Construction spending continues contraction, amplifying yellow flag caution from manufacturing

 

 - by New Deal democrat

I concluded last month’s post on construction spending by writing “Putting this report together with this morning’s other report on manufacturing from ISM, it appears the goods-producing part of the economy as a whole is very slightly contracting. It will be interesting to see if this is reflected in a decline in goods-producing jobs in Friday’s report.”

If anything, this morning’s construction spending report suggests a contraction that is a little less “slight,” as total construction spending (blue in the graph below) declined -0.7%, while residential construction spending (red, right scale) declined -0.5%:

Further, since on a nominal basis both series peaked exactly 12 months ago, on a YoY basis as well as on an absolute basis from peak, they have declined -3.5% and -6.5%, respectively:

Adjusting by the cost of construction materials, which again rose last month, the declines from peak are -8.9% and -10.4%, respectively:

If construction spending has declined more severely from peak, and manufacturing is in contraction as well, that amplifies somewhat the conclusion that the entire goods-producing sector of the US economy is in a downturn. In which regard, here is the latest update to industrial (blue) and manufacturing (red) production, indicating slight declines from peaks several months ago:

In answer to the question posed at the top of this post, last month there was indeed a slight -5,000 decline in goods-producing jobs. On Thursday we will find out if that continues for a second month.

Preliminary economically weighted ISM average for June continues in “recession watch” territory

 

 - by New Deal democrat

As usual, we start out the month with reports on both manufacturing and construction. Ill post separately on construction. Additionally, the May JOLTS reports was posted, but I’ll discuss that tomorrow. So let’s start with the ISM manufacturing report, a recognized leading indicator for the past 60+ years, although of diminished importance since the turn of the Millennium (it was in deep contraction both in 2015-16 and again in 2022 without a recession occurring).

To recap briefly, any number below 50 indicates contraction. The ISM itself indicates that the number must be 42.5 or less to signal recession. For forecasting purposes, I use an economically weighted three month average of the manufacturing and non-manufacturing indexes, with a 25% and 75% weighting, respectively. After both reports were posted one month ago, I indicated they justified a “recession watch.” 

This morning’s report confirms that. While the headline number for June rose 0.5 to 49.0 (still contractionary), the more leading new orders subindex declined -1.2 to 46.4.

Here is a look at both the total index (gray) and new orders subindex (blue) for the past ten years:

Note that both remain better than they were in 2022-23.

Hare the last six months of both the headline (left column) and new orders (right) numbers:

JAN 50.9  55.1

FEB  50.3  48.6

MAR 49.0. 45.2

APR 48.7. 47.2

MAY 48.5. 47.6

JUN. 49.0. 46.4

The current three month average for the total index is 48.7, and for the new orders subindex 47.1. 

As I indicated above, for the economy as a whole the weighted index of manufacturing (25%) and non-manufacturing (75%) indexes is more important. In the non-manufacturing report, the average of the last two months for the headline and new orders numbers has been 51.8 and 49.4, respectively. While the economically weighted headline number remains slightly above 50, at 50.4, the new orders average is 49.0.

In short, pending the release on Thursday of the ISM non-manufacturing report, for the second month in a row the new orders average is forecasting economic contraction in the next few month. Which means that, as of today, the “recession watch” forecast signal continues in place.

Pay close attention to real personal spending on goods

 - by New Deal democrat

 This week data arrives in two batches: a smaller batch (ISM manufacturing and construction spending) tomorrow, and a huge tranche (nonfarm payrolls, jobless claims, ISM services, and factory orders) on Thursday. Which means I might take Wednesday off, and/or delay reporting on some of Thursday’s data until Friday.

In the meantime, let me take a deeper look at Friday’s report on personal spending. That’s because as I’ve written before, spending on goods turns down before spending on services. In fact, spending on services often sails right through recessions without ever turning down at all. Further, spending on goods tends to turn negative before the recession actually hits. And of course since consumer spending is roughly 70% of the economy, it also has a big impact on GDP.

First let me show you the historical YoY% change in real spending on goods (red) and services (blue), averaged quarterly to cut down on noise:

Except for very shallow recessions, real spending on goods has always turned negative YoY, whereas before COVID real spending on services only turned negative during the severe Great Recession. And since real spending on goods is much more volatile, if it is decelerating to a lower level than that on goods, it is a frequent harbinger of recession.

Now here is the post-pandemic look. I’ve broken this out monthly just because it is much easier to see what is going on:

Real spending on goods only turned negative YoY one year exactly after the large 2021 stimulus, which led to a lot of immediate spending on goods. Even with the big miss on Friday, it was still positive YoY by 2.5%.

But as I have often noted, a measure peaks or bottoms before the YoY measure turns positive or negative. Thus, where we have seasonally adjusted data, that is the better way to took for leading turning points.

Since real personal spending on goods and services is seasonally adjusted, let’s look at the % change in each on a quarterly basis. I’ve broken down the 60 years of data before the pandemic into 3 graphs so the important markers are easier to see.

Here is 1960-80:

1981-2000:

And 2001-2019:

Again, note that real spending on services rarely turns down, even during recessions, although usually its growth does declerate below 1% annualized during the quarter just preceding or starting the recession.

Further, if real spending on goods is positive, with exactly one exception in 60 years it either meant an expansion was ongoing, or we were close to the end of a recession.

More often than not, even a negative quarterly number for real spending on goods did not signal a recession either. But if we remove all such cases were quarterly real spending declined less than -1%, we get a much more reliable signal, albeit not perfect. About 50% of the time sharper downturns indicated an oncoming recession, and about 50% just a slowdown. If growth in real spending on services was also decelerating sharply, even if still positive, it almost always meant recession.

Now let’s look at the quarterly post-pandemic data, beginning with Q2 2021:

Although Q2 and Q3 in 2021 were negative, they were a reaction to the huge front-loading of spending in Q1 (which would have dwarfed all the other readings). Only one quarter - Q4 of 2022 - came in with a downturn of more than -1%, a false positive. Q3 and Q4 of 2024 again likely featured front-running. The payback was in Q1 of this year, when the quarterly increase for both goods and services was almost exactly 0.

Now here is the monthly look at the last year:

In the first two months of Q2, total real spending has declined by -0.8%, while services has been basically unchanged. If there is a further decline in June, based on the above discussion that would likely trigger a “recession watch” signal.

Weekly Indicators for June 23 - 27 at Seeking Alpha

 

 - by New Deal democrat

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

There have been a number of reversals since April, most notably the stock market going from a 12 month low to a new all-time high, but also the front-running by consumers apparent in the weekly Redbook report has also reversed, from over 7% YoY to 4.5%, one of the lowest readings of the past 12 months.

Whether this is more than just a temporary reversals will likely play out over the next two months.

As usual, clicking over and reading will bring you up to the virtual moment as to the state of the economy, and bring me a little lunch money.

May personal income and spending: consumer payback for Tariff-palooza! is a B!t©h

 

 - by New Deal democrat

The last significant data for the first half of 2025, personal income and spending for May, was released this morning. It was the first month that reflected the impact of Tariff-palooza!, and boy howdy was it impacted. Not a single metric was positive. One metric was unchanged; everything else was negative. Let’s take a look at the carnage.

Nominally personal income declined -0.4%, and personal spending declined by -0.1%. Since the PCE deflator increased 0.2%, real income was down -0.6, and real spending down -0.3%. Here is this month’s update of real personal income and spending normed to 100 since the onset of the pandemic (as are all other graphs below except for the personal saving rate):

Typically real spending on goods declines before recessions, while real spending on services has increased throughout all but the most severe of them. In May real spending on goods declined by -0.6%, while that for services - the sole relative “bright” spot in this morning’s report - was unchanged:

Further, ant least one important historical recession model posits that evidence that spending on durable turns down before spending on non-durable goods. This month both were abysmal, as the former metric turned down by -1.8%, and the latter by -0.3%:

While on a YoY basis none of the above metrics have broken a trend, it is noteworthy that all of them show a sharp slowdown in growth since last December, from a meager 0.1% increase in real spending on services to a sharp -0.9% decline in spending on durable goods. Some of this can be put down to the effects of front-running earlier this year, but if the trend continues that could indicate a real turning point.

Next let’s take a look at the personal saving rate. Generally, as expansions continue, consumers become more aggressive with their purchasing, and then more cautious immediately in advance of (and partially the cause of) recessions. In May this declined -0.4% to 4.5%, but still well above its low of 3.5% last December:

Again, there’s no indicated break in trend, although it is important to note that such a rate remains below any reading below 1999, and the average between 2000 and 2019 was 5% (not shown).

Finally, there are several important coincident indicators used by the NBER in recession dating in this report.

The first is real personal income less government transfer payments. This declined -0.1%:

The second is real manufacturing and trade sales, which are calculated with a one month delay. In April they declined -0.4%:

Again, neither of them show a clear break in trend, although by the time such a break would be apparent, almost by definition a recession would have already begun.

Last month I concluded that “because of the tariff situation, forecasting based on this report is particularly fraught. What we can say is that the consumer portion of the US economy remained in expansion through April”, and that it would be important to see if the initial evidence of an end to consumer front-running of tariffs would be continued or amplified in May.

This morning we got a clear answer, as the report showed ample evidence of payback, as consumers cut back on spending of almost all sorts, and even spending on services turned flat. Perhaps more concerningly, real incomes declined, even after we account for transfer payments like Social Security. As I wrote above, whether this might mark an actual turning point vs. simple payback for the front-running of tariffs earlier this year will have to wait on another month or two of data. For now, the important point is that in May all of the leading and coincident indicators of personal finance turned down.

Jobless claims indicate employment market continues to weaken, but still not recessionary

 

 - by New Deal democrat

Jobless claims continue to tell us two things: (1) the jobs market continues to slowly weaken, but (2) it is not recessionary.

This week I’ve changed my graphing scheme slightly, to emphasize the less noisy four week moving average of initial claims, to better show the residual post-COVID seasonality, and to put the recent increase in continuing claims in better context.

With that said, initial claims declined last week by -10.000 to 236,000, and the four week average declined -750 to 245,000. With the typical one week delay, continuing claims rose another 37,000 to 1.974 million, its highest level in almost 3.5 years (see extreme left in graph below):

Also note that beginning with the end of 2022, we have seen a pattern where initial claims rise into the summer, then fall back into the winter, a pattern which has continued this year so far.

This residual seasonality makes the YoY% comparisons, which are more important for forecasting purposes, all the more salient.So measured, initial claims are up 1.3%, the four week average up 4.3%, and continuing claims up 7.0%:

This is well within the trend of the YoY comparisons averaging 5.0% +/-5% which we have seen since last October.

Finally, since the 4th of July is Friday next week, the June employment report will be released Thursday, which means this will be our last advance look at what jobless claims are suggesting about the unemployment rate. Here is the YoY% change in both measures of claims as well as that of the unemployment rate:

This suggests about a 4% (percent of a percent) increase in the unemployment rate YoY in the next several months.

Since the unemployment rate was 4.1% last June, rising to 4.2% for several months thereafter, this suggests that the unemployment rate is likely to rise to 4.3% or possibly even 4.4% in the next several months:

We’ll find out a week from today. In the meantime, the message - expecially from continuing claims - is that while new jobs are harder to find, layoffs are not increasing that significantly. Remember that my model requires a 10% YoY increase in jobless claims just for a recession “watch,” let alone a “warning.”