Are you tired of inflation geekery? Me too, actually, but I got curious about something so I'm going to show it to you anyway. Voila:
Let me explain. First I charted the past year of data for four measures of core inflation:
- CPI (Consumer Price Index)
- PCE (Personal Consumption Expenditures)
- PPI Goods (Producer Price Index for goods)
- PPI Services (Producer Price Index for services)
This is not the usual year-over-year metric that I've plotted. It's month-over-month (annualized for ease of viewing) so you can see the inflationary impulse in the economy from moment to moment. Then I let Excel draw trendlines for each. I did nothing to affect it: I just clicked the buttons and let Excel choose the best fit.
Of the trendlines, three appear to have already peaked and are now declining:
- PPI Services peaked in September
- PCE Core peaked in December
- PPI Core Goods peaked around May, or maybe June
Only one measure, core CPI, is still rising.
So . . . what does this all mean? To me it suggests that the inflationary impulse in the American economy is already fading out. I don't know for sure how the Fed should respond to this, but it does seem to confirm that we don't really have any need to panic at the moment.
"PPI Core Goods peaked around May, or maybe June"
Or maybe July Or August, Etc.
Regarding all of this, that a trend has peaked is cold comfort if it has peaked at a large value and hasn't declined much.
Q: What type of regression generated those remarkable dotted-line arcs? Looks like 2nd order polynomial.
If you do linear fits, three more or less match Kevin's curve. PCE Core does not.
My charts are based on eyeball reading of Kevin's chart and best guess which color is for which index. I wish Kevin would provide the data he's using. Here, for spreadsheet fans is a CSV table of my reading.
,pink,blue,purple,orange
month,PPI Service,PPI Core Gooods,PCE Core,CPI Core
Jul,12,8,4,3.5
Aug ,10,5.5,4,2
Sep,1,7,3,3
Oct,2,6.5,5.5,7.5
Nov,12,7,5.8,6
Dec,11,6,6,6.5
Jan,10,10,5.5,7
Feb,6,12,3.5,6
Mar,14,8,4,4
Apr,-1,12,4,6.5
May,7,9,4,7.5
Jun,5,8,,8.5
The trend has to peak before it can go down.
We would not want a rapid descent.
Another dubious set of graphs. Why don't you take for example the last 24 months of data and draw these graphs for say the last 24, 21, 18, 15, 12 and 9 months? I expect you will find the curves changing drastically depending on what period you use meaning they are mostly noise.
I'm not sure what your point is. There are lots of measures that can help us
determine where inflation is going. Why are Kevin's any more "noise" than the latest data on June's consumer price index ? Indeed, choosing 24 months for a timeline, starting at th beginning of the pandemic , would seem to guarantee misleading graphs. I appreciate that Kevin is pushing back, imperfectly (all our measures are imperfect) against all the sky-is-falling screamers on the news.
"I'm not sure what your point is .."
Kevin says things like "... peaked in September .."
My point is, this is an absurd conclusion given the quality of the data he is using. Like predicting the result of an election based on asking 10 friends how they are going to vote.
But the Fed wants unemployment to rise to 6% (sorry, can't find the link to where I read that) so it will increase interest rates by 100 basis points regardless. We must fend off wage increases for the working stiffs, as that is the only and true cause of inflation. /s
I fear you're right.
Alas, the 2% target is writ in stone--though it shouldn't be:
https://www.msnbc.com/msnbc/amp/ncna1297292
(NY Times also has recent article on New Zealand starting the 2% trend)
And they haven't voted right ever since we gave them the vote either :-/
Apropos of nothing, the lowest gasoline prices I saw this morning on my trip to the coffee shop were $3.979 and $3.969.
Safeway this morning: Beef roast $2.77/lb.; Half gal. milk $1.99; Canned veggies, 15 oz, 10 cans for $10. This inflation is out of control.
Okay, I'm off to the store!