Australian dollar vapourised as US jobs go nuts

It was a huge DXY and EUR reversal Friday night:

dxy

AUD was vapourised:

audusd

Markets are still short:

All commodities puked:

brent

Miners (ASX:RIO) did a bit better:

rio

EM stocks (NYSE:EEM) fell:

eem

Junk (NYSE:HYG) stunk:

hyg

The Treasury curve collapsed:

yields

Stocks fell moderately:

spx

The entire risk rally has been built on a falling DXY. But markets are now short that currency and US jobs caught them off guard.  JPM:

The January jobs report was shockingly strong, as the 517,000 nonfarm job growth figure was almost double the December outturn and nearly three times greater than expectations. Just as surprising, average weekly hours surged 0.3 hour to 34.7 hours. The combination of the two—total hours worked—increased a whopping 1.2%, one of the strongest prints outside the immediate post-pandemic period. The rest of the main metrics were strong, but nowhere near as surprising as the two inputs into hours worked. Average hourly earnings increased 0.3%, as expected and in line with the recent trend. The unemployment rate edged down to 3.4%, though that was all due to new population controls, and the participation rate (genuinely)ticked up to 62.4%, matching the highest reading of the cycle. On a non-seasonally adjusted basis nonfarm employment contracted 2.5 million last month. Seasonal adjustment is applied for a good reason, and employment has contracted in every January on record. Even so, the non-seasonally adjusted data can shed some light on what what’s happening under the surface. Last month’s contraction in employment was the smallest for a January going back to the mid-90s. So perhaps rather than a hiring boom, what we see in this report reflects employers’ reluctance to let go of workers, given the perceived difficulty of eventually re-hiring. If so, this doesn’t bode well for profits. Of course, if demand deteriorates enough firms will let go of workers, but this dynamic suggests demand may have to get hit harder before that occurs. Even before today we had been leaning toward adding another 25bp Fed hike in May, and this tips us toward that call, taking our terminal rate to 5.0-5.25%. Some on the FOMC who might have favored 50bp at this week’s meeting may be having a told-you-so moment, but we think any further hikes we see this year will be 25s.

The January headline employment figure was flattered by 36,000 due to returning strike workers in the public education segment, but the ex-strike figure of 481,000 was still a 185,000 acceleration from the prior month. The 443,000 increase in private sector employment was the strongest since July, driven by a 397,000 increase in private services. Some of the biggest gains there came in leisure and hospitality (128k) and health care (79k); temp help swung back into the positive column (26k). Information sector employment was down 5k, so there are some of your tech layoffs. Still no signs of weakness in construction employment (up 25k) or manufacturing employment (up 19k). The three-tick increase in the workweek for all workers was the most for that series, which began in 2007, save for a few months around the pandemic. The workweek for production and non-supervisory workers increased two ticks, and that’s been beat only a few times in a series that began in 1964. The average hourly earnings figures were some of the few that gave support to the soft landing story. The year-ago increase in earnings for all workers continues to trend lower and was down to 4.4% last month. The 0.2% increase for the better-measured production and non-supervisory worker series was the lowest in two years. Private labor income increased 1.5% last month, the most since the summer of’20, which should leave consumers well positioned to pull out of their late-year spending slump. There were three revisions to the establishment survey: benchmarking to the March 2022 census of employment, updated seasonal factors, and an updated industrial classification of workers. We don’t believe any of these biased the major metrics in one direction or another.

The household survey was less confounding. The 3.4% unemployment rate last month was the lowest since the late 1960s (when the Great Inflation was getting underway) though as mentioned above, the tick down from December was due to compositional effect of the annual rebasing to new population controls. Similarly, the 894,000 reported increase in the household measure of employment was also significantly influenced by the new controls. Removing that effect left only an 84,000 increase in employment as measured by the household survey and the smoothed trend in this series continues to run below the establishment survey. The one metric where all should agree that good news is good news is the increase in the participation rate, and last month’s tick up was supported by a three-tick increase in the prime-age (25-54 y.o.) participation rate. The employment-to-population ratio ticked up to 60.2%, the high of the cycle and about where we were in 2017. The BLS doesn’t provide the control effects for the more detailed aspects of the household survey, so it’s difficult to say much about e.g. full-time vs. part-time, etc., but it’s safe to say the labor market remains very tight.

us employment

Monetary policy effects are lagged. Never more so than during a period of powerful fiscal runoff and a pandemic-inspired labor-hoarding cycle.

DXY is likely to run for a while on the strong jobs report and last week’s ECB pivot leaving a wildly over-extended risk rally exposed to higher Fed terminal rates.

AUD will follow.