The shocks to consensus on US industry keep coming: this week we've already had disappointments from April's Empire State Manufacturing survey (just 6.6 vs a consensus of 18), and a second consecutive month where industrial production flatlined (and manufacturing fell 0.2%). The previous couple of weeks saw shocks from regional manufacturing surveys in Milwaukee, Kansas, Richmond and Dallas. It's time to start asking what ails US industry right now: why is another 'soft patch' materialising in the data?
We do not have the dramatic excuses of last year: we've had no environmental catastrophe punching holes in global supply chains; and though WTI oil prices have risen, the movement from around US$100 a barrel in 2H11 to a rapidly fading peak of US$110 is no repeat of the 2010-2011 jump from around US$85 to US$112+. What's more, banks are once again lending modestly (up 5.1% YoY in March and early April) and commercial paper markets are open to non-financial domestic companies (up 17.8% YoY). Finally, throughout 1Q we've got used to housing markets and labour markets, perennial bear-factors, are regularly delivering more positive surprises than negative shocks.
With none of the usual suspects fitting the frame, it's time to try to account for the soft patch from the bottom up – ie, by looking carefully at where US industrial output ends up. And it turns out that this step-by-step accounting approach does yield some answers.
Industrial output must either be bought by someone directly, be added to inventory, or written off. When we look at monthly data for manufacturing and trade sales, we find they remain relatively robust (latest data February), rising 7.6% YoY, and with a positive underlying sequential momentum which remains unchallenged (the 6m trendline for sequential movements is 0.21 standard deviations above the 10yr average). So sluggish domestic sales by themselves don't look to be the problem.
Output which isn't immediately sold to the end user can end up as inventory temporarily: there's a measurable lag of about four months. Total business inventories are matching sales, rising 7.6% YoY in February: there's no story here, since inventory/sales ratios have been essentially unchanged now for a full two years. No pressing inventory adjustment is likely to be generating a 'soft patch'.
But now let's consider the difference in momentum between what's being produced (industrial output) and where it ends up in the domestic economy (sales and inventory). In the chart below the blue line tracks momentum of sales, and the pink line tracks momentum of output minus inventories. In a closed economy which typically operates somewhere near equilibrium, the two lines would track each other closely – as indeed they do usually.
What if they don't match, however? Let's look at the difference in momentum between the two. In the chart below, the thing to remember is that if the line is in positive territory it tells us that the momentum of output is greater than the momentum of end sales plus inventories – in other words, one can and should expect a correcting retreat. Conversely, if the line is below zero, sales and inventories additions have greater momentum than domestic production, and one should expect a positive correction from industrial output.
More, the direction of travel helps us understand the mini-cycles which bubble up in the data. Thus by late 2010 the US had reached a point where for the first time since the financial crisis hit, momentum of sales and inventories was stronger than momentum of industrial output, heralding the unexpectedly strong growth in output during 1Q11. So it's really not surprising that the 'soft patch' of 2011 caught everyone by surprise, nor is it surprising that commodity markets reacted so strongly to the upturn. The shock of supply-side disruptions distorted the picture even more, and it was not until well into 2H11 that output momentum began to catch up once more with sales and inventories.
Notice what's happening now, however, is that that catch-up period has ended – all other things being equal we would expect a modest step-down in pace, unless the pace of sales and inventory momentum lifts.
So far, we have considered US industry operating in effectively a closed economy. But that's wrong. International trade can be seen as a balancing item: when output is rising faster than (sales + inventories), then one possibility is that the surplus output can be exported. Similarly, when output is lagging (sales +inventory) the shortfall in supply to the domestic economy will be made good by imports. As our final chart shows, it doesn't always work out like that – there's more flex in these arrangements than our accounting methods acknowledge.
However, what is clear is that the signals for trouble come when momentum of output minus (sales + inventories) is positive and rising, whilst momentum of exports is negative and fading. That's the point at which you must expect cuts in output – an industrial recession in other words. This was the situation in 2001-2002, and much more intensely in 2008-2009.
Obviously, US industry is not in a similar position now – but it could be getting there. If the trends seen in the last nine months were simply to be extended for a further nine months, the position would become similar: output momentum would be rising significantly faster than (sales + inventories), whilst the ability to export the surplus dwindles as export momentum turns negative. It is precisely to avoid this dynamic that the 'soft patch' is emerging.
What conclusions can we draw from this?
First, the industrial 'soft patch' is no simple mistake or data-blip: rather it is a slowdown necessary to secure something like industrial equilibrium. It can be expected to maintain downward pressure on bond yields, downward pressure on commodity prices (both of which are already manifest) but also, if it lasts, downward pressure on the dollar.
Second, the US industrial cycle is not self-contained, and is not insulated from trends in world trade. In fact, fluctuations in world trade play a crucial role in finding, keeping and maintaining the balance between domestic supply and demand. Right now, that means the US industrial cycle is exposed particularly to potential downturn in Europe, as well as to potential reflation in China.
Third, the imbalances are only just emerging, and industry is acting to ensure they don't curdle into a recession: the modest shocks now ought to prevent worse shocks later. But the exit from the soft patch is likely to need positive momentum surprises in sales, or inventories, or exports – and preferably a combination of the three.