Well we had the ADP employment numbers yesterday–forecast was for 299,000 with actual coming in at 128,000. Of course no one pays much attention to the ADP numbers.
Today we have the ‘official’ government numbers which calls for the creation of 328,000 jobs which is down about 100,000 from last month.
With interest rates hanging in there at around 2.91% we could well see movement if employment is substantially off forecast–say 100,000 or more either way–but likely it will be a non event. We shall see!!
Most macro hedge funds that trade based on the economic cycle focus on YoY numbers rather than taking a MoM number and seasonally adjusting it. And yes, the pandemic put seasonal adjustments out of whack, as a result complicating economic analysis.
A further question: what if seasonality has changed dramatically during the pandemic: would the seasonably adjusted figures not just be wrong? That is, suppose they are adjusting for historical patterns that have not applied recently.
Can someone explain what “seasonally adjusted” means? I haven’t heard the term as much lately, but it has always baffled me.
If there were 300,000 jobs created, what difference does it make what “season” it is? I get it that there’s differences in hiring depending on the time of year – more in summer… more at Christmas… but a number is a number irregardless of the temp outside.
And I also understand comparisons from year to year. I just don’t understand what they are ‘adjusting’.
By adjusting for seasonality it smooths the data out for easier comparisons. Or you can see trend changes more easily. Otherwise you will have seasonal spikes and drops that make examining the chart more difficult for what is going on with a non-seasonal viewpoint.. Since it is hopefully done year after year with the same math it should work out well. I hope I made a little sense.
No, not really.
I understand they are trying to smooth out the data, but with enough data it really shouldn’t matter. If I take a ten year trend line and look at all the July’s, then I can see how this July or that July compares to other July’s.
It just seems to me that they are taking an actual number and making it subjective (of course with the help of reams and reams of data….).
So, maybe I should seasonally adjust my income? I work in construction. I have busier months and slower months. Because I’m so stupid and I can’t understand the peaks and valleys, maybe I need to adjust it so it looks like I make the same every month?
It just seems ridiculous to have to adjust something that starts as a concrete number. I guess that’s why I’m in construction and not an economist or wall street wonk. I think I’m going to start calling out the numbers on my tape measure in “seasonally adjusted” increments. I need this board at 36 1/2 inches, but due to seasonal adjustments, what I really mean is 37 1/4.
Mark – far be it for me to be an apologist for any Govt data and their ability to massage numbers to their liking by the use of such stats as “seasonally adjusted.”, but out of curiosity, I’m wondering if you have a family budget. In other words, even though you have peaks and valleys in your monthly income because of the nature of your work, do you adjust your spending so it looks like you make the same every month? If you do, aren’t you seasonally adjusting your spending to ignore the peaks and valleys of your income?
2WR – good point, but no, I do not have a budget. And I am fortunate enough to not be living ‘paycheck to paycheck’. If I need it, I buy it. If I want it, I think about it, and sometimes buy it. I don’t think I have ever taken the monthly fluctuations into account when assessing a purchase. I have always been pretty frugal with my spending though and know well in advance whether I can afford something or not.
And there’s really only 1 item that adjusts seasonally which is my electric bill. I can opt for a plan where I pay the same amount monthly throughout the year (which would then make it seasonally adjusted I guess – I overpay some months and underpay other months), but the difference is not enough for me to bother. Plus, I like knowing exactly how much my electricity costs. I know it goes up in winter and down in summer…… When there’s the occasional outlier, I get the opportunity to reflect on that to try and understand it. Usually it was just a particularly cold month.
Speaking of temperature, maybe we could seasonally adjust that too. “It’ll be 110 in Denver today, but don’t turn up your AC, because seasonally adjusted, it’s really only 83″ Or when Texas froze – ” I don’t know why your pipes burst, because seasonally adjusted, the temperature was 49.” …..
In the end, I guess I’m just an actual numbers guy. I know, to the penny, how much is in my checking account. I’m going to cut that board to the nearest 16th of an inch, if not even more precise. I just have a personal problem with fudged, I mean ‘seasonally adjusted’ numbers. My brain just doesn’t work that way, and it causes me anxiety.
Is there a way to retrieve *non*-seasonally-adjusted numbers?
I tried looking around some, including bureau of labor statistics, but got impatient and gave up. It’s got to be there somewhere.
I just came out of my non-seasonably adjusted slumber.
Jobs data is like predicting which way the 10yr bond will swing or natural gas price in summer.
Yes, July to July may be analyzable without seasonal adjustment. But June to July and July to August might not. If there was a seasonal spike in July for some data, it might obscure some trends from June to July and July to August.
Not terribly satisfying:
https://www.bls.gov/cps/seasfaq.htm
Looks like un-massaged numbers are available.
Even car makers and home construction figures seem to smooth out numbers, but is probably different.