JOBS!
The US economy added far (far) more jobs than expected last month and annual revisions pushed total hiring in 2023 beyond the three million mark.
Together, January's NFP print and the upwardly revised figures for last year constituted a shock to consensus. And I do mean a shock.
Economists, God bless 'em, expected 185,000 from January's headline. The actual print was 353,000, more than 50,000 above the highest guess from 77 forecasters.
Revisions, including a huge 117,000 bump to December's
Little surprised with all the talk about potential seasonal adj issues nobody mentioning that seasonals added almost 3M jobs. Take away 13% of that, and you get a negative print.
Jobs are jobs…one strong year.
Neil, I appreciate what you’re saying, really I do, but everyone needs to get over the idea that this data (any macro data, really) means anything or can be somehow squared with reality. All that matters is this: BBG blasts out all-caps red heads cataloguing the main numbers, and traders — carbon-based and especially otherwise — move on those headlines. The moves may fade or partially reverse eventually, but not entirely and at the end of the day, if policy and market prices are set/dictated based on the top-line readouts (not the nuance), then it doesn’t matter. Sure, the myriad absurdities contained in various government data updates, releases and reports are worth deriding or lampooning, but that’s not going to stop algos (or anybody else for that matter) from trading the headlines. I’ve never made any money investing based on nuance or rationality.
I do agree with Neil that these numbers are suspect, but H makes a better point. Who cares? Trade the market in front of you.
My larger point is more important: All of this stuff is made up. It’s plainly impossible to accurately tally and tabulate aggregates for an economy the size of America’s. I’d argue that once you get beyond, say, a farmer’s market, getting a strictly accurate read on economic activity is mostly impossible. These releases are just an excuse for people to trade. That’s obviously not what the BLS or BEA or etc. intends, but that’s the reality. It’s just inputs for traders. That’s why market wraps/recaps often read like 1990s Sportscenter: This is just a game we play every day.
And those coaches with broken models just roam halls gnashing teeth until the capital keeps rolling in … next play?
Given the option positioning you touched on earlier this week, it’s likely it’ll be contorted into a buy signal.
That headline graphic might be more engaging with a short authentic sounding (BKHOOoooo) audio clip, for effect. Not that it is lacking for engagement. The explosive dynamics of the subject matter is deserving however, in my opinion.
I bought my daughter a very expensive lunch to see if I could somehow help her. She lost her nice job a year ago when her company got bought a year ago. Her husband lost his job when his industry leading firm got acquired by a firm that just wanted his firm gone and now it is. Both of these folks are early 50s, highly skilled, on in data management and product development and the other in quality management, compliance and contract management. What I discovered is neither is likely to get a job anytime soon. Even previously skilled headhunters can’t help. They are being driven out of business because firms no longer want to pay them. HR departments are turning to AI to make hires and even that approach doesn’t work because none of the 20 somethings setting the AI filter criteria has a clue how to actually do it. Trade that.
As a software engineer I always chuckle when I read about economic productivity indicators — how can you measure software productivity? Lines of code? Domains created?
I thought SLOC (single lines of code / per day) was a measure? Maybe that is just a foible of my particular sub-industry (where I’ve worked as a hardware engineer, then systems engineer, counting my blessings that “SLOC” was not used against me).
The results from your coding work. Though, I sometimes wonder what use the few printouts of Fortran code I wrote actually were used. Or maybe are even still used today as a subroutine in some legacy program.
Yeah. But the report is based on the workforce numbers and they revised this drastically. Apparently 633k men left the workforce, but only 6k women did at the same time…yeah, that’s not suspect. And if you normalize without this accounting gimmick, then suddenly, the beat isn’t really that much at all.
So makes you wonder if the idea isn’t to make the data look however it needs to, in order to support high rates for longer ?