Nonfarm payrolls dropped on Friday. If you traded the first spike in EUR/USD or DXY, you already know how that typically ends.

The headline number is the least useful part of the report for FX. Here's the hierarchy institutional desks actually use, and why the first 30 minutes so often reverse.

---

1. The market is trading the revisions, not the print

The headline jobs number is backward-looking and noisy. What actually moves 2-year yields — and therefore the dollar — is the combination of:

• Revisions to the prior two months. A "beat" that comes with large downward revisions is not a beat. A "miss" that comes with upward revisions is not a miss.

• The unemployment rate vs. the participation rate. Unemployment rising because more people entered the labor force is a completely different signal from unemployment rising because payrolls stalled.

• Average hourly earnings. For FX, wages are the inflation channel. A soft jobs print with hot wages is still dollar-positive if it keeps the Fed on hold. A hot jobs print with cold wages can be dollar-negative if it accelerates cut pricing.

If you only traded the headline vs. consensus, you were trading the least informative line in the table.

---

2. Why the first move so often fades

Two mechanical reasons:

Positioning into the event. Dealers and funds typically cut risk into NFP. The initial spike is thin liquidity plus stop-runs, not a new thesis. Once the full table is digested (usually 15–40 minutes later, once the household survey and revisions are parsed), the real move starts — often in the opposite direction of the headline.

The rates market leads FX, not the other way around. Watch the US 2-year yield in the first hour, not EUR/USD. If 2s rally (yields down) on a "strong" headline, the dollar will follow the 2-year, not the jobs print. If 2s sell off on a "weak" headline because wages were hot, fade the initial USD dip.

The FX pair is a lagging expression of the rates reaction. Trade the 2-year first, the dollar second.

---

3. A simple checklist for the next payrolls Friday

Before you put on a post-NFP FX trade, answer these in order:

1. Did the 2-month revision confirm or contradict the headline?

2. Did AHE (wages) move with jobs or against them?

3. Where did the 2-year yield settle 45 minutes after the print, relative to where it was at 8:29am ET?

4. Only then: does EUR/USD or DXY still have room to catch up to that rates move?

If the pair has already done the full 2-year move in the first 10 minutes, there is nothing left to chase. If it hasn't, that's the trade.

---

The takeaway

Payrolls is not a "buy or sell the dollar" event. It is a rates event that FX prices with a lag. Read the revisions and wages, wait for the 2-year to settle, then decide. The traders who lose money on NFP are almost always the ones who traded the headline in the first five minutes.

---

We walk through the actual data reaction — rates first, FX second — every week in the Sunday briefing. Want it free? Join the weekly macro summary. No cost, no catch.

Here's a pattern that confuses most retail traders: the dollar rallies during a global panic, and then a few months later it rallies again during a US growth boom. Same currency, opposite backdrops, same direction.

That's not a contradiction. It's the Dollar Smile — one of the most useful frameworks in FX, and one almost nobody outside institutional desks talks about.

---

1. The three regimes of the smile

Plot USD strength on the y-axis and "the state of the world" on the x-axis, and you get a U-shape (a smile) with three distinct zones:

• Left side — global crisis. Risk-off, liquidity scramble, everyone needs dollars to cover obligations and unwind leverage. USD rallies on fear, not fundamentals. This is the safe-haven bid.

• Middle — the trough. Global growth is synchronized and stable, risk appetite is healthy, capital flows toward higher-yielding, higher-beta currencies (AUD, EM, commodity FX) and away from the dollar. This is where USD is typically weakest.

• Right side — US outperformance. The US economy and rate outlook are clearly stronger than the rest of the world. Capital flows into US assets for the growth and yield, not out of fear. USD rallies again — this time on strength, not stress.

The mistake most traders make is assuming "dollar up" always means "risk off." It only means that on the left side of the smile. On the right side, dollar up can coincide with a rising S&P 500.

---

2. How to tell which side of the smile you're on

Ask two questions:

1. Is volatility spiking or calm? VIX above ~25 and rising, credit spreads widening, funding stress in cross-currency basis → you're on the left side (fear-driven USD).

2. Is the US outgrowing everyone else on the data? ISM/PMI beats, payrolls surprising up, Fed pricing hawkish relative to ECB/BoJ, and equities calm or rising → you're on the right side (growth-driven USD).

If neither is true — vol is low and the US isn't clearly outperforming — you're likely in the trough, where the dollar tends to grind lower and carry trades work best.

---

3. Why this matters for positioning right now

The costliest mistake is holding a "long USD as a hedge" position through the trough, expecting it to behave like a crisis hedge, and watching it bleed for months while carry currencies rip. The dollar is not a permanent safe haven — it's a safe haven conditionally, and a growth trade conditionally, with a soft middle where it does neither job well.

Before you put on a USD position, identify which of the three zones you think you're in. The trade thesis, the correlation to equities, and the right currency pair to express it through are completely different in each one.

---

The takeaway

"Dollar up" is not one trade — it's three different trades wearing the same ticker. Figure out whether fear or outperformance is driving it before you size the position, or you'll misread the signal every single time the regime shifts.

---

We map which zone of the smile we're in every week as part of the Sunday briefing. Want it delivered free, no cost, no catch? Join the free weekly macro summary.

Most FX traders treat late September like any other week. Institutional desks do not.

Every quarter-end, and especially every year-end, a mechanical bid for dollars shows up that has almost nothing to do with Fed policy, growth data, or the latest headline. It is a funding and balance-sheet event. If you trade EUR, JPY, or EM FX into month-end without accounting for it, you are fighting a flow that does not care about your setup.

---

1. Why dollars get scarce at quarter-end

Banks, asset managers, and corporates all have to show cleaner books on the last business day of the quarter. That means:

• Unwinding leveraged FX positions that consume balance-sheet capacity

• Rolling short-dollar funding (FX swaps, cross-currency basis) before the reporting date

• Window-dressing USD cash so regulatory ratios look better overnight

The result is a temporary squeeze in dollar liquidity. The cross-currency basis typically cheapens (more negative), overnight USD funding rates tick up, and high-beta currencies (AUD, NZD, EM) get offered into the last few sessions of the quarter.

This is not a "risk-off" call. Equities can be fine. The dollar can still grind higher purely because someone needs to own it on the print date.

---

2. The pattern most retail charts miss

Look at DXY and EUR/USD in the final 5–8 trading days of March, June, September, and December across the last decade. You will see a statistically noisy but real bias:

• USD tends to firm into the quarter-end date

• The move often reverses in the first 2–4 sessions of the new quarter as the funding demand disappears

The trap is treating that firming as a new trend. It is frequently just a calendar effect. Traders who chase the dollar higher on September 29 often give it all back by October 3.

The tell: watch the USD/JPY 3-month cross-currency basis and SOFR–FF spreads. If basis is cheapening into month-end while the macro narrative is unchanged, you are looking at a flow, not a thesis.

---

3. How to actually use this next week

• Do not fade a dollar bid into September 30 just because "nothing happened in the data"

• Do not add to USD longs on the last two sessions assuming the trend continues into October

• Size down high-beta FX (AUD, MXN, ZAR) into the final week unless you have a specific, non-flow reason to be in them

• Put a reminder on the first Wednesday of October — that is often when the mechanical bid dies and the real macro tape takes back over

---

The takeaway

Quarter-end dollar demand is one of the few FX regularities that is mechanical rather than narrative. Treat the last week of September as a liquidity event, not a policy event. Get the calendar right and you stop fighting a bid that was never about your chart.

---

This is the kind of flow-versus-fundamentals distinction we walk through every week. If you want the Sunday briefing without paying anything, join the free weekly macro summary.

Every cycle, the same headline shows up: "Yield curve un-inverts — recession fears ease."

Every cycle, that headline is wrong about the timing.

Here's what the data actually says, and why the un-inversion is arguably the more dangerous signal, not the less dangerous one.

---

1. Inversions warn you. Un-inversions confirm you.

The 2s10s spread inverts when the market prices in future Fed cuts faster than it prices in near-term growth. That's the warning phase — recession risk is rising, but the damage hasn't shown up in the real economy yet.

The un-inversion happens for one of two reasons:

• Bull steepening — the Fed actually starts cutting because growth is cooling, and short rates fall faster than long rates. This is the "recession has arrived" steepening.

• Bear steepening — long yields rise on inflation or fiscal concerns while short rates stay anchored. This is a very different, less ominous move.

Look at every US recession since 1970: the curve un-inverted before the recession started, not after. NBER dating confirms it — the average lag from un-inversion to recession start is around 6-10 months.

---

2. Why this matters for your FX positioning

A bull-steepening un-inversion is typically dollar-negative into the easing cycle, but the initial reaction is often the opposite — a relief rally in risk currencies (AUD, NZD, EM FX) as markets price a "soft landing" before the labor data actually confirms it.

That relief rally is where most retail accounts get trapped long risk-on currencies right before the real growth data disappoints and the dollar reasserts itself as the safe haven.

The tell to watch: claims data and the unemployment rate. If claims are still low when the curve un-inverts, you're likely in the bear-steepening, less dangerous camp. If claims are already trending up, the bull-steepening recession scenario is live.

---

3. What to actually track this week

• US initial jobless claims — the highest-frequency read on labor deterioration

• 2s10s and 3m10y spreads — watch the rate of change in the un-inversion, not just the level

• Credit spreads (HY OAS) — if credit isn't confirming the curve story, be skeptical of the curve story

---

The takeaway

Don't treat "the curve un-inverted" as a green light. Ask why it un-inverted. That single question separates traders who get chopped up in the relief rally from those who position for what actually comes next.

---

We break down setups like this every week with full context and levels inside TheMacroDesk Pro. Want the same analysis delivered straight to your inbox for free every Sunday? Join the free weekly macro summary — no cost, no catch.