Newsletter · 2026-09-18 · Ashutosh Agarwal

The Fed Hiked and EM Held the Line - EM FX Weekly - Week of September 18, 2026

EM FX Weekly for the week of September 18, 2026: new Fed chair Kevin Warsh delivered a hawkish, unanimous 25bp hike, the dollar finally rallied back above 100 on the DXY, and J.P. Morgan's EM desk graded its own homework, arguing EM held day one and that the dominant driver of these currencies is now oil, not the Fed, with the Bank of Japan the live event still to come.

EM FX Weekly

Week of September 18, 2026: The Fed Hiked and EM Held the Line

Last week the whole EM world was holding its breath for one 48-hour window: the Fed on Wednesday, the Bank of Japan the day after. Well, the first shoe dropped. On September 16, new Fed chair Kevin Warsh delivered the first US rate rise in three years (25 basis points, and, more startling, a completely unanimous vote), then stood at the podium and made clear he thinks he has more to do. Bond yields briefly punched through 5% for the first time since 2007, stocks sold off, and the dollar did something it hadn't managed in almost two months: it actually went up.

And here's the part that matters if you run emerging-market money. The day after the hike, J.P. Morgan's emerging-markets desk sat down and asked the exact question we posed last week (is the coast clear now that the Fed has moved?) and their answer was calm, specific, and a little surprising. Not only did EM hold on day one, but the desk's whole framing has quietly shifted: the thing driving these currencies now isn't really the Fed at all. It's the price of oil. That reframe is the spine of this issue.

(A few plain-English definitions we'll lean on: the carry trade means borrowing cheaply in a low-yielding currency and parking the cash in a high-yielding one to pocket the interest-rate gap. The dollar index (DXY) measures the greenback against a basket of big rich-world currencies. A basis point is one-hundredth of a percentage point. And carry just means the yield you get paid to hold a currency.)

TL;DR

The Fed hiked 25bp to 3.75%-4.00%, unanimously, and Warsh signaled more. It was the first hike since July 2023 and the first under Warsh. The vote was 12-0, the dot plot showed 16 of 18 officials expecting at least one more hike this year, and market-implied odds jumped to 54% for October and 90% for December, per CNBC's Steve Leisman on Squawk Pod.
EM held the line on day one, and J.P. Morgan says the Fed was never the main event anyway. On J.P. Morgan's own podcast, EM strategist Aneska Kristovova said the recent moves in EM currencies have tracked energy prices far more closely than Fed pricing: oil "has probably played an even larger role than the Fed repricing." The high-yielding, high-carry currencies stayed resilient, exactly as the desk expected.
The dollar finally rallied, last week's big "tell" flipped. The DXY reclaimed 100 for the first time since the end of July, its first real move up in almost two months, driven by a breakdown in the euro (Market Maker; Macro Voices). For a soft-dollar carry trade, that's the one thing to watch.
The 10-year yield hit 5.04% intraday (a near-20-year high) then reversed back below 5% once the hike landed. J.P. Morgan reads the hawkish, credible hike as putting a ceiling under long-term rates, which would be good news for EM.
J.P. Morgan thinks the EM rate sell-off is close to exhaustion. Over 16 years, EM rates only fall on about 55% of days in these sell-offs; the desk sees momentum near the point where the move usually subsides, and expects a near-term "rest bite."
The Bank of Japan is the live event now, and the pods didn't yet have the verdict. As of Wednesday night's recordings, the BOJ decision was still hours away. The worry, per Ed Yardeni on Bloomberg, is that a more hawkish BOJ speeds up the unwind of the yen carry trade and pressures global bonds.

What's new

The one that matters most: J.P. Morgan's EM desk answers last week's question.

Last week we flagged that J.P. Morgan's EM strategists had ranked the survivors and casualties of a Fed hike. This week (the morning after the hike actually happened) the same desk reconvened to grade its own homework, and it's the clearest EM-specific discussion of the week by a mile. On J.P. Morgan's At Any Rate, "EM Fixed Income: Is the coast clear after the Fed?" (Sep 17), EM sovereign credit head Ben Ramsey, local-markets strategist Aneska Kristovova and rates strategist Mike Harrison walked through what the hike actually did.

Ramsey set the scene: the realization of the hike, even though it was widely expected, was "at least initially a stabilizing factor for markets." The 10-year was back below 5%, the dollar was "giving back some of those gains" from its knee-jerk pop, and EM credit was "opening up with a firmer tone."

Then came the counterintuitive part. Kristovova pointed out that for emerging markets, a credible, hawkish Fed is a double-edged sword:

"For us in emerging markets, Fed credibility is a double-edged sword... because we are, to some extent, a comparative asset class to U.S. assets. So actually, for us, if the Fed was less credible, my conjecture would be we would have traded even better... we would have probably gained more versus the dollar, not less."

In plain terms: because investors weigh EM currencies against US assets, a Fed that looks firmly in control can actually pull money toward the dollar and away from EM. So a hawkish Fed is, on balance, a negative for EM, but, crucially, "mainly through the EMFX channel," and the degree of that hit is "just smaller than in previous historical episodes" thanks to healthier starting points: better balance of payments (a country's money-in-versus-money-out with the world), light positioning, high local real yields, and growth that's firming everywhere at once.

The real reframe, and the single most useful idea this week, is what Kristovova said is actually moving EM currencies:

"The recent repricing in EMFX has had quite large correlations to energy markets... it actually seems to us that that has been probably the dominant factor for FX compared to the Fed repricing... the oil price has probably played an even larger role than the Fed repricing."

That flips the story. For weeks the question was "which currency survives the Fed." The desk is now saying: watch oil. And there's a hopeful wrinkle: both oil and natural gas have now "exceeded the guidance from our commodity team," which the desk sees as a possible mean-reverting force. If energy cools, EM currencies can keep trading resilient. If it pushes higher, the worry moves from the Fed to growth.

One more thing that held exactly as advertised: the high-carry currencies did the heavy lifting. "Higher carrying currencies have been our favorite. And that has really shown through in recent weeks... And I continue to expect that going forward."

On EM rates, the message was "this sell-off is nearly done."

Mike Harrison made the point that the recent sell-off in EM local rates is not an EM problem, it's a global one. The spread of EM rates over US rates "is essentially unchanged throughout this whole sell-off," meaning the market is repricing a more reflationary world everywhere, not punishing EM specifically. His read on how much further it can run is worth quoting because it's the kind of concrete history that's hard to find:

"If I look at previous examples over the last 16 years of EM rates and these kind of multi-month sell-off periods, you only sell off on 55% of the days. It just happens to be that when you sell off, the rate sell-offs tend to be larger than on the days when you rally."

His conclusion: near-term, "we're probably in for a bit of a kind of rest [bite]," a pause in the sell-off. He described the world as "bimodal": the market is pricing two fat tails, one where central banks deliver more than four hikes over the next year, and one where they don't hike at all and even cut. Over the past month the "no-hikes-or-cuts" tail got priced out hard and shifted into the "four-plus hikes" bucket. The swing factor, again, is energy: a sharp drop in oil would let the front-end hikes get priced out fast (bullish for rates, good for high-yielders), while an energy-driven stagflation scare is the nastier path, where lower-yielders price out hikes and, the risk for a carry book, high-yielders finally underperform if growth expectations crack.

On EM credit, a possible turn for the better. Ramsey noted that the treasury sell-off fed almost one-for-one into EM sovereign yields but "not into spreads": spreads have stayed range-bound at very tight levels because fundamentals have improved and countries used the last two years of market access to push out their near-term debt maturities. His tentative but genuinely new observation: on day one after the hike, EM credit didn't rally hard but it stabilized, with spreads "a bit tighter... notwithstanding the fact that we see treasuries rallying." If the Fed's credibility really has put a ceiling on rates, he thinks "we could be back in the universe where spreads could start to grind tighter again," a possible inflection out of the range.

The dollar finally rallied, and that's the tell that flipped.

For weeks the defining puzzle was a dollar that refused to strengthen even with hot inflation and $100 oil. That changed the moment the Fed moved. On Market Maker, "Why the Fed Is Raising Rates Again" (Sep 17), Anthony Cheung and Piers Curran walked through the reaction: two-year yields rose (highest since June 2024), the 30-year "actually fell a touch" as bondholders regained a little faith the Fed will curb inflation, and the dollar "gone back above 100 on the dollar index... the first time we've been above 100 since end of July." The technical crowd flagged the same thing on Macro Voices, "MacroVoices #550 Harley Bassman: In FED We Trust" (Sep 17): the dollar "bullishly broke out on the DXY," driven by "a breakdown in the euro," and was "the single most interesting thing to watch" out of the Fed. The important caveat from the same discussion: "one day does not make a new trend," and there isn't yet enough follow-through to confirm a fresh dollar bull market. But after weeks of a soft dollar being the quiet engine under EM carry, this is the datapoint to respect.

The hike itself: hawkish, unanimous, and about credibility.

The mechanics are worth getting right because they explain the reaction. On Prof G Markets, "Fed Hikes Rates For First Time In 3 Years" (Sep 17), FT commentator Robert Armstrong called the unanimity the real story: several members had "made slightly dovish noises going in," so getting everyone on board (or joining them) "puts them in a strong position politically" and sends a message on Fed independence: "if you screw with one of us, you're screwing with all of us." Warsh's mantra, repeated from Jackson Hole: "We must be confident that underlying inflation is moving to our objective clearly and at sufficient speed." On the fear that hiking can't fix an oil-driven supply shock, Armstrong relayed Warsh's own answer: the Fed can't "create hydrocarbon atoms," but it can stop high oil prices from "infecting" wages and spreading into broad inflation.

On Squawk Pod (Sep 17), CNBC's Steve Leisman noted the detail markets didn't love: not just the unanimous vote, but "16 of 18 members" seeing at least one more hike, with implied odds of "54% for October and 90% for December." The Dow fell about 1.2% (600-plus points) on the day, then bounced more than 300 the next morning; the two-year jumped 13 basis points; the VIX still closed at a sleepy 16. Warsh's most-quoted new phrase, that the Fed had "removed a dose of accommodation," got flagged as a genuine hawkish signal on Monetary Matters (Sep 16), where former Fed staffer Joseph Wang called it a "bombshell": it implies Warsh still thinks policy is accommodative, which is why Wang expects "two more hikes, not one." Wang also flagged the higher-for-longer signal buried in the dots: the 2028 median rate rose to 3.9% from 3.4%.

The dissent that wasn't: on Halftime Report (Sep 17), a clip of DoubleLine's Jeffrey Gundlach argued the Fed should have gone bigger, "I actually would have dissented and voted to raise rates 50 basis points today... stun and done," to close the gap with the two-year yield in one move. Nobody thought 50 was truly on the table, but it captures the "there's catching-up to do" camp.

The Bank of Japan is the live event, and the verdict wasn't in yet.

Here's the honest limit of this week's podcasts: they were recorded before the BOJ decided. On Saxo Market Call, "Warsh hawkish, but market quick to reverse. Now comes BoJ." (Sep 17), John Hardy noted the market had already largely unwound its knee-jerk reaction to the Fed, and turned to the main event still ahead: "this Bank of Japan meeting tonight... I have an inkling they're going to be quite hawkish. The question is, is that priced in?" He noted traders were using the hawkish Fed as an excuse "to get short of yen crosses once again." The stakes were laid out most clearly by Ed Yardeni on Bloomberg Surveillance TV (Sep 17): he's "particularly concerned about the unwinding of the carry trade in Japan," because US officials, Treasury Secretary Scott Bessent chief among them, keep pressuring Tokyo to hike faster, which strengthens the yen and stimulates the unwind. His rule of thumb: "the more hawkish this BOJ is... the more vulnerable the global bond market." A quarter-point move "will keep the carry trade from unwinding faster"; a surprise half-point "might be more of a shock." We'll have the actual outcome in next week's issue.

The debate

This was a genuinely two-sided week, and the fault line runs straight through the carry trade.

The bull case (EM held, and the Fed was never the real threat). J.P. Morgan's EM desk is the standard-bearer, and it's the closest thing to a specialist institutional voice on EM this week, not a TV pundit riffing on macro. Their argument: the hike stabilized rather than shook EM; the sell-off in EM rates is a global reflation story with the EM-over-US spread unchanged, and it's near exhaustion; four Fed hikes is probably peak pricing; the high-carry complex stayed resilient exactly as expected; and, the reframe, the dominant driver has been oil, not the Fed, with energy now stretched beyond the house commodity forecasts and therefore a candidate to mean-revert lower. Add a credible Fed putting a ceiling under the 10-year, and EM credit could even grind tighter from here.

The bear case (the dollar turned, the funding leg is turning, and the BOJ could light the fuse). The single cleanest datapoint cuts against the bulls: the dollar finally rallied, back above 100 for the first time since July. A rising dollar is the classic headwind for high-yield EM. On top of that, the cheapest fuel in the carry tank is getting pricier: the BOJ is hiking into US pressure, and, per Yardeni, the global bond route already bears "the clear tracks or fingerprints" of a yen carry unwind that "may not necessarily be over." Layer on a hawkish Fed with two more hikes possibly coming (Joseph Wang), a 10-year that touched 5%, and $100 oil keeping every EM central bank's job harder, and you have the cocktail that historically flushes a crowded, one-way trade.

The honest read: day one went to the bulls: EM held, credit firmed, rates stabilized. But the two things that would hand it to the bears both moved this week: the dollar broke higher, and the BOJ is about to act. As Kristovova herself put it, she remains constructive but the whole thing now hinges on energy: if oil mean-reverts, carry keeps working; if it grinds higher and finally dents growth, the high-yielders she likes are the ones most at risk. Watch oil and watch the dollar's follow-through.

The trades in play

The desks were concrete this week, so here's the actionable read:

Stay in the high-carry complex. J.P. Morgan's clearest EM call is to keep owning the higher-yielding currencies whose carry cushions them, the part of the market that "has really shown through in recent weeks" and where the desk expects continued resilience.
Express EM rates through relative-value structures, not outright. Harrison said the desk is "sticking with RV structures for the most part" and sees the skew toward "some bull steepening" (betting the front end falls faster than the long end in different scenarios) rather than a clean directional bet, given how bimodal the outlook is.
Treat oil as the real EM swing factor. Both the bull and bear paths for EM now run through energy. A sharp drop in oil is the bullish trigger (front-end hikes get priced out, high-yielders outperform); a further energy spike that hits growth is the bearish one.
Watch the dollar's follow-through before adding EM risk. The DXY reclaiming 100 is the first real dollar strength in two months. If it confirms with more upside (a euro breakdown is the driver to watch), that's the signal to trim; if it fades back into range, the soft-dollar carry backdrop is intact.
The BOJ decision, not a price level, is the near-term pivot. A quarter-point keeps the yen-carry unwind slow; a surprise half-point is the shock scenario that would pressure global bonds and, by extension, EM.

Read-throughs

Broad dollar, the tell flipped. After weeks of refusing to rally, the dollar broke above 100 on the DXY (first time since end-July), led by euro weakness (Market Maker; Macro Voices). Not yet a confirmed new trend, but the soft-dollar floor that EM carry rides on just got tested for the first time in a while.
EM local debt (EMB / GBI-EM). J.P. Morgan reads the treasury back-up as a global reflation repricing, not an EM-specific risk premium, and thinks the EM rate sell-off is near exhaustion. Currency gains have been carrying local-bond returns; a 10-year that has stopped rising would take real pressure off the whole complex.
EM sovereign credit. Spreads stayed pinned at very tight levels through the treasury sell-off and even firmed slightly on day one after the hike. If the Fed's credibility caps rates, Ramsey sees room for spreads to "grind tighter again," but the risk he named is that persistently higher rates eventually shut lower-rated (BB) sovereigns out of the market.
LatAm and the high-yielders. The desk's favored high-carry names did their job. Notably, Harrison flagged Brazil as the one place where the market is not pricing higher end-2027 policy rates, a reminder that the BCB is already carrying one of the fattest yields in the basket. Single-name desk commentary on the peso, real, and the rest was thin this week; the story was told at the basket level.
Oil, now the master variable for EM. Brent and WTI both held above $100 even after Saudi Arabia found extra capacity to route crude through Oman (Macro Voices), with US diesel at a record $6.40/gallon (Squawk Pod). J.P. Morgan says energy has overtaken the Fed as the dominant driver of EM currencies, and, with prices now above the house forecast, a pullback would be an outsized tailwind.
Yen and the carry funding leg. The market used the hawkish Fed to short yen crosses again (Saxo), but the bigger risk is a hawkish BOJ accelerating the carry unwind that Yardeni says is already visible in the synchronized global rise in bond yields. This is the cheapest fuel in the tank getting more expensive.
EM Asia / the yuan. Still quiet on the pods as a price story: no dedicated yuan, PBoC-fix, rupee, won, peso, real, rand, lira or CE3 desk episode this week. The EM conversation ran through the J.P. Morgan basket lens, not single-name calls.

What changed

Plenty moved from last week's issue, and it cuts both ways:

The event happened, and EM passed day one. Last week the question was "which EM currencies survive the Fed hike." This week the Fed hiked, hawkishly and unanimously, and EM held: credit firmed, rates stabilized, the high-carry complex stayed resilient. Question answered, at least for the first 24 hours.
The dollar's "won't rally" tell flipped. The defining feature of the last several weeks (a dollar that wouldn't strengthen no matter what) reversed the instant the Fed moved. The DXY back above 100 is the biggest single change, and it's the one that argues against carry.
J.P. Morgan reframed the whole driver: it's oil, not the Fed. A real analytical shift. Last week's frame ranked currencies by Fed sensitivity; this week the desk says energy has been the dominant force on EM currencies, and, because oil and gas are now above the house forecast, the next big move may be a cooling that helps EM.
The 10-year peaked and then reversed. Last week the bond revolt was the escalating threat; this week the 10-year touched 5.04% (a near-20-year high) intraday and then fell back below 5% once the credible hike landed. J.P. Morgan now sees a possible ceiling under long rates.
The Fed sell-off in EM rates is judged near exhaustion. New this week: the desk's concrete history (EM rates fall only ~55% of days in these sell-offs) and its call for a near-term "rest bite."
The BOJ went from preview to live. Last week we were reading a BOJ preview; this week the decision is upon us, with the yen-carry unwind now the front-of-mind risk, but the verdict lands after the podcasts we heard, so it carries into next week.

The trade that runs on cheap yen funding and a soft dollar spent this week watching both of those supports wobble, and, once again, held its ground on the day. Whether it keeps holding now depends less on the Fed, which has shown its hand, and more on two things the podcasts kept circling: the price of a barrel of oil, and what the Bank of Japan does next.