Slack Durables Demand; Giant Inventories; Cratering Goods Prices
The US economy is in a very precarious place right now. You've probably been reading that GDP growth remains strong, and if you don't bother to look past the headline numbers, that looks true. But it doesn't take a very far trip down the splits to see one giant problem, and host of smaller ones.
The giant problem is this (annualized real dollars):
- The US economy grew by $396 billion from July 2018 through March 2019.
- Of that, $331 billion, or 84% went into non-farm inventories and were still there at the end of March, despite a fierce March retail frenzy cued by steep discounts.
- The real top-line growth for the three quarters was 2.86% annualized. Without inventory growth, the entire rest of the US economy grew at only 0.47%, near recessionary.
- There is so much extra stuff, even after a downward revision, the number one contributor to nonresidential structure investment in Q1 was warehouses.
About a third of the inventories are light trucks and SUVs, split evenly between new and used. Demand for new vehicles from consumers has been particularly slack. Corporate investment in light trucks has kept the sales up. Likely many of those "new" vehicles sitting on lots are last years' models and will only sell at deep discounts.
Of the other two-thirds of inventory growth, it is mostly consumer durables, as well as many categories of capital and industrial goods.
Fun fact: many of these industries are the exact same ones that will see prices pressures from across-the-board tariffs on Mexican imports and Mexican retaliation. Just when they are trying to empty out inventories with prices cuts, their margins will be thinned more by a 5% tax.
Another area of giant inventories is clothing. Like autos, this is last years' fashion and will only sell at deep discounts.
Giant inventory growth is a down payment on either slower future growth, at best, or systemic liquidity collapse, at worst. The current Atlanta Fed GDPNow Nowcast has real Q2 GPD growth pegged at 1.2%, with inventory reduction pulling 1.3% off the top line. Without the reductions, the top line would be 2.5%, not nearly as shabby. At the currently measured rate, inventory reduction will pull out about half of Q2 growth.
But that's just the beginning. At that rate it will take an additional year after Q2 of similar inventory reductions to get back to where we started at the end of June 2019 - a roughly 2-year cycle of buildup and drawdown.
The worst-case is a systemic liquidity collapse, where correlated high inventories lead to a collapse in their prices, and the balance sheets of their owners. The first such example is the Dutch tulip craze in the 17th Century. The last one was the housing/financial crisis in 2008. History is littered with more.
Last week brought the usual month-end barrage of quarterly revisions and new monthly numbers for April, so let's see where we stand. I already covered the advance print in extreme detail, but here I'm going to focus on revisions and the new April numbers.
Q1 Major Revisions
While large revisions on the top line are unusual, this is not so the farther you get into the splits, and often the revisions can change a picture significantly. These are more reliable numbers, and the "final" ones come in a month ("final" number are also sometimes revised). You can see the challenge for policy-makers like the FOMC, who have to make decisions about today based on either unreliable, or stale data.
I always start by looking at incomes, as that drives everything else. Right off the bat, we see large revisions downward in both Q4 and Q1. The two-quarter annualized growth rates (except savings rate):
The big riddle of the advance report was that income growth had remained strong, but consumption really cratered from December to February. It had looked like people were saving their money instead of consuming, but in fact consumption was following reduced growth in incomes, which we will see are now not keeping up with real expenses in housing and health care, and other consumption is being crowded out.
Moreover, only DPI per capita is inflation adjusted, the rest are nominal, so you can pull 1.5% off the top of those growth rates. This solves that riddle a bit, so let's look at the large consumption revisions. Annualized growth rates:
You can see why we dig down on revisions. These are much more accurate numbers, and can turn a bad couple of quarters around, like with securities commissions.
While vehicle demand is still terrible, the revised Q4 and Q1 numbers make it a little less bleak. Furniture and household durables went in the other direction. These are key areas of the inventories story, so we will be spending more time with them when we get into April.
Turning to structures investment revisions. Annualized growth rates:
The most notable revision turns manufacturing construction from ugly to pretty. But even with a substantial downward revision, warehouse construction still leads Q1 contributors to nonresidential structures investment growth, accounting for 96% of the growth in that line.
Equipment investment, annualized growth rates:
More large revisions here, most notably taking communications equipment to negative. We'll dig deeper into investment later on.
The best revision in the report, is a small downward revision of inventory changes by $3.2 billion. A drop in the bucket, but at least it's something.
Incomes were up in April after a rough 1st quarter, but this was mostly driven by large increases in dividend, interest and asset income. After a long period of growth, wages are starting to flatten out, and small business income has been having a very rough go of it since January.
Annualized growth rates for the trailing 3 months:
All these rates are for nominal figures, so think of the core PCE inflation line as the baseline.
The first thing we see is that the revisions have changed the picture dramatically. In the advance report, November and December had looked like huge Christmas bonus months, but now they look very subpar after revisions. Wages have recovered from the year end low, but are fading now.
Zooming in on wages:
We can see that the top line for private industry wages is largely driven by services. Manufacturing wages have been fading since October and is growing below the inflation rate.
Overall, the giant revisions in the incomes picture really clears up why we saw such a collapse in consumption from December through February.
April Consumption and Prices
The wild swings in goods consumption that began in November took a month off in April. The numbers look a little weak, as did the advance April retail, but it is more an effect of the tough comp with March. Annualized MoM real growth rates:
Goods (red) had such wild swings that it is literally off the chart in November, December and March. Along with preloading inventories before tariffs, this is why buyers had such trouble managing their inventories.
Digging down into the splits, we see that services, especially housing and health care and eating up everything, and durables goods consumption is taking it on the chin. Steep price declines in March and April are driving the retail recoveries in these sectors. Annualized PCE growth rates:
Durables and clothing wholesalers and retailers have prevented a complete collapse in demand with the oldest solution there is: steep price cuts. This is how they are looking to clear out inventories.
Vehicles again provide a problem for the auto makers. Dealers have huge inventories of used light trucks, and also many "new" trucks that are last year's model. They are going with very steep price declines to clear out their used inventory and it is working. Auto makers see none of this revenue.
On the flip side, dealers are sticking with increased prices on this year's model trucks, and no one is biting. Instead, they see the great deal on a 2016 F-150 just off a lease with low miles and grab that.
Furniture and major household appliances are the other areas of low demand and high inventories where we are seeing big price cuts. It's working in furniture, but not the appliances.
Clothing is a real mess, and margins must be razor thin at this point for retailers.
Resetting the Investment Picture
We don't get monthly numbers on investment, but there were large Q1 revisions all over the splits here, so let's push reset and see where we were at the end of March. Also, we must discuss Boeing (BA), because they are all over this report.
The revised structures investment picture is much improved in factory construction and health care facilities, and the not-terrible top line growth rate here is 1.71% instead of -1.97% we saw in the first print. But pretty much all of that is unusually growth in warehouse construction. You know, because there's so much stuff, there's no place to put it all. Annualized Q1 growth in billions of real dollars:
That's all we need to say. It's all inventories nowadays.
Residential construction is in the toilet, but that's a whole article to itself, so I'll skip it here.
Shifting to equipment investment, as we saw there were many downward revisions, and overall the picture looks bleaker in many categories outside of computers and servers, which had a big recovery quarter. Many of the large inventories are in these industrial/capital goods categories
But the big driver here is Boeing's problems which show up all over the report, but especially here:
So the top line number is -1.05% real growth, but subtracting the cratering of civilian aircraft, equipment was up 0.71%. Certainly not good, but at least not shrinking.
Also of note, despite downward revisions, corporate investment in vehicles remains very high, which is offsetting the cratering of consumer demand somewhat.
Industrial equipment, where we also see very high inventories, had the worst quarter of any of the major subcategories, down -2.90%.
IP investment had some pretty big downwards revisions, but remains very strong, so we'll skip it.
I already wrote this up with the advance numbers, but let's do a quick check on the amazing amount of influence this one company's problems has on the entire GDP report. Annualized numbers:
Add it all up, and Boeing snatched 27 bps from the top line. Egads.
Inventories: The Big Kahuna
BEA does not publish April numbers here, so we will have to look at the Census Bureau surveys from last week. Starting with wholesale and retail:
As you can see, the entire chart is green. After March when at least retailers were able to draw down inventories, we are back to the same YoY growth rates in inventories. I'm not sure what is going on.
Similarly, when we turn to manufacturers, they are also mostly seeing higher inventories and inventory-shipment ratios.
While the overall picture is not terrible, in the select industries where the problems lay, only computer manufacturers are digging themselves out of this hole. Where you see green in those last two columns, the problems are continuing or getting worse.
Yes, the auto makers. Stay away.
The Atlanta Fed has a "nowcasting" model called GDPNow. The idea is to use the same data and algorithms as BEA uses for their advance GDP report, and update the nowcast as each new data point is released. Here's the current table for Q2:
GDPNow Columns are annualized real growth, except the rightmost two, which are in annualized real billions of dollars.
As each of those reports in the second column is released, the nowcast is updated. The difference between this and a forecast is that it tells us what the quarter would look like if the rest of it was exactly the same as up till now. There is no human judgement or intervention in the number - just the raw data and BEA's algorithms.
So I am confused. The inventory data from Census (incorporated into the model on 5/24 and 5/30) shows inventories rising everywhere, but GDPNow has inventories being reduced by -$60 billion annualized real dollars. I tried to track down their math, but it seems to indicate an increase of $57 billion, so I'm not sure what's going on here exactly.
But for argument's sake, let's assume that -$60 billion turns out to be right.
The current GDPNow has Q2 pegged at 1.2% growth, with inventories sucking 1.3% off the top line - without inventories, their nowcast would be 2.5% growth, a huge difference. Clearing inventories will suck out half of the growth in the quarter.
The column I have highlighted is CIPI, or changes in private inventories. This includes all inventories, including farm inventories, so it is not strictly what we're talking about, which is just non-farm inventories.
But for the sake of argument, let's say that farm inventories are flat, and that number is just the reduction in non-farm inventories. At this rate, it will take over a year to clear out the added inventories of the July to March buildup, with 5.5 quarters in a row where 1.3% of growth is snatched from the top line. That takes us all the way to the end of August 2020 - a 2-year cycle of inventory buildup and then drawdown.
The revisions to Q1 and the new April numbers have not changed the picture appreciably, with the exception income and wages.
- The big spike in December incomes was revised away, and now the collapse in consumption from December through February makes much more sense than before.
- Small business owners' incomes and manufacturing wages are driving the slowing trend.
- Even with steep price cuts, some businesses still cannot clear giant inventories. Some have been more successful, but have a long way to go.
- The auto industry has been bifurcated between used light trucks and everything else. People seem to be choosing them over new model year vehicles in a big way.
- Goods prices are deflating rapidly as a reaction to slack demand. Consumers reacted in a big way in March, but backed off a bit in April. Still, the consumption picture is far better than it was in December through February.
- If increased tariffs on China and Mexico go through, with anticipated retaliations, this will strike a serious blow to these durables industries that are already under tremendous pressure.
- Investment is slacking badly, with the exception of IP investment which blows out every quarter. IP saw some big downward revisions in Q1, so that may be the beginning of a trend.
- Boeing's problems are all over the GDP report.
- While the Census Bureau reports show inventories rising all over, GDPNow shows them dropping. I'm not sure what's going on under the hood.
- At the rate currently predicted by GDPNow of inventory reduction in Q2, it will take until August of 2020 to return inventories to June 2018 levels, with GDP growth reduced by 1.3% each quarter along the way.
As I said up top, that's the good scenario. The ugly scenario is that the growth of inventories pushes down prices to such an extent, that they all become next to worthless. This is what happened in Holland with tulips in the 17th Century, and with empty subdivisions where no one wanted to live in 2008.
I'll go with the slow growth.
I Regret Nothing
I totally stole the "ex-Boeing" framing from Hale Stewart, whom you should follow. I regret nothing.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.