So let us look at US manufacturing new order data, at first in terms of the frame, which bounced back on a monthly basis in April:
We can see from the following chart that annualised real growth rates over rolling 10 year time frames (using new order data adjusted for PPI and smoothed) have been heading downwards since the start of the decade. The real annualised rate of growth over rolling 10 year time frames shows that the current cycle is much weaker than the pre 2008 cycle.
We can see the step down in growth more easily if we just view the high water mark dynamics (which ignores the current index level if it remains below the HWM).
In reality given the declines in flows since 2014, new orders for manufacturing adjusted for the PPI stand below levels reached at the end of the 1990s, and yet, here we are all worried about the risk of recession when the risk of something far greater has already occurred. Tied into the frame is the key dynamic of global rebalancing: cheap labour in emerging/developing economies had offshored a significant amount of manufacturing output from the late 1990s onwards (the weak order data partly reflects this) and the expected maturing of developing economies consumers and rising wage costs was anticipated by many to create a rebalancing of global economies. This is presently at risk a) due to the time frame of transition being longer than many expect and b) given that technology is driving out labour in key industries.
And in the next chart we see nominal annualised rates for new orders over rolling 10 year time frames: clearly the structure and balance and dynamics of the US economy and the attendant global economic dynamics are unable to support the type of growth the economy had experienced in post WWII years. The primary issue we are facing today is not one of a prospective recession but of a weakening frame:
US durable goods order data out yesterday suggested a strong rebound in new orders in April led principally by transportation orders. But I am not going to talk about the new data (at least until I get the full manufacturing new order data set due shortly), I want to talk about the difference between the new and the old order data:
Data to March 2016 had been revised down by some 2% for manufacturing orders and the rate of revision was similar for both durable and non durable goods.
However the wider picture shows that transportation, in particular motor vehicles and parts orders showed significant upward revisions:
On the other side of the variance we see non defence capital goods excluding aircraft, computers and electronic goods and primary metals heavily negatively revised:
The picture confirms the relative strength we have seen in auto led consumer credit growth on the one hand and weakness in exports and capital expenditure on the other.
Nominal retail sales data is typical of a recessionary environment, but much of this is due to declining gas prices. Manufacturing output and new order data is also typical of recessionary conditions. Motor vehicles and parts sales/new orders/output are still strong data points albeit showing signs of weakening, especially in the auto components. Cycle to cycle we see retail sales, orders and output all failing to establish a clear positive post crisis fundamental growth trajectory. That said there does not appear to any abrupt collapse in the data which is not necessarily a positive.
There are some interesting patterns and trends in US data: so I do ask myself, are we at the peak of the current cycle, are we as far as debt and low interest rates can take us?
US income growth has long been acknowledged to have weakened considerably yet recent data shows that the trend has indeed been weaker than first thought. Note the following chart showing pre and post revisions to chained per capita personal disposable income:
Yep the employment numbers looked reasonable but we are going to have to see better wage growth going forward and much less reliance on consumer credit if we are to believe that the economy is no longer skating on thin ice:
Motor vehicles and parts new orders have kept on rising:
But non revolving debt is growing at its hottest pace since 2000/2001
And as a % of disposable income, much higher:
And relative to income growth, well, historically high levels again:
I just do not feel comfortable with these kinds of fundamentals underpinning growth expectations.
Low savings rates, high levels of consumption relative to income and a heightened dependence on current transfers for income growth looks to be a poor frame for consumption growth. Too much reliance is being placed on “auto loans” and a steady if unspectacular growth in employment.
US consumer debt has been growing at a reasonable pace, and non revolving debt particularly so. In fact if we relate annual average growth in non revolving debt to growth in personal disposable income (using 3 year rolling averages) we find that non revolving credit growth is at historically high levels relative to income growth.
The US economy lies somewhere between boom and bust as shown by the following graphical representation of real GDP growth. Nevertheless, there are aspects of US economic growth that have boom type characteristics/risks; these are found primarily in the significant increases in auto focussed consumer credit and automotive production/capacity
Short term data has varied wildly of late; such can often obscure the underlying trend: what if we adjust for inventories and changes in consumer credit? Well we see less noise for one, but we also see a slower underlying growth profile – yes, credit creation is part and parcel of growing expenditure but I still feel we are in a high debt/deleveraging and weak income growth dynamic that needs to be especially sensitive to growth in credit/debt.
We know the main PMI index weakened in April and the two most recent regional PMIs also disappointed (Phil Fed/NY Empire State) and March industrial/manufacturing output confirmed a weaker manufacturing picture. The most recent NFIB report, while showing improvement, was still decidedly gloomy.