Saturday, March 25, 2023

Weekly Indicators for March 20 - 24 at Seeking Alpha


 - by New Deal democrat

I’ve neglected to put this up for the past several weeks, but by now you know where to find my latest Weekly Indicator post at Seeking Alpha.

Probably unsurprisingly, in the week after the Silicon Valley Bank failure, just about every financial stress indicators suddenly spiked. In other words, credit conditions, which had already tightened by the end of last year, tightened a lot more in the past several weeks.

Like I wrote yesterday, there are only a couple of things still holding up the economy from falling into recession. As usual, clicking over and reading will bring you up to the moment, and reward me a little bit for putting in the effort.

Friday, March 24, 2023

There is now only one significant manufacturing datapoint that is not flat or down - but it’s the one the NBER relies upon


 - by New Deal democrat

I am increasingly of the opinion that at the moment, the only two economic data series that are important are nonfarm payrolls and the personal consumption expenditure deflator. That’s because almost every other important metric of the economy is either flat or declining. But payrolls keep chugging along (as evidenced by yesterday’s initial claims report showing that layoffs are figuratively non-existent). And the PCE deflator, which covers a broader spectrum than the CPI, keeps helping two coincident indicators important to the NBER, namely real personal income and real manufacturing and trade sales, remain in the plus column. 

This morning’s report on manufacturers’ new orders for durable goods for February is evidence of that. The broad measure (blue in the graphs below) declined just shy of 1%, while the core measure (red), which is less noisy, increased by 0.2%:

Both are below their peaks in December and August 2022, respectively. In the past few months the broad measure has been trending down, while the core measure has been basically flat.

New orders for durable goods have long been recognized as a leading indicator. Here’s the longer term view for the past 25 years:

Their record certainly isn’t perfect. The core measure continued to rise well into the Great Recession, and both declined sharply in the industrial recession of 2015-16 (that was not an actual recession because consumers kept spending merrily away, and so employment kept rising as well).

Anyway, in the below graph I’ve also added the value of manufacturers total actual shipments, and shipments for “core” capital goods, as well as *nominal* manufacturers’ sales, all normed to 100 as of October 2022:

The slowdown in growth, followed by an actual downturn in the broad measures of orders and shipments, is apparent, as is the flattening in both “core” measures, up only 0.1% and 0.2% since October, respectively.

Nominally, manufacturers sales (gold) through January are also down -0.8% since October. This is where the PCE deflator comes in. Because while there is no breakout of “real” manufacturers sales only, real manufacturing and trade sales (which also includes wholesale and retail sales, and is relied upon by the NBER) was up 0.6% as of its last reading in December.

In other words, the entire panoply of goods production and distribution in the US economy (including industrial production as well) is either flat or down - with the exception of those two measures (real personal income and real manufacturing and trade sales) which take into account the big deflation in producer prices since June, and the PCE deflator. More on that next week, as we wait for the February PCE report next Friday.

Thursday, March 23, 2023

New home sales for February increase; likely bottomed last July


 - by New Deal democrat

Most of what you probably read elsewhere focuses on new home prices, which after finally declining -0.7% YoY in January, rebounded to +2.5% YoY. As is usual, prices  follow sales YoY with a considerable lag (note since prices are not seasonally adjusted, this is the right way to make the comparison):

In fact if you’ve been reading me and following my rule of thumb, the peak occurred  months ago, once the YoY gains had decelerated by over 50%.

But the most important news was actually in the seasonally adjusted sales, which at 640,000 annualized increased 7,000 from a seriously downwardly revised (by -37,000) January. Big deal, you say? Here’s a graph of the last several years:

While revisions can still be made to the last several months, it is apparent that the bottom for new home sales was last July, at 543,000 annualized. There has been an almost consistent monthly increase since.

This joins existing home sales, which likely bottomed in December; and housing permits and the three month average of starts, both of which possibly bottomed in January. As I’ve written many times before, while new home sales are very noisy and heavily revised, they are frequently the first housing data to turn. And it appears they have. 

Nope; nobody is still getting laid off


 - by New Deal democrat

Initial jobless claims declined -1,000 to 191,000 last week, while the more important 4 week moving average declined 250 to 196,250. Continuing claims, with a one week delay, rose 14,000 to 1.694 million:

All of these remain excellent numbers. In particular, the higher numbers at the left end of the graph would have been considered excellent at any previous time in the past 40 years. So the paradigm that almost nobody is getting laid off remains intact.

The YoY comparisons are against the best numbers of last year, so they look comparatively bad. Weekly claims are up 15.1% YoY and continuing claims are up 9.8%. The more important 4 week average is up 8.6%, still below the 10% threshold for even a yellow flag, and bear in mind that it is only important if it lasts a month or more at that level:

This suggests that the unemployment rate in the March payrolls report will be unchanged or close thereto, and at very least there will not be a “bad” headline jobs number.

Wednesday, March 22, 2023

Updating 3 high frequency indictors: no recession yet, but no paucity of yellow flags

- by New Deal democrat

Aside from the Fed’s rate decision which will be announced this afternoon, it’s a slow economic news week. In general, the punditry which I read seems to be settling on a consensus that we are going to manage to have a soft landing. With the exception of jobs and payrolls, the rest of the indicators I track in sequence continue to indicate a recession is near. Let me update three high frequency indicators I’ve been paying particular attention to.

1.  Redbook consumer spending

This is a weekly update of same store spending, measured nominally YoY. The trend here is pretty self-explanatory:

YoY consumer spending was holding up in double-digits until the middle of last summer. Since then the trend, albeit with some waxing and waning, has been pretty consistent. The average of the last 3 weeks has been just below 3% YoY.

For comparison, here is monthly nominal retail sales YoY since last March:

The trend is similar, but the decline in Redbook has been considerably more pronounced. If the monthly series follows, March nominal retail sales will decline further YoY. And as indicated that’s before consumer inflation is taken into account, which as of February was 6.0% YoY.

Further, if the trend continues, even nominally Redbook will be negative YoY by midyear.

2. Temporary employment

Temporary employment is a leading sector of the jobs market. The American Staffing Association has been posting a “Staffing Index” since 2006. Typically the Index slowly increases during the year, with major seasonal fluctuations around the 4th of July, Thanksgiving, and Christmas-New Year’s. Here’s what it looks like since January 2022, and the onset of the Great Recession, 2006-07, for comparison:

The 6% YoY downturn in February through early March this year has been the biggest since the series was begun. That being said, there was a similar downturn in late 2015 (not shown) without the economy coming close to recession.

For comparison, note that there has been a similar YoY downturn in termporary employment in the official payrolls report:

The last few weeks suggest this downturn in the monthly jobs report will continue.

3. Payroll tax withholding

This is a decent coincident proxy for the total jobs market. Almost everybody pays payroll taxes, and it stands to reason if wages and/or jobs growth stagnate, so will the taxes from those paychecks.

Matt Trivisonno of the Daily Jobs Update keeps track of these graphically, and makes a free graph available with a 90 day delay. Which means that the below graph includes the period up through roughly December 15. By way of further explication, the comparison is taxes paid during the entire previous 365 day period, vs. the entire 365 day period before that:

YoY payroll taxes peaked at about 22% in March 2022. Remember that my rule of thumb for non-seasonal adjusted data is that it probably has peaked if the YoY comparison has declined by more than 50% within the ensuing 12 months.

As of December 15, this had declined to about 8.5% - i.e., more than half, suggesting that, if we could seasonally adjust, we would find that payroll taxes paid had turned negative.

Because the data is public at the Department of the Treasury’s site, I can further report to you that as of March 20, the entire previous 365 day period resulted in 5.9% more payroll taxes paid than the 365 day period between March 20, 2021 and March 20, 2022. This is almost a 75% YoY decline.

But the data is a little more nuanced. During the 1st fiscal Quarter of 2022-23, payroll taxes paid were only 1.2% higher than Q1 of 2021-22. So far in the fiscal 2nd quarter this year, taxes are up 7.1% compared with Q2 of 2022 through March 20. That’s a pretty strong rebound.

And the rebound accords with information provided by California’s Treasury Department, shown below:

After a steep YoY decline in late 2022, there has been no further YoY decline in 2023 so far. Since California includes about 1/8th of the entire US population, and an even bigger share of the economy, this is good information.

Interestingly, the Treasurer’s Office for California indicated that they believed the big shortfall in 2022 was due to the downturn in the stock market, which meant that stock options tied to an increase in share prices could not be cashed. The stock market recovered pretty nicely between November and February, so likely some of those stock options were now in the money and could be cashed.

Probably the closest monthly analog is aggregate payrolls in the jobs report, shown YoY below:

Again, there is deceleration, but no nominal downturn yet.

Put the three data series together, we see that two short leading indicators, temporary help and consumer spending adjusted for inflation (since consumption leads employment) have turned down, but no coincident downturn in overall employment. In short, no recession yet, but no paucity of yellow flags.

Tuesday, March 21, 2023

February existing home sales confirm prices have declined, but bottom in sales and construction may be in


 - by New Deal democrat

There were only two noteworthy takeaways from the February existing home sales report:

(1) like mortgage applications, permits, and starts, existing home sales responded to lower mortgage rates (a decline from just over 7% to just above 6% between last October and January):

(2) As usual, price changes lag sales; for the first time since the pandemic, the median house price actually declined -0.2% YoY from $363,700 to $363,000 (remember: this data is not seasonally adjusted):

This simply confirms the data we have gotten from the more important new home construction data. I do think there is some good news here, in that we may very well have seen the bottom in this metric as well as in permits, starts, and mortgage applications. This does not mean a recession is not going to happen; but it does suggest it will not be prolonged and may not be very deep.

Monday, March 20, 2023

Average and aggregate nonsupervisory wages for February


 - by New Deal democrat

There’s no significant economic news today, so let’s update a couple of income indicators important to average American working households. Namely, because we now have the inflation report for February as well as payrolls, we can update average and aggregate nonsupervisory wages.

Average hourly earnings for nonsupervisory employees increased 0.5% on a nominal basis in February, tied for the strongest reading since last June. But since consumer prices increased 0.4%, real average hourly wages only increased 0.1%:

The good news, as indicated above, is that this is tied for the highest reading since last April (as I’ve noted many times, a $2 decline in gas prices can do wonders for economic statistics). The not so good news is that the above graph is normed to 100 as of January 1973, the record high water mark for nonsupervisory wages prior to the pandemic. Which means that we remain below that level, as we have been for almost a year.

Second, aggregate real nonsupervisory payrolls are an excellent way of viewing the health of the American middle/working class as a whole. Nominal aggregate wages were unchanged in February, but in real terms declined -0.3% in February from January’s record high:

Typically real aggregate payroll growth slows down sharply in advance of recessions, and usually stalls out during recessions. In fact, YoY negative growth is a very good “fundamentals” mark of recession, because when average American households have less money to spend in the aggregate, they cut back. And a cutback in consumption leads to a cutback in jobs.

And the YoY trend in real aggregate payrolls, while not negative, has declerated sharply in the past year, and is currently at 1.4%:

In the below long term graph, I subtract 1.4% from YoY growth so that it shows at the zero line:

As you can see, this level is consistent with a sharp slowdown (e.g., 1967, 1994, 2016) as well as an oncoming recession. *If* consumer inflation continues to ebb, then to indicate a recession, aggregate payrolls will have to decelerate faster than they have so far.