One feature which is common to 3Q national accounts for the US and for those parts of Asia for which we have a detailed GDP breakdown is an unusually sharp build-up of inventories. These have inflated nominal and real GDP growth on both sides of the Pacific in 3Q, disguising a slowdown in final spending. The 3Q inventory build-up potentially represents a time-bomb under GDP growth in coming quarters. However, assessing the timing and strength of that threat will partly depend on what prompted 3Q’s build-up in the first place. Was it US-Sino trade threats to supply chains, or commodity prices? Or both?
First, a methodological warning: when it comes to inventory, I use the nominal numbers, rather than the deflated ‘real’ numbers. There’s a good reason for this: change in inventory behaviour, and by extension changes in inventory prices are at the very heart of business cycle dynamics. The claim that deflators can be accurately calculated is not just implausible in practice but, I think, flawed in principle. Similar reasons lie behind my use of nominal investment tallies when it comes to calculating my Return on Capital Directional Indicators. Let’s first see where we can see inventory-building inflating 3Q GDP growth significantly:
In the US, the $ 80.4bn nominal additions to inventories in 3Q reversed a $10.4bn drawdown in 2Q. This was the biggest quarterly turnaround in inventories since 4Q12, and accounted for 0.4pps of the 1.2pps qoq rise in 3Q nominal GDP growth. That's the highest proportion since 1Q15. Nominal GDP grew 4.9% annualized in 3Q, but final spending on domestic product (ie, GDP less inventory changes) rose only 3.1% (vs 8.6% in 2Q).
In Hong Kong, 3Q saw the biggest nominal rise in private inventories in 3Q for years, adding 1.9pps to nominal gdp growth of 6.7%. Perhaps more clearly, the rise in gross fixed capital formation was HK$19.12bn yoy, whilst the change in private inventories HK$12.86bn yoy. Nominal GDP growth came in at 6.7% yoy, but final spending on domestic product rose only 4.8%.
In Thailand, 3Q saw the biggest addition to private inventories since 1Q13 - an addition which added 5.7pps to the nominal 5.5% growth! So nominal GDP growth 5.5%, but final spending on domestic product actually fell 0.2% yoy.
In Singapore, the story is less obviously dramatic, but still material. Inventory additions accounted for only 1.5pps of 3Q’s 4.5% yoy nominal growth, but accounted for slightly more than all of the qoq growth. Nominal GDP rose 4.5% yoy, with a quarterly rise 0.2SDs above historic seasonal trends; final spending on domestic product, however, rose only 3.1% yoy with the quarterly gain 0.8SDs below trend.
Taiwan’s first estimate of 3Q GDP does not break out an estimate of inventory changes, but the extraordinary jump in gross capital formation (which includes inventory changes) strongly suggests the build-up was dramatic. In constant dollar terms, 3Q GDP rose 2.28% yoy, within which investment spending jumped 17.5% yoy, and accounted for 336bps of the 228bps of 3Q GDP growth.
I commented at the time: ‘It is an extraordinary number. Not only is it the first positive yoy comparison since 2Q17, it also the highest since the rebound-year of 2010, and before that since 2004. What accounts for it? . . . . Gross capital formation includes changes in inventory. Starting in 3Q17 and carrying on through to 1Q18, Taiwan was losing inventory fast. The sharpest inventory-dumping came in 3Q17, with the change in inventories equivalent to 1.4 percentage points of GDP growth. However, in 2Q18 there was a small addition of inventory, and if this continued into 3Q18 this could have contributed powerfully to overall investment growth. Indeed, even if there were no inventory building, but only a quarter-on-quarter standstill, this would still have represented approximately 6.3 percentage points of growth for total yoy investment spending.’
And in Europe, too, it seems something similar happened in Germany. In nominal terms, Germany’s 3Q GDP rose 1.8% qoq and 3% yoy, with gross capital formation rising 19.3% qoq and 12.4% yoy. However, within that, gross fixed capital formation (ie, investment excluding inventories) rose only 1.2% qoq and 6.3% yoy. The implied inventory growth is dramatic, and accounts for 1.4pps of 3Q’s 3% yoy growth. Nominal GDP grew 3% yoy, final spending on domestic product rose only 1.6%yoy
What accounts for it? There are two obvious contenders, both of which may be at work:
first, the scramble to secure and deliver supplies in anticipation of escalating US - Sino trade frictions;
second, a response to the the belief that the dollar is weakening and (thus) commodity prices rising.
US-Sino trade frictions explanations are consistent with when and where the inventory-builds are most obviously found: ie, in the US, Hong Kong, and probably Taiwan. However, It is not obvious why this should have been extended to Thailand (but not Indonesia). It is also noticeable that there is no obvious inventory build implicated in the UK’s strong 3Q GDP result (0.6% qoq), whilst the Eurozone’s disappointing 3Q GDP (0.2% qoq) result leaves little room to hide an inventory bulge.
Movements in commodity prices, and expected movements in commodity prices, regularly produce significant inventory shifts. Indeed, inventory additions and clearances are a major factor in commodity prices whipsawing at inflection points.
There are good reasons to expect that these inventory additions were in part the result of expected commodity price rises - namely movements in the dollar, in inflation, in bond market inflationary expectations, and in direct movements in commodity prices.
Throughout 2017 and the first half of 2018, the dollar had been almost continuously weakening against the SDR: between Jan 2017 and mid-April, it had lost 8.1% against the SDR basket of currencies. More often than not, the corollary of a weakening dollar is strengthening (dollar-based) commodity prices: and as the chart shows, this time was no exception, with the CRB index rising 26.7% between June 2017 and May 2018.
Those expecting these trends to continue had plenty of evidence going for them. Although it was not difficult to spot that in yoy terms global inflationary trends had peaked around July 2018 and were likely to retreat, the data showing that would only have arrived around late-August and early September, and it required a degree of confidence to recognize it at the time. Not least because between May 2018 and early August 2018, global inflation announcements had been sharply more inflationary than consensus expected, as shown in my global Shocks & Surprises inflation index.
Moreover, in the US, bond markets were at that stage signalling no retreat in inflationary expectations, with the inflation risk premium (10yr Treasuries minus 10yr TIPS) rising from a low of 167bps in late June 2018 to a peak of around 220bps in late May, from where it did not noticeably retreat until mid-October.
Which of these two possible reasons for 3Q’s inventory build was dominant matters because they imply different trajectories in the near term.
If the dominant factor was simply a reaction to the rise in commodity prices and the expected continued rise in commodity prices, then the strengthening of the dollar, the retreat of global yoy inflation prints, the moderation in inflation shocks & surprises, the retreat of commodity prices and finally the retreat also of inflation assumptions embedded in US bond markets all suggests that inventory holders will be dumping inventory overhangs as quickly as possible - and consequently that we can expect this to show up in noticeably weaker GDP growth in 4Q.
If the dominant factor is a desire to secure certainty of short- and medium term supply in the face of US-Sino trade frictions, then it is less clear that the 3Q buildup will be revoked quickly, or at all. For if the expectation is that trade frictions are likely to be a continuing feature of US-Pacific trade relations, it is likely that this will be answered by a longer-term increase in working capital tied up in inventory for both buyers and sellers.