Stale data watch: manufacturers’ new orders soared in November — more evidence of AI data center building?

 

 - by New Deal democrat

With another likely government shutdown looming at the end of this week due to DHS funding, we are still playing catch-up from the last one that ended in November. This morning’s edition of stale data was durable goods manufacturing for November.

One of the stories of the latter part of last year is that manufacturers appear to have adjusted to the increased tariff regimen imposed by Washington. That was apparent in this morning’s data, as new orders for durable goods (blue in the graph linked to below) increased a strong 5.3% in the months, while core capital goods orders (red), which convey more signal and less noise, increased 0.7%:

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

This is in stark contrast to new orders for consumer goods (gold), which, while they have not been decreasing, completely stalled over the past 2 years. 

Note that this is in contrasst to the ISM manufacturing index, which has been in contraction since February of last year:

 https://tradingeconomics.com/united-states/business-confidence 

Since the ISM metric is a diffusion index, meaning that the number of respondents reporting contraction have outnumbered those reporting expansion, this suggests that the increase in new orders for manufacturing is concentrated in relatively few industries. The biggest driver, per the report, appears to have been transportation equipment, although we know that the trucking industry is suffering greatly. Beyond that there is little information, but my suspicion is that we are seeing a byproduct of the big build-up in AI-related data processing centers.

Weekly Indicators for January 19 - 23 at Seeking Alpha

 

 - by New Deal democrat

My “Weekly Indicators” post is up at Seeking Alpha.

The trends from last year are continuing so far this year. On the plus side, initial jobless claims are very low while stock prices continue near all time highs, and consumer spending is on a tear. On the minus side, the US$ started sliding again, and commodities, and in particular precious metals, are soaring - a bet against the US and its economy on a global scale.

As usual, clicking over and reading will bring you up to the virtual moment as to all these trends, and bring me a penny or two for collecting, collating, and presenting the data in organized fashion.

Economic cycle indicators update: the Autumn Shutdown Stall

 

 - by New Deal democrat

Now that we have the very important personal income and spending data through November, here is a look at the widely acknowledged monthly indicators that the NBER has highlighted in determining whether the economy is continuing to expand, or is contracting.

The below link includes nonfarm payrolls (blue), industrial production (red), real personal income excluding government transfers (gold), and real manufacturing and trade sales (dark green, through October). Also included are manufacturing production (light red) and real disposable personal income (yellow). All are normed to 100 as of last July: 

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

It’s fair to say that all of these stalled through October, with only real personal income higher by 0.1%; and to some extent through November, with industrial production higher by 0.1% and real personal income higher by 0.2%. At the same time, there clearly was not any significant downturn that would warrant a recession declaration.

So call it the Autumn Shutdown Stall.

Through December, we only have the jobs data, which continued to stall, and industrial production, which rose to 0.4% higher than July, primarily on utilities (likely data center construction for AI purposes). Although it isn’t included in the above, as indicated yesterday real personal spending continued to increase at a solid pace through November. An update for manufacturing and trade sales is scheduled for next week, but personal income and spending for December isn’t scheduled to be released until late February.

No mini-recession during the government shutdown, as consumers raided their piggy banks to spend

 

 - by New Deal democrat

The news from this morning’s personal income and spending report for October and November was almost all good - with one major exception.

To cut to the chase, real income rose slightly, but real spending rose substantially. Which, if you are following basic math, means that consumers dipped into their savings in a very significant way in order to make it happen - which is not a good sign for the future.

Let me start with that last statistic. The personal saving rate for Americans declined -0.3% in October and another -0.2% in November, down to 3.5%. As shown in the graph linked to below, this is the lowest saving rate in the past 60+ years outside of parts of 2022 and 2005-07, plus two months after the 2001 terrorist attacks:

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

Nominally, personal income rose 0.1% in October and another 0.3% in November. However, because the PCE deflator increased 0.2% in each month, the net result was a real increase of only 0.1% at the end of the two months. Further, once we exclude government transfers, the numbers for the two months are 0.0% and 0.1%, for a similar tepid increase:

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

This is only a 0.2% on net since July, and it remains -0.2% below its peak in April. By itself, this statistic would be consistent with a recession starting last spring.

But the story is completely different as to spending. Nominally that rose 0.5% in each month, which after accounting for the 0.2% deflator for each month, means real personal spending rose 0.3% in each month.

Further, the spending was not limited to services, which tends to rise even during most recessions. Rather, the figures for both total goods spending and durable goods spending were also strong, at 0.3% and 0.% respectively in October and 0.6% for both in November:

https://fred.stlouisfed.org/graph/fredgraph.png?g=1QMZ9&height=490 

Finally, although I won’t bother with the graph, the 0.2% increase in the deflator for each month means that the YoY% increase for the PCE deflator was 2.8%, in line with most of its readings over the past 12 months. In other words, so measured inflation is not decelerating any more, but is has not been accelerating either.

The takeaway for this very important set of statistics - virtually equal in my view to the monthly jobs report - is that there was no mini-recession during the government shutdown, but that was largely due to consumers reducing their savings (likely another manifestation of the “wealth effect” of sharply rising stock prices) in order to spend on both goods and services. But this doesn’t mean that the economy was doing well. Rather, it was figuratively keeping its head above water as consumers exposed themselves to more risk from an adverse shock by raiding their piggy banks.

More evidence of a positive “regime change” in jobless claims: is it tied to the immigration crackdown?

 

 - by New Deal democrat

We will get the very important personal income and spending updates for the government shutdown months later this morning, which should finally tell us whether or not there was a mini-recession during that time. In the meantime, let’s take our regular weekly look at jobless claims, as to which, to reiterate, there appears to have been a positive change of regime starting half a year ago - and which continued this week.  

To wit, initial claims rose 1,000 to 200,000 for the week, still among the lowest numbers in the past 50 years, The four week average declined -3,750 to 201,500, except for several weeks in 2022 and one week in 2024, the lowest number since the 1960s (tied for one week in 2019). Continuing claims also declined -26,000 to 1.849 million, the 2nd lowest number since last April:

 https://fred.stlouisfed.org/graph/fredgraph.png?g=1QMv0&height=490 

The above graph includes the last three years, since there still appears to be unresolved post-pandemic seasonality, whereby even after adjustment, claims have risen during the first half of the year, and declined in the second half towards a nadir in January.

This issue is even more apparent in the YoY% changes, in which initial claims were down -9.9%, the four week average down -5.5%, and continuing claims down -2.1%:

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

Initial claims and the four week average are virtually identical to what they were two and three years ago. In other words, the issue may be a relative increase in claims one year ago. It has been suggested that this may have to do with the wildfires in Los Angeles last January - but claims only spike for one week (January 18) one year ago in California, so that does not appear to be the main explanation.

More importantly, as I’ve noted several time in the past month, initial claims have been negative YoY most weeks since the end of last June. This suggests that there is more going on: a “regime change” as I’ve called it. My best speculation, and it is only that, is that it has to do with the push for massive deportations that has been undertaken during that time. If many thousands of immigrants (including those legally here) have quit or simply stopped showing up for work, then that would mean a large decline in the number of new unemployment claims.

In any event, if the big decline in claims lasts another week or two, that will put it well beyond the range for which Holiday seasonality could be an explanation.

Finally, this is also a positive for the unemployment rate going forward for the next month or two, as shown in the graph linked to below:

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

The sharp decline in jobless claims since last summer’s peak strongly suggests that the 4.5% unemployment rate in December was also the peak for that number. A rate steady at 4.4% or even moving down towards 4.0% appears more likely.

Stale data watch: construction spending for October — more “green shoots”?

 

 - by New Deal democrat

Originally housing permits and starts through December were supposed to be updated this morning, but that has now been put off another week. In the meantime, we did get some data, albeit stale, about the important leading sector of construction, in the form of the construction spending report through October. And it added to the very small patch of evidence suggesting that some “green shoots” may be forming.

To wit, as shown in the graph linked to below, in October nominally total construction spending rose 0.5%. More importantly, the long leading sector of residential construction spending rose 1.3%. Since the cost of construction materials in the last PPI report declined -0.2%, in real terms spending rose 0.7% and residential spending rose 1.5%:

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

In absolute nominal terms, both series rose to close to 10 month highs:

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

In real terms, residential construction also made a 9 month high:

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

Keep in mind that construction spending is one of the latter housing series to turn, although it generally turns before housing units for sale and residential construction employment. Across almost all the more leading housing data, the trend in the past few months has been a leveling out or even some small improvement. Even prices (which follow sales) may be leveling out.

All of this is downstream of the 3 year lows in mortgage rates. Of such small things are “green shoots” made (subject, of course, to longer term interest rates blowing out to the upside due to T—-p’s blowing up NATO).