS sits at 2.75pp is not fear. Credit is priced for nothing to go wrong. What is being hedged is the fiscal and duration side, and that hedge does not pay out through EM currencies.

The inflation data deserves more weight than the market is giving it. Core CPI rose 0.724 on the index in July against 0.245 for headline. Core running well ahead of headline is not an energy story and it is not a base effect. It is services, and it is stubborn, and breakevens at 2.34 are not pricing it. Payrolls fell 23 thousand while unemployment dropped a tenth to 4.1, which is participation leaving, not labour tightness. A Fed reading a soft core would be cutting. A Fed reading that split holds, and holding is what keeps the long end elevated and the carry trade dead.

Which is the part worth naming. Jakarta and Johannesburg did not cause sticky US services inflation, and their currencies will spend the next two quarters absorbing the consequence of it anyway. USD_TRY up 0.38% is idiosyncratic and Turkey owns that one. The rest of the complex is flat because the transmission channel that would lift it runs through Washington and is closed.

I do not expect the catch-up. If US core stays where the July print put it, EM FX is weaker from here, not stronger, and the euro-bloc move retraces first.

The dollar did not slide. EUR did the work, and the broad trade-weighted index tells you so: 118.2548, last print moved -0.078. That is not a currency being sold. What happened over five sessions was EUR_USD +1.29% with PLN, HUF and NOK trailing behind it, and those three are not independent observations. They are EUR beta with a spread. Poland, Hungary and Norway moving 1.08%, 0.90% and 1.75% against the dollar while Mexico moves 0.41% is one trade reported four times.

So the question is not why EM carry failed to fire. It is why anyone expected it to. Carry prices off the front end and off vol. The front end did nothing. DGS2 sits at 4.19 with a zero change, effective funds at 3.63, also unchanged. What moved was the back: 10y to 4.69 and 30y to 5.23, both up four basis points, 10s2s out to 0.5. A steepening driven by the long end is a term premium event. Term premium events do not pay carry. They punish duration, and EM local curves are duration.

Look at the crosses and the funding leg becomes obvious. EUR_JPY +1.01% over five days, GBP_JPY +0.79%, USD_JPY up 0.58% on the day at 159.069. The funder was yen, not dollars. This was short JPY into European assets, and it never needed to touch MXN, ZAR or CNH, which is exactly why those three sat within half a percent of unchanged. High yield OAS at 2.75pp confirms nobody was reaching for spread. They were reaching for a specific regional expression.

The commodity leg fits the same read rather than the EM one. Gold +3.85%, silver +5.61%, WTI +7.05% and Brent +5.86% against a crude build of 4405.0 MBBL excluding SPR. Crude rallying into inventory accumulation is not a demand print. It is the same real-rate trade that sold the 30y, with breakevens at 2.34 and up four basis points. Metals and the long end are telling one story. EM FX was never asked.

The position that costs money here is short yen against Europe. It is levered, it is crowded, and the exit is the same door everyone came through. USD_JPY at 159.069 and EUR_JPY at 185.768 are the levels where a BoJ repricing stops being an academic question. When that unwinds, the EUR leg goes with it, and the PLN, HUF and NOK beta goes faster because it is the same trade with worse liquidity.

EM carry is not being discounted for something nobody has named yet. It simply was not in the trade. The pain is in the funding currency, and it is not priced.