In Q1, bank conditions for loans appear to have darkened

 

 - by New Deal democrat

Until Thursday we are once again in a data drought this week. In the meantime, there are a few points I want to address, including the very important Moody’s downgrade of US debt.

But there was one important piece of data that came out last week that I didn’t discuss yet: the quarterly Senior Loan Officers Survey published by the Federal Reserve.

The ease or difficulty in obtaining a loan is an important long leading indicator. Banks generally ease credit terms earlier in the cycle, and tighten them as they become incrementally more cautious about loan repayment. In general they turn relatively cautious more than 12 months before a recession.

I have not placed a lot of weight on the long leading indicators for several years, because their information was confounded by the massive kinking and then unkinking of the supply chain during COVID. While that ended at the beginning of 2023, the problem for, e.g., interest rates, has been whether I should base a forecast during this entire expansion including the supply chain problem years, or only since the beginning of 2023? There is simply no good answer.

But the Senior Loan Officer Survey does not have that conundrum. Since the beginning of 2023, there either has or has not been more demand for loans, and banks either have or have not tightened terms and conditions since then. So I can safely look at the trends over the past 2+ years.

Many of the old metrics from this release were discontinued some years ago, and others do not have an extensive history, but there are two important metrics that have been reported consistently for 35 years. 

The first of those is demand for loans from producers. More demand is expansionary; less is constractionary. In the below graph, the thick lines are for loan demand from big firms. The narrower lines are demand from small firms:

Note that these turned down over a year before both the 2001 and 2008 recessions. They also turned down later during the 2010’s expansion that may or may not have been cut short by COVID. As indicated above, they also turned down during the period of COVID supply chain tightness.

But over the last several years the situation looked very much like the early recoveries from both the 2001 and 2008 recessions. Demand was still not strengthening, but it had stopped declining in relative terms. This needless to say was good.

Now let me focus in on the last 5 years of this data:

After being positive in Q4 2024, it turned down in Q1 of this year. Only one quarter, but if it does not turn back positive this quarter then we have likely broken the improving trend, and this metric becomes a negative for the economy one year plus out.

The second indicator with a long history of being leading is whether banks are tightening or easing loan terms for firms. In this metric a number above zero indicates more tightening and so is a negative for the economy:

There is less noise in this indicator, and only one significant false positive, in 2016. Like demand, it was getting better in 2023 and 2024 after the supply chain issue stopped, and looked very much like an early recovery chart.

But in Q4 of last year the decline stopped, and it reversed higher in Q1 of this year. This is significant tightening, a sharper increase than in 2016. Which means it is already a negative for the economy in 2026.

Finally there is one important caveat. The Chicago Fed publishes weekly figures for financial conditions, which while noisier in the past have generally tracked with the quarterly Senior Loan Officer numbers. These are another set of series in which a negative number means loosening, so good; a positive number tightening, so bad. 

In any event, they have not tracked with the most recent Senior Loan Officer Survey this year:

The weekly numbers indicate continued loose conditions, with only a very slight move to “less loose” in the past several months. I would expect these weekly numbers to turn positive (i.e., bad) significantly before the start of any recession.

Weekly Indicators for May 12 - 16 at Seeking Alpha

 

 - by New Deal democrat

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

Changes due to Tariff-palooza! are happening very slowly. Most noteworthy this past week, rail traffic is still running ahead of rail traffic in the same week one year ago. But when we focus just on the intermodal container traffic, which is the main type coming from overseas, the growth rate of the volume - while still higher cumulatively than the first 4.5 months of 2024 - has slowed down comparatively almost every week since late March, suggesting that very slowly at least the backlog from front-running is being resolved.

As usual, clicking over and reading will bring you up to the virtual moment as to the economy, and bring me a penny or two in lunch money.

Housing permits and starts still rangebound, but with units under construction down almost -20%, is the last shoe finally dropping?

 

 - by New Deal democrat

In April total permits (dark blue in the graph below) declined -69,000 on an annualized basis to 1.412 million, while the less volatile single family permits (red, right scale) number declined -50,000 to 922,000. The slightly lagging and much more volatile starts number (gray, narrow) rose 22,000 to 1.361 million annualized:

The same data on a YoY basis demonstrates how it has been rangebound:

This is of a piece - and largely caused by - mortgage rates (YoY change, inverted, *10 in the graph below), which have also been rangebound between roughly 6% - 7%:

You may recall several years ago, even though starts and permits had declined sharply, the number of housing units under construction - the closest proxy for the actual economic impact of new housing construction - continued to levitate at all-time record levels. But ultimately they declined sharply as well, Once that happened, ever since the beginning of 2024, I have paid ever more attention to how deeply it would decline. Typically it has taken about a -15% decline to be consistent with a recession. Once that happens, the last show to drop is the number of employees engaged in residential building construction (red, right scale in the graph below). In April, housing units under construction dropped another -9,000 to 1.382 million annualized, a -19.6% decline from their October 2022 peak, while residential construction employment finally did decline as well, if only by -700:

Last month I wrote that “Since the significant downturn in units under construction began about 18 months ago, I suspect the turn in employment will take place within the next few months.“ I suspect April did indeed mark the turn.

To better show the trend, here is the same data on a YoY% change basis, together with manufacturing employment (gray):

With the exception of one month in 1995, any time both housing units under construction have been joined by residential construction employment as YoY negative, a recession has followed within 12-18 months. When manufacturing employment is also down, recession has been inevitable.

If April did indeed mark the turning point for residential construction employment, a loss of only -7,000 jobs in that sector over the next six months would be enough to set of recession alarm bells.

Industrial and manufacturing production suggest front-running production has peaked

 - by New Deal democrat

The final datapoint for today is industrial production, including its important manufacturing component. 

Last month I wrote that “I suspect the big increases in February and March in manufacturing, like this morning’s retail sales numbers, were about front-running T—-p’s tariffs. Which means that like retail sales, production might have been pulled forward from the next few months, which may lead to whipsaw declines.”

That probably started to happen in April, as total production (blue) was unchanged, while manufacturing production (red) declined -0.4%:

But improvement continues to show on a YoY basis:

This data was partially supported by the first two regional Fed manufacturing reports for May, from New York and Philadelphia, which came in at -9.2 and -4.0, respectively. But the new orders components of both the NY and Philadelphia surveys improved, however, to +7.0 and +7.5, respectively - which were sharp improvements from -8.8 and -27.2 last month.

I think it is safe to suggest that the front-running of tariffs on the production side may have peaked; but on the other hand there is no significant evidence of contraction beyond what may be monthly noise. The expansion continues, for now.

Real retail sales turn down in April, but continue to reflect consumers’ front-running of tariffs

 

 - by New Deal democrat

Next up in today’s slew of data is retail sales. This is one of the most important indicators I look at, because it tells us so much about consumers, and since consumption leads employment, it gives us information about the trend in that as well.

In April, nominally retail sales rose 0.1%. But because consumer prices rose 0.2%, real retail sales declined after rounding by -0.2% (blue in the graphs below). In recent months I have also been calculated real sales excluding shelter, because that has been distorting the CPI. This month the result was the same: real retail sales ex-shelter were down -0.2% (gold). In the below graph I also show real personal consumption expenditures for goods (red), which tends to track real retail sales well, but won’t be reported for several more weeks:

With rare exceptions - one of which was in 2023-24 - when real retail sales are negative YoY, a recession has followed shortly. In the past 12 months, real retail sales YoY have been positive, and was so again in April, up 2.8%. Excluding shelter, real retail sales were up 3.7%:

These are strong positive readings, and as so much I have reported on in the past few weeks, almost certainly have been affected by consumers front-running price increases and shortages anticipated from tariffpalooza.

Finally, let’s compare the YoY% changes with their potential effects on employment (red):

The good news is that these imply that the YoY% change in employment should hold steady or even improve a little bit in the next several months. Given that all but one month last spring and summer shoeed under 150,000 gains in employment, this implies job gains in the 150,000-200,000 range.

Jobless claims: more of the same

 

 - by New Deal democrat

After a long data drought, there are many releases today. I’ll start with jobless claims.

Initial claims were unchanged at 229,000, while the four week moving average rose 2,250 to 230,500. With the typical one week delay, continuing claims rose 9,000 to 1.881 million:

On the YoY% basis more important for forecasting purposes, initial claims were up 3.2%, the four week average up 6.1%, and continuing claims up 5.1%:

These YoY numbers are in line with what we have been seeing for the past eight months. They imply a relatively weak but expanding economy.

Finally, let’s take our first look at what this might imply for the unemployment rate in the next several months:

There is no upward pressure from either initial or continuing jobless claims, implying the unemployment rate will stay in the 4.1%-4.2% range.

Average and aggregate nonsupervisory real April wages continued to fuel the consumer

 

 - by New Deal democrat

Now that we have April’s consumer inflation data, let’s update real wages for average American families.

In April average hourly wages for nonsupervisory employees increased 0.3%, and aggregate payrolls for nonsupervisory employees increased 0.4%. Since CPI increased 0.2%, in real terms wages (light blue) increased 0.1% and aggregate payrolls (dark blue) increased 0.2%:

In the case of payrolls, this was a new all-time high. In the case of wages, it was an all-time high excluding April and May 2020, which were distorted by layoffs that concentrated on low wage service workers.

Here are the same metrics as YoY% changes:

Real hourly wages are up 1.7%, while real aggregate payrolls are up over 3%. 

The bottom line is that in April ordinary American consumers had more to spend in real terms, which is good for confidence and also means they had more of an ability to front-run tariff impacts by purchasing goods in advance.

In preparing this post, I wondered how much it was a feature of earlier recessions that low wage employees bore te brunt of layoffs. So the below two graphs compare real average hourly wages (light blue) and real aggregate nonsupervisory payrolls (dark blue) since the 1960s.

Looking in reverse chronological order, we see that low wage workers appear to have borne the brunt of recession layoffs in both the 2001 and 2008 recessions as well:

But in the 1970s through 1991, both aggregate real payrolls and average real hourly wages moved more or less in tandem:

Note by the way that over time aggregate payrolls increase more than wages, because of populations and labor force increases. In other words, if real wages are unchanged, but more people are earning those wages, then the aggregate goes up while the average does not. And when we are talking about whether the economy as a whole is improving or contracting, the aggregate amount is more important.

In any event, the above suggests that those earlier recessions hit the spectrum of wage earners more equally; but it is also possible that it is not a coincidence that this earlier period is when women entered the workforce in huge numbers, so that recessions exacerbated the securlar downtrend in real wages that lasted until women were fully absorbed into the labor force by around 1995.

In any event, the news for April suggests that American consumers are not ready to roll over into a cautious recessionary ball.