US - Anything But Normal, Part I

Bloomberg’s published consensus expects no change from the ‘new normal’ as far as US growth is concerned over the coming two years. But the new political leadership has been washed into Washington precisely because the ‘new normal’ produced unacceptable results. It is therefore likely to strain every sinew to break out of it. And the current state of both returns to the strategic factors of production, and shorter-term potential cyclical accelerators mean there is a good foundation for the breakout to be made. 

Since 2010, quarterly GDP yoy growth has averaged 2.1% with a standard deviation 0.7 percentage points. This expansion has been a sustained but grinding affair, never seriously threatening to develop the usual cyclical dynamics either for recession or boom. And for the next two years, the consensus expects more of the same, settling on 2.3% GDP growth, with a 15% chance of recession.  

But to expect ‘the new normal’ to linger indefinitely no longer seems cautious. Radical change in political, financial and ultimately economic attitudes and choices have made it onto the US agenda. The desire to break out of the new normal is likely to frame major policy choices made this year and next. Whether that is possible or actually desirable will be discovered only in the long-term. But an attempt will be made.  

What would signal such a break-out? Let’s say three consecutive quarters of 3%+ growth.  This will only happen if the economic and financial groundwork for such a breakout has been made.  The major factors of product - returns to capital and labour - are increasingly friendly towards such a breakout.  Similarly, a series of sub-cyclical indicators which can act as accelerators to business cycles, including overall the overall inventory position, the capital goods cycle, and labour market indicators, all look better at the beginning of 2017 than at the beginning of 2016. 

None of this means that a break-out from the ‘new normal’ is imminent or inevitable. But it is certainly more possible and more likely than the steady 2.3% GDP consensus accepts. 

Strategic Position - Factors of Production

Over the longer term, the main dynamics powering growth are movements in returns to the factors of production - ie, labour and capital. If return on capital is rising, companies tend to invest more, and tend to have more cash to invest anyway. If output per worker is rising, companies tend to employ more labour - particularly if the rise in output per worker is not simply owing to a rise in the capital deployed. Both of these factors are increasingly favourable, suggesting increased investment and a more positive labour market. 

First, return on capital. I generate a return on capital directional indicator by expressing nominal GDP as a stream of income from a stock of fixed capital, and estimate that stock of fixed capital by depreciating all private non-residential investment spending over a 10yr period. As the chart shows, during the new normal, that ROC directional indicator essentially plateaued and then declined gently for four years from 2012. That gentle decline caught up with capital spending sufficiently that by the beginning of 2015 growth in the stock of capital was slowing. However, as growth in capital stock slowed, so raising asset turns became less demanding, and consequently the ROC indicator bottomed out in 1Q16, and has been rising since. Even in 2009, it took only a year after the upturn in ROC indicator for a recovery in capital stock growth to start.  As a result, we should certainly be pencilling in an upturn in the capital spending cycle for 2017, quite possibly as early as the first quarter of 2017. 

There is a similar dynamic at work in the labour market. The modest slowdown in employment growth during 2015 and 2016 be seen as a response to the fractional loss in labour productivity (real GDP per worker deflated by capital per worker) which first emerged at the beginning of 2014. That erosion, however, moderated in 2H16, which should backstop the slowdown in employment growth. If anything, it suggests a modest acceleration in employment growth during 2017, more likely in the second half of the year than the first. 

The early indicator to watch will be job openings, since inflection points in growth of job openings have quite closely echoed movements in this measure of labour productivity, as the chart below shows. Interestingly, an upward inflection in the rate of job openings looks to have been taking shape as early as 4Q16. 

The progress of the main factors of production will determine the general trajectory of GDP growth. But within an expansion, there are also a number of sub-cyclical disequilibria which can determine the dynamics of that expansion. Under certain conditions these can power a genuine business cycle (and ultimately undermine it, too). Throughout the new normal grinding expansion these have repeatedly failed to power a genuine business cycle upswing. 

The main subcyclical accelerators can be found in a) the inventory cycle; b) the capital goods cycle; c) the labour market cycle; and d) the credit cycle. 

The inventory cycle can provide some dynamic push to a business cycle. Additions to inventory positions, if made voluntarily, anticipate future demand, and by doing so, create additional current demand. Conversely, an involuntary build-up of inventory occurs when final demand disappoints expectations, and the determination to get rid of inventory supplies additional demand at exactly the point at which the market is oversupplied. 

The trend growth of business inventories has been slowing uninterruptedly since the middle of 2014, and in the 3m to November rose just 0.8%, whilst manufacturing and trade sales rose 1.7%. Within manufacturing, inventories were down 1.3% yoy 3ma, whilst shipments were down only 0.1% yoy 3ma.  The result is that the inventory/shipment ratio began to retreat early in 2016, and maintained that retreat all year. 

At the same time, manufacturers’ book-to-bill ratio has been steady, or perhaps in very modest recovery. 

The position of the inventory subcycle is, in this reading, perhaps only modestly positive. But the shifts in the inventory cycle have an obvious link with price movements and price expectations.: if prices are expected to rise, it makes sense to build inventory now (which will tend to raise prices); if prices are expected to fall, it makes sense to offload inventory as quickly as possible (which will tend to erode prices). 

It is in this light that the current rise in industrial prices matters. By December final demand PPI had accelerated to 1.6% yoy, the highest since September 2014, reversing the deflation and disinflation seen in 2015 and 2016. Just as that deflation/disinflation was both an expression of the attempt to curb inventory growth and a result of it, so it is now reasonable to expect the dynamic to reverse. Which means a greater tolerance or desire to hold inventory, which means a boost to immediate demand in the near future.  To cut a long story short, the underlying equilibrium in inventories has been gently adjusting to disinflationary expectations for years, and must now consider the implications of rising prices. 

The capital goods cycle is similar, in that if producers encounter slightly greater demand than expected, they may choose to invest in order to satisfy that demand. In doing so, their spending creates more demand, which generates further feedback loops which can accelerate cyclical forces within an expansion. 

The equilibrium within the capital goods sector is increasingly positive. In particular, the book-to-bill ratio , which spent 2015 to mid-2016 below 1, has been sustained above 1 since June 2016.  In addition, the inventory/shipment ratio, which rose almost continuously since 2012, appears to have peaked out in August. 

As with the inventory cycle, however, the adjustments have been made in the face of underlying weakness, not strength. Although the book to bill ratio may have improved, by 4Q orders for capital goods (non-defence, ex-air) were still down 1.2% yoy, whilst shipments were down 3.8%. The capital goods cycle also shares with the inventory cycle a sensitivity to underlying price movements: specifically, if capital goods are priced ‘correctly’, their price reflects movements in the price of revenues expected from them, discounted forward.  Consequently, if prices of  the goods they produce (consumer goods) are expected to rise, the price of capital goods will rise faster. Conversely, if price of consumer goods are falling/expected to fall, the price of capital goods will fall faster. 

As a result, the capital goods/consumer goods terms of trade are an important indicator of inflationary/disinflationary forces.  An interpretation of the chart below would include
i) 1980 to late 1980s - inflationary expectations still embedded, albeit in retreat
ii) late 1980s to around 2000 - erosion of inflationary expectations, replaced by modest disinflationary expectations
iii) 2000 to 2011 - deflationary expectations
iv) 2011 to 2015 - the new normal
v) 2015 onwards - tentative expectations of ‘a recovery in pricing’.

If this interpretation is correct, then even without an inflationary breakout, the foundations for an upswing in the capital goods cycle are made. 

So far, this piece has concluded that the long-term strategic factors are favourably positioned for more economic rapid expansion, and that in addition, two important subcyclical factors, having undergone downward adjustments in the last couple of years, are also more likely to show upward volatility than continued downward pressure, particularly if confronted with any upward demand or pricing surprise.  

These provide grounds for thinking a breakout from the ‘new normal’ is possible in the coming two years. But there are other factors to consider which are important and could yet scupper the break-out attempt. They are i) household saving/dissavings behaviour; ii) the wider monetary cycle; and iii) labour market dynamics, and in particular, potential movements in labour participation rates. These leads will be pursued in Part II.