The dollar rose, and all risk assets plunged after US nonfarm payrolls increased by 162,000 in August while the unemployment rate held steady at 4.1%, the Bureau of Labor Statistics reported. The result blew past the consensus forecast of 55,000 — nearly three times higher — and was five times the average monthly gain of the prior 12 months, which stood at 31,000.

Revisions proved even more important. June's figure was revised up by 11,000 (from 20k to 31k) and July's was revised up by 44,000 (from -23k to +21k). Together, employment over the two months was 55,000 higher than previously reported. In other words, the July jobs decline that prompted so much discussion simply didn't exist.

This report diverged from all leading indicators at once, and I regard the magnitude of that divergence as the day's key fact. ADP counted just 38,000 private?sector jobs — the weakest since January. The ISM services employment index remained in contraction for a second month at 47.8. The manufacturing ISM showed employment at 51.2 versus 52.8 a month earlier. All three signals pointed to slowing hiring, while the official data produced the best result of the year.

Who's right? I'm convinced the difference is not a survey error but methodology. ADP covers only the private sector and relies on its payroll client base, whereas local government education — which added 42,000 jobs — is outside that scope. ISM indices measure the share of firms expanding payrolls, not absolute hires; if hiring is concentrated in a narrow set of industries, ISM can still show weakness.

The composition of the gain supports this view. Accommodation and food services added 59,000 jobs versus a 12,000 monthly average over the year; local government education added 42,000; manufacturing 16,000; construction 22,000; healthcare 13,000. Almost all the increase was provided by two sectors, one of them public. That narrowness favors those arguing for a pause at the Fed and undercuts those who read the headline number as clear overheating.

Wage dynamics remain subdued, which is crucial for the Fed. Average hourly earnings rose by $0.10, or 0.3%, to $37.75, with a 12?month increase of 3.1%. Average weekly hours rose 0.1 hour to 34.4. The lack of wage acceleration means second?round effects from higher energy prices have yet to materialize in the labor market.

This report materially changes the outlook for the September meeting. Christopher Waller had said he leaned toward holding rates if he saw inflation progress, and John Williams cited signs of disinflation. Both arguments relied on a weakening labor market, and that assumption has just fallen apart. I expect the odds of a 25?bp hike to move back into roughly the 65–75% range, with the CPI report on September 11 becoming decisive, since the dovish wing now lacks an employment?based argument.

Yet, despite the multi?faceted labor report, I will venture that the Federal Reserve still may raise interest rates on September 15–16.

Technically, for EUR/USD, the key task for buyers is to hold above 1.1641 — only that would open the way to test 1.1657. From there, the currency pair can reach 1.1673, though without participation from large players such a move seems unlikely. On a decline, I expect serious buying interest only near 1.1621; if demand does not appear, it would be wiser to wait for a new low at 1.1601 or look for longs from 1.1584.

For GBP/USD, the technical picture centers on nearby resistance at 1.3545, which pound buyers must first take. Only then can the pair target 1.3573, above which further gains will be difficult; the more distant target is 1.3596. On the downside, bears will attempt to seize control of 1.3521. If they succeed, a range break would deal a significant blow to bulls and, in my view, push the instrument to 1.3501 with a prospect of reaching 1.3480.