Friday, January 16, 2026

Industrial production sets new post-pandemic high in December - but mainly due to utilities

 

 - by New Deal democrat


Industrial production is much less central to the US economic picture than it was before the “China shock,” but it remains an important if diminished economic indicator, particularly since the month it has peaked in the past has typically been the month the NBER has chosen as the economic cycle peak.

In December, headline industrial production (blue in the graph linked to below) rose 0.4%, with previous months revised higher 0.2% on net, establishing a new post-pandemic high, although it remains -1.3% below its 2018 all-time high.  Manufacturing production (red) increased 0.2%, and prior months were also revised higher by 0.2%, but it remained slightly below its September 2025 post-pandemic high:


The difference between the two is mainly due to utility production, which rose 2.6% for the month, and was higher by 2.3% YoY. And all of 2025 on average set new all-time records for production, most likely driven by AI data center needs:


Despite the influence of utility production, this was a positive report, adding to the evidence we have seen in durable goods orders and regional Fed manufacturing reports in the past few months indicating that manufacturing production in particular has been improving. This in turn is most likely due to the lack of new tariff gyrations, and producers having found a modus operandi to deal with the effects of previously imposed tariffs.

That being said, the next comprehensive report on personal income and spending will be crucial to determining whether the autumn lull or downturn in important coincident economic data ended after the end of the government shutdown or not.


Posted by at

Thursday, January 15, 2026

Important scenes from the (recessonary?) December jobs report; was July a cycle peak?

 

 - by New Deal democrat


Last Friday I summarized the jobs report as “show[ing] a contracting jobs market in all important metrics except the headlines (which, for the record, were positive).  …[A]lmost] all of the important leading metrics … were negative, [including the] goods-producing sectors - manufacturing, construction (including residential construction), and temporary jobs - declined, as did the goods-producing sector as a whole. [And] “…[To] be clear: the jobs market is being entirely held up by service providing jobs, which tend to rise even in the earliest stages of recessions. [In short,] This is a jobs report which is ringing the alarms for imminent recession.” 


Let me elaborate on that with several important graphs as linked to below.

Since last April, the total number of jobs in the economy (pending benchmark revisions) has grown by a whopping 93,000. That’s under 12,000 per month! On a YoY basis, total jobs have increased only 0.4%. Going all the way back to WW2, only once has YoY job growth decelerated to such a paltry level (in July 1952) without there being a recession:


Indeed, with only one exception during WW2 (1944), by the time job growth has decelerated this much, a recession had already begun.

Goods-producing jobs have always led the way. These peaked in April, and have declined by -90,000 since. They are now down YoY -0.3%. Only three times since WW2 - in 1952, 1967, and 1986 - have there been such declines without a recession, and in all cases where there was, with the same exception of 1944, the recession had already begun:


A similar situation obtains for aggregate hours worked by nonsupervisory personnel. These are up only 0.7%. This series started in the early 1960s. With the exception of 1967 and single months during 1986 and 1996, before the pandemic such paltry increases had always meant recession:


To be fair, since the pandemic there have been 6 equivalent or worse YoY comparisons without a recession occuring.

Next, let’s compare all three of the above series. What I want to show you in this link is the order in which the declines have typically occurred:


Historically, the pattern has been: first, goods-producing jobs turn negative YoY (red); then aggregate hours worked (gold); and finally total employment (blue). Interestingly, 2025 has been somewhat unique in that YoY hours worked have held up better than total employment - but the pattern will not be broken if hours decline more precipitously from here than jobs. As noted above, YoY goods producing jobs have already turned negative.

Finally with regard to the employment report, real aggegate nonsupervisory payrolls did decline in December from a record high in November:


These had grown only 0.3% from March through September, but jumped 0.5% higher as of November, due to a strong 0.7% nominal advance in payrolls, plus the kludged CPI numbers for those months, that added only 0.2%. Had shelter been more accurately calculated in that CPI report, it is likely that real aggregate payrolls would only have advanced 0.2% or even 0.1% instead of 0.5%. So while it is fair to say that this metric is not recessionary through December, a more accurate reading for the past 9 months may be closer to flat.

Which brings me to a link to one final graph, which is the most updated values for the 4 most important series the NBER takes into account when calculating recessions: payrolls, industrial production, real income less government transfers, and real manufacturing and trade sales, all of which have been normed to 100 as of July. The graph also included nominal total business sales for reasons I will describe below:

https://fred.stlouisfed.org/graph/fredgraph.png?g=1QuhH&height=490

Note that several of the series have not been updated beyond September or October. The point is, only two of the series - payrolls and real personal income - have exceeded their readings in July, both in September, and both by only 0.1%. Further, since total business sales declined in both September and October, and they do not take inflation into account, it is almost certain that real manufacturing and trade sales did so as well.

In other words, there may have been at least a small cycle peak in July, with at least a shallow downturn during the autumn, and in particular during the government shutdown. Whether if so it was pronounced enough, or will last long enough, to qualify as a recession  (pending revisions!) is completely unkown. But the leading metrics in the December employment report are not auspicious.


Posted by at

Jobless claims continue to be very positive, near multi-decade lows

 

 - by New Deal democrat


First, usually the week following the employment report is very quiet, and I put up “scenes from the report” with some important graphs. With all the releases catching up on old data this week, I haven’t done that; but because jobless claims are the only significant data this morning, I intend to put up a very important update on those “scenes” later this morning.


With that out of the way, let’s take our usual look at new and continuing jobless claims. I’ve noted a couple of times lately that there has been a “regime shift” from the end of last June towards lower YoY numbers. And that very much continued in this morning’s data.

Initial claims declined -9,000 last week to 198,000. Aside from a few weeks in the past 3+ years, there have been no numbers under 200,000 since the end of the 1960s! The four week average also declined -6,500 to 205,000. Similarly, aside from the last 3 years, 2018 and 2019, this is the lowest number in over 50 years. Finally, with the typical one week delay, continuing claims declined -19,000 to 1.884 million:
. 

There is a significant caveat, in that as shown in the graph linked to above, this is *very* similar to the post-pandemic unresolved seasonality we have seen in the past few years, notably exactly two years ago. 

All that being said, as usual it is the YoY comparisons that are more important for forecasting purposes. In that regard, initial claims were down -8.5% and the four week average down -3.5%. Only continuing claims remained higher, at 1.8%:


The analysis remains that *very* few people are getting laid off (possibly some of this is due to immigrants in some industries either quitting or getting deported), but those who are laid off are having a more difficult time finding new jobs. I’ll have more to say about that later this morning.

Finally, although I won’t bother with a link to a graph this week, the lower numbers portend a decline in the unemployment rate in the next several months. One year ago the unemployment rate was averaging 4.1%-4.2%, vs. the 4.4% in the December report, so I am expecting at least a small further decline ahead.

The bottom line is that jobless claims continue to forecast a growing economy in the months ahead.
Posted by at

Wednesday, January 14, 2026

December existing homes sales add evidence to the “green shoots” thesis for sales, while inventory still has a long ways to go

 

 - by New Deal democrat


Although the government shutdown is long over, the most recent government housing updates have been for October, I.e., two months stale. Thus the NAR’s existing home sales report has temporarily become among our best look at housing sales, prices, and inventorythe housing market. 

As per my context all this year, after the Fed began hiking rates in 2022, mortgage rates also rapidly rose from 3% to the 6%-7% range, where they have remained ever since. Since sales follow mortgage interest rates, existing home sales rapidly declined to 4.0 million annualized, and have remained in that range, generally +/-0.20 million for the past 3.5+ years. Since September, mortgage rates have been at the bottom of their 3+ year range, and in December existing home sales predictably reacted, breaking out of that range to the upside, at 4.35 million units annualized, the highest number since March 2023:

https://tradingeconomics.com/united-states/existing-home-sales 


In the past several years I have been looking for the new and existing homes markets to rebalance. Existing home inventory has been removed from the market for over 10 years (likely due in part to absentee rental owners buying increasing chunks of inventory), and really accelerated during the pandemic. This caused an acute shortage of houses for sale, which in turn led to bidding wars among buyers and a spike in prices.

A rebalancing of the market more than anything would require an increase in inventory at least to pre-COVID levels, and a deceleration of price increases, or even outright decreases. Which means that the level of sales themselves was far less important than what the median price for an existing home and inventory are telling us about the ongoing rebalancing of the housing market.

The secular decline in inventory reached a nadir in 2022. This series is not seasonally adjusted, so it must be looked at YoY. In December inventory declined sharply, as it does every year, to 1.18 million, exceeding every December level since 2019, when its level was 1.39 million:
 
 

Since inventory was typically in the 1.7 million to 1.9 million range before the pandemic, the chronic shortage still exists, although it is very slowly abating.

For inventory to fully adjust, so must prices. As shown in the below graph, the median price of an existing home rose about 45% between July 2019 and July 2022 and another 5% from there through July of this year, before seasonally declining:

 https://tradingeconomics.com/united-states/single-family-home-prices 



With seasonal adjustments are not made, my rule of thumb is that a peak (or trough) occurs when the YoY% change is less than half of its maximum change in the past 12 months. Here are the comparisons in the past 12 months:

December 6.0%
January 4.8%
February 3.6%
March 2.7%
April 1.8%
May 1.3%
June 2.0%
July 0.2%
August 2.2%
September 2.1%
October 2.1%
November 1.2%
December 1.4%

While YoY price comparisons have crept up since July, they remain well below their past 12 month peak of 6.0%, so the fair conclusion remains that, if we could seasonally adjust, house prices are softer than they were last spring.

My last report on existing homes sales concluded that “the rebalancing of the [new vs. existing housing] market is a long slow slog. Yesterday’s existing home sales report is another data point of very slow progress towards that rebalancing.” The December report adds evidence to the “green shoots” thesis for sales on top of the housing construction and new home sales data we got earlier this week. But the rebalancing remains a long slog, with the pandemic era low inventory almost totally reversed, but yet far below the 2016-18 levels.


Posted by at

Monthly retail sales sharply higher in November, but flagging YoY real sales spell further trouble

 

 - by New Deal democrat


Real retail sales, one of my favorite broad-economy indicators, was updated through November this morning, making only one month stale. This, along with real personal spending, is one of the two most important indicators which have been missing, as we know the jobs and real income have been stagnant, but in terms of important expansion vs. recession metrics, what of sales and purchases?

Let me cut to the chase: in terms of nominal spending, it confirmed the strength we have seen in the weekly Redbook and daily restaurant reservations reports beginning in November. Specifically, in nominal terms retail sales rose 0.6% in November after -0.1% downward revisions for both September and October. In real, inflation adjusted terms, however, the story is different.

Real retail sales are more problematic, in part because there was no number for October, and November’s reading was marred by the shutdown kludge, particularly for shelter. With those important caveats noted, in November real retail sales were higher by 0.3% compared with September, and up 0.6% YoY. The below graph, through September, shows YoY real retail sales (blue) and the similar measure of real spending on goods (gold ), with the most recent reading of each subtracted so that it =0:

 https://fred.stlouisfed.org/graph/fredgraph.png?g=1QtjV&height=490 

If you believe, as I do, that the shutdown shelter kludge removed about 0.2% from consumer inflation, that becomes a tiny 0.4% increase YoY, the smallest such gain since October 2024. Also, recall that real personal spending has not been updated yet beyond September.


Going back 75 years, a decline in YoY real retail sales has almost always meant a recession (but both the exception in 2023!). Neither they nor real personal spending on goods are negative as of their last readings,, but real retail sales have decelerated sharply since their YoY peaks in early spring. Should the trend continue, they could be negative YoY in their December or January reports.

Finally, because consumption leads employment, here is the update of YoY real sales (/2 for scale) together with employment (red), updated through the December jobs report:


[Note that, since I can’t show the November real retail sales “dot,” you’ll just have keep in mind that there was further YoY deceleration] This sharp deceleration in YoY growth in consumption forecast the slide in employment, and suggests that the jobs reports in the next several months will get no better.



Posted by at

Tuesday, January 13, 2026

October new home sales: also pre-recessionary, also with signs of possible “green shoots”

 

 - by New Deal democrat


New home sales were updated for the second time since the shutdown, with data only through October - i.e., stale. The silver lining is that this is a long leading indicator, so it remains of value.

Normally I save inventory for last, but in view of the importance of new homes for sale (red in the linked graph below) in providing housing’s final pre-recession signal, here is that number compared with new houses sold (blue, right scale):


For sale inventory was unchanged month over month, and only up 1.7% YoY. As I wrote yesterday, once that has gone negative YoY, it has typically signaled the imminence of a recession. In that regard, if inventory simply has remains unchanged through December, it will have turned negative YoY.

The silver lining is that, like housing permits, actual single family home sales turned higher in September and October, down only -0.1% monthly in the latter month, but higher by 1.8% YoY suggesting that lower mortgage rates may be laying the groundwork for a recovery. As per usual, I caution that this series is very noisy and heavily revised.

Finally, as I typically note, prices follow sales with a lag, and that continued to be true as the median price for a new home was down -8.0% YoY:


Note that this series is not seasonally adjusted; hence the focus on the YoY change.

In short, much like housing permits, starts, and units under construction, this stale data looks very pre-recessionary, but also with some signs of “green shoots” thereafter.


Posted by at

December consumer inflation: a return to pre-shutdown trends and still affected by the shutdown shelter kludge

 

 - by New Deal democrat


We finally got our first “regular” CPI report since September this morning. Caution is still warranted, however, because the October-November kludge is still present in the base from which December’s monthly change was calculated. But the bottom line is that the series’ all reverted to their pre-shutdown trend, with the headline number up 0.3% and core inflation up 0.2%, but also with the shelter kludge still affecting the headline YoY comparisons of 2.7% and 2.6% respectively. 

As per my usual practice for the past several years, let’s start with the YoY numbers for headline inflation (blue), core inflation (red), and inflation ex shelter (gold), which was only up 2.4%:


The good news is that CPI less shelter decreased -0.2% in December,  and CPI ex shelter was the lowest since July, suggestion significant *disinflation.* Notably the previous uptrend in non-shelter inflation and a smaller but notable increase in headline inflation, with no deceleration in the past 12 months flat YoY core inflation, was clearly broken by the shutdown kludge. If shelter had increased its previous 0.3% monthly during those two months, both headline and core consumer inflation would be over 3%. 

Nevertheless, shelter inflation has decelerated YoY per the latest measure, down to a 3.2% increase, with rent up 2.9% and Owner’s Equivalent Rent up 3.2%, the lowest increase since September 2021 except for last month’s kludge:


As usual let’s compare that with the YoY% changes in the repeat home sales indexes, which lead by about 12-18 months (/2.5 for scale), to CPI for shelter (red). YoY home price increases are near or at multi-year lows, each at roughly 1.5%, and shelter inflation has followed. The graph linked to below includes several years before Covid to show that this is well within its 3.2%-3.6% range during the latter part of the last expansion:


Needless to say, this is not only good news, but because of the leading/lagging relationship, we can expect further deceleration in the shelter component of inflation during this year.

Another bright spot is that gas prices declined -0.5% for the month, resulting in a -3.4% YoY decline, which is welcome news to consumers:

 https://fred.stlouisfed.org/graph/fredgraph.png?g=1QqH9&height=490 

Let’s take a look at a few other areas of interest.

First, new car prices continue to be largely unchanged, flat for the month and up only 0.3% YoY, while used car prices reversed their shutdown increase, declining -1.1% in December and up only 1.6% YoY. The graph linked to below shows the post-pandemic trend by norming both series to 100 as of just before the pandemic:


Every month I check the detailed breakout for “problem children,” I.e., sectors that have increased in price by 4% or more YoY. This month included several minor irritants including non-alcoholic beverages and tobacco, as well as fuel oil. Another recent problem child for inflation has been transportation services, mainly vehicle parts and repairs as well as insurance. Of these, only repairs and maintenance are still problematic, as while declining -1.3% for the month, they remain higher YoY by 5.4%:



Finally, electricity prices have also become a significant problem, likely a side effect of the building of massive data centers for AI generation. These declined -0.1% in December, but on a YoY basis are up 6.7%, the highest increase since 2008 except for the shutdown kludge and the immediate post-pandemic inflation:


As I wrote last month, this has already created a backlash, and I expect that backlash to intensify.

In summary, on a monthly basis December consumer inflation was relatively tame, with shelter cost increases slowly abating and only a few other problem children. I would continue to treat both headline and core YOY numbers with extra caution, since they both remain affected by the situation with shutdown shelter kludge. More likely YoY inflation is roughly steady in the 3% range, above the Fed’s target and with employment growth dead in the water.


Posted by at

Monday, January 12, 2026

September and October housing construction consistent with government shutdown recession, but also the possibility of “green shoots”

 

 - by New Deal democrat


On Friday housing permits, starts, and units under construction were finally reported for the first time in four months, since September’s report for August. The bad news is that the report only updated through October, so we are still two months behind. The very qualified good news is that, since housing is a long leading indicator, even with this lag the report still gives us insight into where the economy is likely to go in the next eight months.
 
When I last updated this information in September, I wrote that “a puzzling relationship this year has been that the housing data has been classically recessionary for a number of months, and yet the economy has not rolled over.” That May no longer be true, in that the government shutdown may have caused at least a brief economic contraction, but we won’t know that - even even if it was probable - until real sales and consumption are updated for last autumn later this month.


So let’s start by reiterating the basics: mortgage rates lead sales, which in turn lead prices, which in turn lead inventory.

Mortgage rates have fluctuated in a range between just over 6% to 7.6% in the past 3+ years (red, left scale in the graph linked to below), and housing permits have similar been rangebound between 1.330 and 1.620 million annualized (blue, right scale) over that same period:


In the past several months, interest rates have been near the bottom of their range, and permits responded in September and October by increasing from their August post-pandemic low. 

In more detail, total permits increased 82,000 to 1.412 million during that two month period. Single family permits, which convey the clearer signal, increased 18,000 to 876,000. Meanwhile the much noisier and slightly lagging housing starts declined -45,000 to 1.246 million units, their lowest number since the pandemic:


The decline in starts is unsurprising, since permits made their post-pandemic low in August, and as stated above, starts tend to follow within several months.


Even with these gains, permits and starts remain in territory below their peaks sufficient to be consistent with a recession. On the other hand, all three measures are down less than -10% YoY, where in the past it has taken a more severe decline of greater than -10% to be consistent with with a recession:


Let’s turn next to the number of housing units under construction. As I have written many times in the past several years, it is the best “real” measure of the economic impact of housing (blue in the graphs below). In September and October they remained almost exactly unchanged from their post-pandemic low in August, up only 2,000 units, but still down -23% from their peak:



The above graph shows how they have followed single family permits (red), as expected. More often than not in the past by the time a decline in units under construction had declined by as much as they did in August - and September and October - a recession had already begun. The only two exceptions were the late 1980s, where the pre-recession decline was -28.2%, and 2007, where the pre-recession decline was -25.6%. 

Now let’s update housing units under construction with the typical final shoes to drop before recessions, houses for sale (gold) and residential construction employment (red), in comparison with units under construction, all normed to 100 as of their respective post-pandemic peaks. Both the number of employees in residential construction and new one family homes for sale peaked in March and have declined almost uniformly since:



On a YoY basis, with the exception of 1974 and the COVID recession, houses for sale and (once available) employment in residential construction had turned down YoY before the recessions had begun:

 
As of their last update for August, houses for sale were still higher by 4.0% YoY, but as of last Friday’s employment report for December, residential construction employment is now down -0.1%. 

In September I concluded that August’s report was “very much recessionary, although in some YoY comparisons, I would expect further damage before the actual onset of one. But that could easily occur within the next four to six months.” Indeed, per my last paragraph above, to the extent available, some of that has already happened. Additionally, as I pointed out several weeks ago, employment, industrial production, and real manufacturing and trade sales, as of their last reports, were all below their respective spring and summer peaks. On the other hand, the fact that permits did rebound for two months and units under construction did not decline further argues for the possibility of a bottom in the housing market and the proverbial “green shoots.”

The big missing piece remains real personal spending, and we won’t know anything about that even for autumn until later this month.


Posted by at
Subscribe to: Comments (Atom)