The Campaign Against Prosperity

A useful axiom: bad policies result in bad economies, which in turn result in more bad policies. By contrast, good policies result in good economies, which in turn generate the space for more good policies.

We live in an age when growth in world trade has slowed to a crawl. Between 2000 and 2007, Northeast Asia’s exports rose an average 17.7% a year. After the violent shocks of the GFC are shaken out of the yoy data, between 2012 and 2018, those exports have grown only 2.6% yoy on average. And it is against this background of stalled world trade growth that future trade growth is compromised still further both by rising US-Sino trade frictions, and the threatened collapse of UK-EU trade relations.

What is little realized is that these potential overt trade dislocations do not come out of the blue. Rather, they are the logical extension of an aggressive campaign against international trade which has been waged by all major trading economies, and with increasing intensity, since the GFC. Bad policies have resulted in bad economies, which in turn are provoking even worse policies. This, then, is the story of trade wars foretold.

neax lt.jpg

One can reach for purely economic reasons to account for this dramatic curtailment of trade growth: changing patterns of domestic saving and cross-border investment clearly have a part to play. But using economics as a sole explanation overlooks a simpler and powerful explanation: the slowing of global trade growth to a crawl is (also) a result of deliberate policy aggressively pursued.

Trade has slowed in part because regulators in every major economy have raised obstacles to trade in goods, in the form of Non-Tariff Barriers (NTBs). NTBs are not usually noticed much, because they don’t immediately lend themselves to direct economic quantification. ‘Trade policy can take many different forms: tariffs, quotas, non-automatic licensing, antidumping duties, technical regulations, monopolistic measures, subsidies etc. How can one summarise in a single measure the trade restrictiveness of a 10% tariff, a 100-ton quota, a complex non-automatic licensing procedure and a $1 million subsidy?’ (Estimating Trade Restrictiveness Indices - Kee, Nicita, and Olarreaga, 2009)

Nevertheless, econometric calculation allows the authors to construct an Ad-Valorem Equivalent (AVE), which compares the impact of NTBs directly with direct trade tariffs. They conclude: ‘The importance of NTBs as a protectionist tool is substantial, especially considering that in 55% of tariff lines subject to core NTBs, the AVE of core NTB is higher than the tariff.’ ‘On average, they add an additional 87% to the restrictiveness imposed by tariffs. Moreover, in 34 out of the 78 countries in our sample, the restrictiveness of NTBs is larger than the restrictiveness of tariffs.’ The malign impact rises as a country grows richer, with the average AVE of NTBs rising with GDP per capita.

Those conclusions were reached using data from the early 2000s, when the US, EU, China and Japan had deployed comparative NTBs to impede trade. Since then the rate at which they have growth, and continue to grow, is extraordinary: back in early 2000, between them the US, EU, Japan and China had put up only 2,028 NTBs. As of end-November 2018, there are 12,612 NTBs either initiated or in force. More than two new NTBs are added by these economies every working day.

The means to track it are publicly available: the NTB tallies in this piece were all taken from the WTO’s Integrated Trade Intelligence Portal.


This wild proliferation of specific trade barriers has continued unabated even as the trade which they are designed to hinder has almost stopped growing. This looks like a classic case of bad policies producing bad economies, which in turn produce even worse policies.

One way of illustrating the interaction between NTBs and trade volumes is to chart the number of NTBs to be overcome by every US$1bn of monthly imports made by these countries.

ntbs per billion.jpg

The number of obstacles set in the way of every US$1bn of imports prior to the crisis fell gently from 13.6 at the beginning of 2004 to a low of 11.2 in 3Q08.

Since then, the rise has been dramatic, albeit disrupted by the business cycle. Still, it reached a peak of 21.8 per billion in 2016, and currently stands at 19.2. In other words, the number of barriers to trade has roughly doubled since the crisis. If the AVE of these NTBs is merely as expensive as a tariff on the goods, then it seems likely that post-GFC, we have seen the equivalent of a doubling in import tariffs.

This is our generation’s equivalent of Smoot-Hawley.

Yet this relentless campaign against trade - against prosperity - is for all intents and purposes invisible. It attracts no headline, courts no controversy. So I think I shall start to include monthly checks on the numbers of NTBs in my Shocks & Surprises work.

Which countries, then, are currently the worst prosecutors against world trade, and where are the instincts to damage world trade most virulent at the moment? Using the same metric (NTBs per US$ billion of imports per month), there is a quite clear order: Japan is by far the most protective, with 37.2 NTBs per US$1 billion of imports. The least active on this regulatory front - much to my surprise, I must admit - is the EU, with 11.1 NTBs per billion. The US is one of the most litigious economies in the world, so it is perhaps not surprising that it is heavily protective, with 27.8 NTBs per billion, whilst China, which only really started playing this game by WTO standards, has no reached 15.2 NTBs per billion.

ntb regions.jpg

The other thing this chart shows is that the campaign against trade really intensified dramatically in 2015 and 2016, when these economies added 947 and 687 new NTBs respectively. And whilst everyone was at it - this was the point at which trade regulators were gaming the system, presumably driven by Prisoner’s Dilemma rationalizations - the two most fervent gamers were Japan and China. Japan started 2014 with 1,240 NTBs, but by end -2016 they had 1,556. China’s regulatory activity was similar: at end-2014 it had 2,031 NTBs in force; by end-2016 it had 2,527.

This is a dark picture, and darkening by the day. But even in the darkest picture there is some relief, and in this case, it is that the regulatory enthusiasm seen in 2015 and 2016 abated somewhat in 2017, which produced only 503 new NTBs. As markets twitch to the latest twist in US-Sino trade diplomacy, it is clear that without a de-escalation in the NTB trade wars, the prospects for global trade, and global prosperity, will remain deliberately and profoundly compromised. Bad policies make for bad economies, which make for worse policies.

US & Asia: 3Q's Big Inventory Build

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.

dollar crb.jpg

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.

ss inflation.jpg

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.

inflation risk premium.jpg

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.

China: Exploring the Shadowlands

A few weeks ago I was challenged to produce a 'bull case for China'. It's a tough ask, and the best I could do was show how we can expect that profits growth will be strongly leveraged to the recovery, (when it arrives). This is because the efficiency gains (efficiency of finance, of asset-usage, and of labour productivity) which were won in the early stage of the credit squeeze will endure for at least the early stages of the recovery (when it arrives). (This is despite the evidence that the continued squeeze is, right now, eroding those initial gains.)

But when will it arrive? The belief that recovery will arrive sooner rather than later is made tougher by the run of data, as October's monetary, industrial and demand data showed this week.

Not that any of October's industrial or demand data was exceptionally grim. Rather, it was just uniformly disappointing: industrial output rose 5.9% yoy (0.3SDs below trend), electricity output rose 4.8% (0.5SD below trend) and exports (in Rmb and volume terms) came in 0.1SD below trend. For demand indicators, retail sales rose 8.6% yoy (0.8SDs below trend), auto sales fell 13% (0.6SDs below trend), urban investment rose 5.7% ytd (0.3SDs below trend), PMI employment indexes were 0.7SDs below trend, and only the real estate climate index managed a positive result (0.2SDs above trend). Industrial indicators have leaked momentum for four of the last five months; demand indicators for the last two months.

remi cn.jpg
ddmi cn.jpg

But then there's October's monetary numbers. Interpreting China's monthly monetary and finance data is becoming a bewildering puzzle, because although we know that great efforts are being made to scale back or close down the 'shadow banking' sector, we have no way really of telling how this is affecting banks' ability to lend, companies' ability to get credit from the banking system, and, moreover, how it is affecting government finances.

The key point is this: whilst the 'shadow banking system' of wealth management product, trust loans and entrusted loans started out with a good claim to be vehicles which could usefully circumvent interest rate ceilings for savers and institutional credit hurdles for companies, they morphed into something quite different.

In recent years, 'shadow banking' operations have tended instead to act primarily as ways in which certain types of commercial banks could sidestep overall credit limits, essentially by offloading loans to special vehicles (such as trust companies) in exchange for participation in 'Trust Benefit Receipts' which count as investments rather than loans. Over time, this has developed into a series of tightly interlocking financial relationships between various parts of the banking system and parts of the 'shadow banking system' which are very largely opaque.

To get a really good idea of the issues involved, I can recommend the BIS Working Paper No 701 'Mapping shadow banking in China: structure an dynamics.' This is a brilliant paper, and, if nothing else, one should take a look at the BIS's schematised map of China's shadow banking system, starting with the 'easy' map on p4.

BIS writes: crucially, 'direct shadow credit to ultimate borrowers has slowed considerably in recent years, whereas the volume of shadow funding as well as structured shadow credit intermediation has grown at a fast pace.' I think this is BIS being diplomatic: what it means is that in recent years, the principal driver for 'shadow banking' activity has been the need or desire to circumvent banking regulations by converting 'expensive' on-balance-sheet loans into 'investment receivables' from the trust company in which those loans have been parked.

That's good news and bad news. The good news is that winding down the 'shadow banking activity' as seen in the monthly aggregate financing numbers probably hitting banking operations more than credit creation. In other words, the slowdown in total aggregate financing probably overstates the slowdown credit created for ultimate users. The bad news is that we simply have no idea how those 'tight interlinkages' are going to be unwound, and what 'unknown unknowns' will make themselves unavoidably 'known'.

However, one conclusion - and no conclusions in this case are 'obvious' or even certain - is that during the expansion of the shadow banking system, the monthly scores of total aggregate financing are likely to have been inflated by an unknown degree of double counting. If so, it is quite possible that the current slowdown in total aggregate financing (just Y729bn in October, with bank lending accounting for Rmb714bn) with the contractions consistently recorded in trust loans and entrusted loans, is correspondingly less dramatic than it seems. Quite how the dis-integration of the shadow banking system and the banking system will affect, or is affecting, final credit provision is, I think, impenetrably obscure.

However, some light is shed by the inclusion this year of 'special local government bonds' as a separate line-item in total aggregate financing. What seems fairly obvious is that these bonds have been issued at a pace directly to offset the fall in the 'shadow banks' trust and entrusted loans.

During Jan-Oct this year, there has been a net repayment of Rmb1.847tr of trust and entrusted loans. Since trust and entrusted loans were the favoured vehicle for local governments to finance themselves, it looks like it is local governments which have been doing the repaying: during the same period, the issuance of 'special local government bonds' has amounted to Rmb1,782tr. The high-interest rate (and for banks' customers, high-risk) trust and entrusted loans are being repaid by the proceeds of the local government bonds.

shadow and bonds.jpg

But whilst these bonds are looking after the interests of local government finances, it is far less clear what the knock-on effect on final credit provision for other economic entities must be.

Nevertheless, it has the further consequence that we are getting a less unrealistic picture of the state of China's government finances. Consider, for example, that the total published fiscal deficit for Jan-Sept 2018 came to Rmb1.749tr, whilst the new local govt bond issues for Jan-Oct came to Rmb1,782tr. At some point we should start counting these new debts as evidence of previously unregistered deficits (rather than assuming they represent current fiscal gaps).

For those with long memories, the rather baroque system of 'tight interlinkages' which now characterise China's shadow banking system is reminiscent of S Korea's financial system prior to its 1997 crisis. There, the fundamental problem the system was trying to solve (or evade) was that the legacy of a command-banking system meant not only that credit where available was mispriced, but also that great swathes of S Korea's commercial economy couldn't access official credit markets at all. For small companies, the problem wasn't the price of credit, it was the lack of all access to official credit, whilst for chaebol, credit was price too cheaply. There are distinct echoes of this in the difficulties China's major banks still have in the preferential lending to SOEs and similar.

In S Korea, the financial system first developed derivative-institutions which attempted to address this problem. However, whenever economy-wide cashflows faltered, those institutions tended to fail, because a) they were not yet good at credit-pricing; b) they tended to have major liquidity mis-matches, and c) when cashflows into the financial system got crimped, these derivative institutions would be banks' first to get their lines cut. In one form or another, this sequence of events was repeated about three times before S Korea began to solve the underlying problem.

In China's case, the two underlying problems are the unacknowledged fiscal problems at local level, and the dominant giant banks' historic mistrust of lending outside the traditionally favoured areas. The blooming of the shadow banks represented a dangerous attempt to solve (or evade) those problems. Even if the current efforts to de-fang the mistakes doubtless made by the shadow banking system are successful, we should expect some sort of re-iteration until the twin problems are solved.

US Productivity - Out of the Slow Lane?

Most economists and policymakers believe that since around 2005 the US has been stuck in a 'low productivity growth' regime in which labour productivity oscillates around an average of 1.33% yoy, compared with the previous 'high productivity growth' regime which held from early 1997 to late 2004 in which productivity averaged 2.9% a year.

The assumption that there’s a ceiling on US labour productivity growth, coupled with an assumption of a demographically-constricted ceiling on employment growth feeds into calculations about sustainable growth rates. And those calculations will go a long way to determining how the FOMC feels obliged to act.

But the low-productivity assumption is under challenge now, with recent readings knocking up against the ceiling: in 2Q labour productivity rose an annualized 3%, and the first-estimate of 3Q this week came in at an annualized 2.2%.

Taking real GDP per worker as a crude but easily calculable measure of productivity, the 1997 to 2004 average comes out at 2.2% with a standard deviation of 0.8%, whilst the 2005-current average comes in at 0.9% with a standard deviation of 1%. If we exclude the exceptional 'recovery period' of 4Q09 to 3Q10 (when this measure of productivity growth averaged 3.5%), then the low regime of productivity growth averages 0.7%, with a standard deviation of 0.7%.

If we now consider the last four quarters performance (4Q17 1%, 1Q18 1%, 2Q18 1.2%, 3Q18 1.3%), we can see that this is now consistently pushing towards the ceiling of this 'low productivity' regime.

productivity regimes.jpg

There are two possible interpretations of this, and both are in different ways destabilizing. One view is that there is no reason to expect the low-productivity regime to be challenged. In which case, we must expect to see the performance of the last two quarters as exceptional peak-cycle phenomena which will be swiftly followed by a cyclical decline. This is as good as it gets.

The alternative view is that there is nothing sacrosanct about the 'low productivity regime' and that the gains of the last two quarters are presaging a reversal back to a higher productivity regime.

What view you, and policymakers, take of this is absolutely critical, because it will be one of the assumptions which determine the possible sustainable growth trajectory of the US, and consequently what is the 'natural rate of interest' around which the Fed will attempt to guide the economy.

Despite the ambiguous/ambivalent structure of 3Q’s 3.5% GDP growth, I think there is a good reason to believe that the US is transitioning to a higher productivity regime. That good reason is to do with how capital stock per worker helps determine labour productivity growth.

As the chart below shows, between 2010 and late 2016, growth in labour productivity was constrained by negative or historically very weak growth in capital stock per worker. In fact, when one stripped out the impact of capital per worker, the performance of output per worker was markedly better than in the so-called 'high labour productivity' period.

Since 2016, growth in capital per worker has accelerated back towards the lower boundary of pre-crisis normality, but with surprisingly little deterioration in output per worker less capital per worker. Even into 3Q18, this measure remains noticeably better than what was usually achieved prior to the crisis. By itself, this chart gives no reason at all to expect any deterioration in overall output per worker in the near future. In other words, no reason to discount the exit from the post-2004 'low productivity regime'.

elements of productivity.jpg

UK Regional Inequality: Is the UK Still a Coherent Sample?

The previous piece (UK - The Curse of Regional Inequalities) ended by warning that if regional income inequalities continue to widen, at some point the UK loses coherence as a statistical sample. Or, to put it plain English, the inequalities can become so great that in economic terms we're no longer really talking about a single integrated society.

For better or worse, these judgements need not be made on the basis of prejudice or political outlook. Rather, it is the data which does the talking. The underlying premise is that many social and economic phenomena show a Zipfian pattern of distribution. (How to describe Zipfian distributions? It is a mathematical pattern which was first observed in the frequency of words in a language. Wikipedia to the rescue:'Zipf's law states that given a large sample of words used, the frequency of any word is inversely proportional to its rank in the frequency table. So word number n has a frequency proportional to 1/n. Thus the most frequent word will occur about twice as often as the second most frequent word, three times as often as the third most frequent word, etc.'

Zipfian distributions are also regularly observed in economic geography, most famously in the size/frequency distribution of cities within a nation (which, by the way, is how you can tell that London is the capital of Europe, rather than the UK).

We can use it to test whether the rising regional inequality in the UK is threatening its coherence as a society. As before, I am using the ONS's exemplary survey of household per capita disposable income, available annually from 1997 to 2016. In this case, I am using the NUTS3 breakdown of the UK into 179 different regions (NUTS stands for Nomenclature of Territorial Units for Statistics and is based on areas of local government). Basically, I am counting the number of NUTS3 units which fall into per capita annual income bands, stepping with increments of £500. The chart then logs both the frequency and the income (income along the x-axis, frequency along the y-axis).

In a single society one would expect the relationship between per capita income and frequence to be very regular - in fact, one would expect it to be roughly Zipfian. So how's it looking? First, this is the distribution pattern in 1997:

1997 coherence.jpg

There are three, maybe four, outliers of significant poverty, and three, maybe four, outliers of significant wealth. But these are comparatively rare, with the income/frequency of most of the regions clearly cleaving to a strong log/log trend. You might (or might not) be worried about the outliers at either end, but overall the sample looks strongly coherent.

And so on to 2007.

2007 coherence.jpg

The pattern hasn't changed much. The positive outliers are perhaps a little more obvious, and the negative outliers slightly more frequent, but the chart is still dominated by a very clear and strong trend linking income and frequency in a predictable relationship. The centre, we might say, is holding.

You cannot say the same thing for 2016's data-set.

2016 coherence.jpg

One can still see a rough sort of relationship at work, but it takes an uncomfortable effort of will. Had you not seen the pattern of 1997 and 2016, one would be tempted to see two conflicting trendlines at work here: one sloping upward and to the right, and the other sloping downward and to the left. Or one might see, perhaps four bands of frequency (at roughly 2.5, 2, 0.7 and zero).

I think the 2016 chart is very unsettling. Not primarily because of the evidence of inequality, but because it suggests that the complicated and extended network of economic relationships which normally result in a coherent pattern of economic activity and hence income distribution are looking extremely frayed. That suggests that the economic fortunes of different parts of the country no longer have the same impact or implication (positive or negative) on other parts that they used to. A rise of income in, say, Hounslow, no longer means much to, say, Wolverhampton or Leicester. Conversely, even a dramatic fall in fortunes in, say, Southampton, may not longer have any noticeable implications for Oxfordshire. So why should either region have any interest in the economic welfare of the other? Once again, putting things horribly simply, if the society no longer expresses (among other things) a complicated chain of commercial co-dependence, what does it express?

UK - The Curse of Regional Inequalities

I rarely write about the UK economy, because I fear delusion and confirmation bias. (Economists suffer badly from confirmation bias.) Nevertheless, there is one thing that needs to be understood by anyone who claims to know what the future holds for the UK. It is this: regional inequalities in income and wealth within the UK, and most particularly within England, are worse than at any time for which I have the data.

For decades, these inequalities have been widening and widening without significant political consequence. They may finally have become so wide that they constrain all sorts of policy-choices (for example, I suspect the Brexit vote is one consequence). There are a couple of economists who begin to be concerned - Jim O’Neill is one, and Bank of England’s Andy Haldane are both on the roll of honour. But I expect that the London-based financial services industry have neither the knowledge nor interest to prevent themselves from being repeatedly blindsided by the policy implications.

This is short-sighted, because without acknowledging the deep economic foundations of political dissatisfaction, analysts are unlikely to understand its durability or its longer-term consequences. The assumption that the political settlement which has produced this inequality can continue unchallenged must, surely, be naive.

The available data is unambiguous and damning. It is easiest to demonstrate by using the gap between London and elsewhere in the UK as the benchmark. The ONS has been producing extremely fine-grained survey data on household income by region and city since 1997. Back in 1997, the average London disposable per capita household income came to £13,183, compared to a UK average of £10,817 - a multiple of 1.22x. By 2018, London’s average had risen 106% to £27,151, whilst the UK’s average had risen only 80% to £19,432 - the multiple had risen to 1.40. There’s no reason to believe the multiple isn’t even higher today.

london uk income multiple.jpg

But this is not just a story about London’s success, it is also, alas, a story about failure elsewhere. In the North East, the rise in per capita income since 1997 is just 68%, and its 2016 £15,595 average is below anything London has had to get by on any time this century! (London’s income is 1.74x that of the North East).

The regional income inequalities continue to worsen. There are two ways one can illustrate this. First, one can express the standard deviation of regional disposable income as a percentage of the mean: the higher the number, the greater the relative dispersal of regional income. In 1997 the standard deviation of incomes was 13.7% of the mean: by 2016 it had risen to 18.6%.

standard deviation to mean.jpg

Second, by comparing the median to the mean, one can capture the degree to which the sample is skewed at the upper end. The same story emerges: back in 1997 the mean reading of per capita disposable income was 106% of the median - a level which it maintained with little change until around 2004. Peak-finance prior to the crash lifted the mean to around 108% of the median, where again it stayed, with wobbles, until around 2012. However, since then the ratio has taken off once again, and by 2016 it stood at 109%. In short, the regional dispersal of household income is increasingly skewed towards the richest regions.

mean to median.jpg

(Nor is it simply a story of London vs the rest: within London itself, the inequality has burgeoned. Consider: between 1997 and 2016, average disposable income in Hackney & Newham jumped 199%, whilst Bexley and Greenwich had to settle for a mere 66.3%. In 1998 the mean income of boroughs was 103% of the median, by 2016 that had risen to 122%.)

When it comes to wealth, the story is of course similar. The most obvious evidence comes from the property market, and in particular the way in which London house prices diverged from the rest of the UK in the aftermath of the GFC, and in particular after the introduction of quantitative easing by Bank of England in 1999. Between 2007 and 2016, average UK house prices (ie, including London houses) rose 13.9%, whilst during the same period London house prices rose 58.5% (the data comes from the ONS house price index series).

At the beginning of 2007, the average London house price was 1.5x that of the national average: by end-2016 the multiple had risen to 2.2x.

Since the national average price includes London itself, the expansion of the London multiple is unrealistically muted. Consider, then, the divergent course between prices in London and in Yorkshire (where I live). In the period that London’s prices rose 58.5%, Yorkshire’s prices rose. . . 1.6%. The London/Yorkshire multiple rose from 1.9x to about 3.2x. Or, to look at it from a Yorkshire point of view, when the City helped blow up the world financial system, quantitative easing gave Londoners a free house. Still, the current fall in London house prices provides the first suggestion that the epic divergence is discovering its limits.

london house multiple.jpg

Issues of inequality are easier to lament than either to diagnose or correct. Measures of disposable income and property wealth are neither complete or sufficient indicators of regional disparities. This piece could be extended to cover regional disparities of public spending on education, health and infrastructure (transport!). Conversely, it would also be possible to mount a rearguard defence of the widening inequalities as merely the result of a free market.

For now, though, the point is simpler: underlying every society is an assumption that it is coherent. In statistical terms, we can say that the distribution of resources and outcomes represent a coherent, usually Zipfian, sample. When inequalities become sufficiently great, they threaten to disrupt the coherence of the sample. If you don’t expect the disintegration of the UK, then you should expect policies designed - successfully or not - to close the regional inequalities which have opened up since the financial crisis.

China: Time's Up for Credit Squeeze - Upward Inflection Mid-2019

The race to pass judgement on China’s economy peaks four times a year, when China releases its quarterly GDP estimates. It involves a spectacular suspension of disbelief: does anyone really believe it is possible to measure an economy as large, complex and opaque as China’s to within a tenth of a percentage point?

There is little sensible to be said, therefore, about last week’s news that 3Q GDP growth slowed to 6.5% from 6.7% in 2Q.

But there are two conclusions one can draw about the structure of China’s economy, and its likely medium-term trajectory.

The first is that time’s up for the credit squeeze. The determination to improve the discipline of credit allocation was well-founded - back in 2011 every 100 yuan in new aggregate financing was associated with approximately 60 yuan in extra nominal GDP; but by 1Q16 this addition to GDP had fallen to just 26 yuan. From an economic point of view, credit creation was running into the sands.

However, two years into the attempt to discipline the allocation of credit the question China’s authorities have to face is this: does the efficiency of finances continue to rise, or has the negative economic impact of restraining credit growth led to an overall deterioration in financial efficiency?

Between 1Q16 and 2Q18 the credit crunch undoubtedly produced positive results in terms of the impact of extra finance upon nominal GDP: in the 12m to 1Q16, 100 yuan of extra aggregate financing was associated with only 26 yuan of extra nominal GDP; but by 2Q18 this had risen to 47 yuan. These sorts of gains are destined to be more difficult in the second year of a credit crunch as the easy gains have been made, the improvised sources of extra financing will already have been mobilized, and more companies will be hitting the wall. This is now showing in the 12m to 3Q, with the amount of extra GDP associated with 100 yuan of extra aggregate financing falling to 45.5 yuan.

And that, I think, accurately indicates the point at which maximum pressure for a policy loosening is felt, for not only is the pain of continued stringency now laid bare, but in addition, it becomes obvious to banks and monetary policymakers that the erosion of credit quality will itself tend to constrain further credit growth beyond what had been initially envisaged. We are now past that point.

financial efficiency.jpg

Now that we have now entered the phase where policies are set to limit the unforeseen/unexpected damage done by the previous credit restriction, what results can be expected from the current and likely ongoing policy relaxation?

I construct a monetary conditions indicator for China monitoring monetary growth (M1 and M2), changes in real interest rates and yield curves, and the size and historic volatility of changes in the SDR value of the Rmb. Significant changes in that indicator (on a 12ma basis) have in the past been associated with similar changes in the growth of private nominal domestic demand nine months later.

mci pri domestic demand.jpg

This chart suggests three things:

First: even on a 12m basis, China’s monetary conditions are now being loosened, quite aggressively, with 3Q18 being the upward inflection point.

Second: although the slowdown in growth of private domestic demand is moderating, it still has a further six to nine months before we should see a convincing upward inflection point. We can expect good news as a result of policy loosening . . . but not just yet

Third: each time monetary conditions have been loosened, dramatically, the positive impact on nominal GDP growth has weakened. The very sharp upswing seen in 2008 topped out at 26.1% yoy; in 2010 and 2011 it topped out at around 21%; in 2013 a similar loosening resulted only in nominal private sector domestic demand growth topping out at 13.8%. The positive upturn to be expected in 2H19 will probably be weaker still. If so, the upturn in 2019 may seem more like mere stabilization.

Investing, Fast & Slow. Part 3: Slow

Investing: Fast gave some hints as to how we can explore the likely market impact of System 1 type of thinking about economies and markets. The long-term results, I think, speak for themselves, but they are the result only of cumulative short-term investment tactics. System 2 type investment thinking aims at improving longer-term strategic positioning, allowing a longer investment horizon, less costly in terms of day-to-day effort and cost.

If the energy devoted to analytical and deliberative thinking about investment were easily repaid, the ‘strategic advice’ industry would be even bigger than it is.

Conversely, and ironically, it is the very sustained failure of these attempts that keeps the industry afloat: if anyone ever managed to think their way to consistent investing success, they would keep very quiet about it. But the history of technology demonstrates that key breakthroughs often have multiple discoverers who enjoy the same Eureka moment quite independently of each other around the same time. So even the most determined IP lawyer would be unable to stop the market-busting breakthrough finding a wider audience. At which point, there would be no point in markets anyway, since everyone would ‘know’ the information the market discovers beforehand.

So far, so negative. However, whilst no System 2 process will deliver a strategic method guaranteeing 100% success, there are economic tools which which quantitative analysis confirms do add value, systematically raising your chances of being invested in the right places at the right time. Here I’ll introduce two of them (albeit in conceptual terms only) which have survived the quant analysts’ autopsies: first, return on capital; and second, profits.

Both sound not mere familiar but bone-headedly obvious: why would anyone think of investing in equities if they didn’t have a view on return on capital and profits? Answer: because they’re macro-economists, and macro-economics historically has been relatively incurious about returns and profits.

Taken together these two not only help form the foundation of understanding where a particular economy is in its business cycle, but also, of course, builds a basis for comparison between the position and likely medium-term trajectory of other economies competing for your investment.

Return on Capital

I construct a Return on Capital Directional Indicator (ROCDI) by treating GDP as a flow of income from a stock of capital, and as that income fluctuates relative to the stock of capital. This is a simple idea, in macroeconomics related to Incremental Capital Output Ratio (ICOR) and in equity research to the asset turns (Sales/Total Assets) element of the Dupont Analysis decomposition of return on equity.

The key difficulty is estimating the size and growth of capital stock. I make an estimate by depreciating the gross fixed capital formation line-item in quarterly national accounts over a 10yr period. Whilst such a straightforward approach refuses to engage in a number of obvious (and interesting) questions, in practice the results concur closely with the results of more scrupulous surveys, where available.

Crucially, these calculations are made using only nominal data (rather than real): ‘real’ counts of investment in capital goods seem to me to be hopelessly compromised by a) the issues arising from the Cambridge Capital Controversies; and b) the key characteristic of capital goods is that their price fluctuates dramatically with business cycle, but in balance sheet accounting terms they remain on the books at purchase cost. That difference between capital goods market price and book price is, quite obviously, a key element in the business cycle dynamics we want to understand.

The result is that we can see not only the fluctuations in trends in return on capital, but also how investment activity fluctuates (and is likely to fluctuate) in response.

Here, for example, is how it looks for the US.


Kalecki Profits

Getting a grip on profits is less simple, but one can treat changes in corporate profits, or corporate saving, as a function of the changes in net investment plus the changes made in savings flows from the household sector, the government sector, and the ‘rest of the world’. To put it crudely, one sector at a time:

if the household sector’s savings ratio (defined as compensation minus consumer spending) diminishes and everything else is held constant, then corporate savings (ie, profits) will rise by the same amount;

If the government’s net (dis)savings position changes (ie, the fiscal deficit rises or narrows), and everything else is held constant, then corporate savings will fall or rise by the same amount;

If the rest of the world’s saving position with the economy (ie, the trade balance in its wide description) changes, and everything else is held constant, corporate savings will fluctuated by the same amount.

This is, in fact, nothing more than an economic accounting identity. To my knowledge, it was first articulated by a Polish Marxist-influenced economist Michal Kalecki in the 1930s, who initially used it to analyse the exploitation of the workers. Whilst it is not necessarily always easy (or possible) to calculate it with certainty, the results are extremely useful in observing the trajectory and force of business cycle dynamics.

kalecki profits.jpg

A Scamper Through Gold's Monetary History

If you find money mysterious - and I think anyone who has thought hard about it must - then monetary history is a constant source of revelation. One strand which is retreating into history concerns the role of gold. I have a piece on ‘Gold as Monetary Arbiter’ in the Global Dispatches magazine - it’s a quick scamper through some monetary history which you might not be familiar with.

More broadly, over the last few months I have been trying to read through the entire accumulated transcripts of the FOMC from the post-war meetings where the Fed is trying to finance the Cold War, to the 1971 abandonment of the dollar gold standard, to the extraordinary practical, philosophical and academic debates which engulfed the FOMC during and after the Great Financial Crisis.

It is the history of highly intelligent, highly motivated and informed people constantly discovering their mistakes, and repeatedly finding that even the basis on which they made those mistake were themselves mistaken.

On one level, it is an on-going demonstration of Hayek’s contention that economic policy-making faces impossible information problems. On another level, it is a history of people grappling with the fundamental contradiction of any reserve currency: too little money inhibits its use as a medium of exchange, too much money erodes its use as a store of value. But more mundanely, it is confirmation that no theoretical economic policy, and no econometric approximations, survives contact with unfolding reality for long. Everyone is constantly doing their best, and only fitfully finding that their best is, for the time being, good enough.

It is easy, then, to run through the history pointing to the errors, to the hubris of FOMC efforts. But truly this would be dishonest. The general tenor of FOMC discussions throughout the decades is one not inviting hubris, but irony: earnest irony. Even during what from current perspectives were the FOMC’s finest hours, board members quite openly acknowledge that they aren’t entirely sure what they are doing. Indeed, some of the most dramatic policy decisions are ventured despite profound uncertainty about whether the outcome might be good, or disastrous.

The thread which connects all the attempts of FOMC’s personnel to determine policy is the tension between reacting to current circumstances whilst at the same time protecting the value of money. This tension is always present, and even when seeming financial emergencies command the FOMC’s urgent attention, the philosophical backlash arrives rapidly after the immediate emergency retreats. What are we doing? What have we done?

In a free-floating monetary world, the search for a standard, a lever from which to move the world, goes on. Quite possibly it always will.

Investing, Fast and Slow. Part 2: Fast!

I don't mean frantic day-trading, chasing micro-trends before they evaporate. Rather, I mean working to create insight into what drives System 1 thinking about investment. If a portion of market-pricing at any one time is the result of System 1 thinking, we should expect it to be the result of impressions and feeling and instincts which eventually surface as the explicit beliefs expressed in the deliberate choices of System 2 thinking.

But these impressions and feelings will not be random. To quote Kahneman, System 1 type thinking 'has also learned skills . . . and understanding nuances of ... situations. Some skills . . . are acquired only by specialized experts. Others are widely shared.' The approach I adopt, then, is to work out what generates those impressions and feelings, and the understandings they generate. If my beliefs about the inputs into System 1 investment thinking are correct, it opens the possibility that a careful tracking of those inputs will allow me to understand and anticipate the results of System 1 investment choices.

This is what I have done, using as feedstocks the shocks & surprises indexes generated by my unusually broad and detailed tracking of global economic data. The very name 'shocks and surprises' conveys the hint - that these are precisely the data which System 1 type thinking is likely to notice and react to. Yet as Kahneman warns 'the automatic operations of System 1 generate surprisingly complex patterns of ideas,' and the intuitive belief that movements in asset prices will simply mimic the trajectory of shocks & surprises indexes turns out to be wrong. Such a simple interpretation will, alas, lose you money (so DO NOT TRY THIS AT HOME).

Nevertheless, further exploration has borne fruit. Alternating investment choices among the world's major stockmarkets using models approximating System 1 thinking has allowed a consistent and ultimately rather spectacular outperformance of the MSCI World Index since I started this at the beginning of 2014.

system 1 model.jpg

Since that period, the MSCI World Index has gained 25.2%, whilst my model has gained 80.7%. The average weekly gain for the MSCI World Index during that period is 0.11%, with a standard deviation of 1.74%; the weekly average for my System 1 model is +0.26%, with a standard deviation of 1.89%.

Conclusion? Investing Fast isn't about turning over your portfolio at rapid speed, it's about understanding that System 1 thinking is part of how market prices emerge in the short term.

Investing, Fast & Slow. Part 1

My starting point is that, one way or another, investing is and will remain a fundamentally human activity.

To be more specific, investing is and will remain a cultural phenomenon that has evolved in a way which contributes to the growing role of intelligence in the universe. Fairly obviously, that evolution is itself evolving different tool sets to help do the job. Quantitative modelling has developed a whole sert of tactics, strategies and expressions which are new branches on the tree, but the tree remains fundamentally the same.

That the task has remained fundamentally the same is not obvious: as each new tool has arrived, intelligent people have feared that the game is changed. (Here's how Robert Mayo, then head of the Chicago Fed discussed with the FOMC the rise of the options market in the 1970s: 'My elbow is still uneasy about this. There is a counterculture, if you want to distinguish it that formally, developing now with the evolution of puts, that will counterbalance the calls, and the market will be better again. But this is getting into real mystique. I don’t think anyone really knows what he’s talking about.')

Even when computer-driven trading represents the majority of stockmarket activity, I'm prepared to take it as axiomatic that investment will remain a fundamentally human activity. That's partly because no matter how much data is fed into the algorithms, market pricing will retain its ability to discover information which can be discovered in no other way (Hayek's thesis). Indeed, if computer programs alone were able to remove unprogrammable surprise element from pricing, then there would be essentially no purpose to markets in the first place - we'd be better off simply allowing the computers, not markets, to allocate resources. We know how that ends, and it's not well (ironically, we might say the data is in on it.)

The ultimate impact of computer-driven trading in all its variety is mitigated by the enduring fact that at both the top and bottom of the investment-decision tree, humans get involved. At the bottom of the tree, we have humans making decisions about whether to invest and how to invest. Having made the decision to save, and save in equities, what governs the choice between single stocks, mutual funds, ETFs, family offices, hedge funds, or CDFs and derivatives of various kinds? Price, fashion, advice?

When an institutional vehicle is chosen, the way that institutional vehicle runs itself is constrained by irreducible human factors, including fashion, 'respectability', theoretical justification (portfolio theory etc). the regulatory environment, and the sensitivities of investors and trustees. And finally, at the top of the investment decision, the entire investment environment is affected by the decisions of the 12 people on the FOMC. History provides abundant evidence of human intelligence and fallibility in their decisions.

Essentially then, investing is a human activity, and therefore an expression of human thought.

Which brings us to Daniel Kahneman and Amos Tversky and the book 'Thinking, Fast and Slow'. My guess is that you'll have recognized the reference in this piece's title. Kahneman's book gave birth to the field of behavioural economics and in its wake there have been a large number of usually unsuccessful attempts to deploy the insights of behavioural economics (ie, loss aversion etc) into investment strategy.

But it may be more useful to consider what he has to say more widely. At the heart of 'Thinking, Fast & Slow' is the description of two types of thinking. System 1 thinking is fast thinking, what we do, very often unknowingly or instinctively, to make snap decisions about situations we encounter all the time. By contrast, System 2 thinking is slow and deliberative, demanding an expenditure of energy in conscious effort. (And System 2 thinking is expensive: as someone said, 'Most of the world's problems can be solved with five minutes thought. But thinking is hard, and five minutes is a long time.')

It strikes me that since investing is an essentially human activity, we should understand that the results will always represent a mixture between the results of System 1 thinking and System 2 thinking. The suggested corollary is that if you rely solely on System 1 thinking (ie, you are a day-trader) you will sooner or later get beaten out of the market by the results of System 2 thinking/investing. But it's probably no less true that if you rely only on System 2 thinking, investing only in line with your carefully worked-out strategic insights, you will spend a lot of time getting killed by the 'irrational' activities of System 1 day-traders (who might well turn out to be algos anyway).

In other words, if investing is a fundamentally human activity, you need to have an approach which acknowledges and incorporates both System 1 and System 2 investing. You need to think about Investing, Fast & Slow.

I’ll start doing that in Part 2.

Korea Households Now in Net Debt

S Korea's 2Q flow of fund accounts, released today, at first sight look rather dour as far as the household sector is concerned. On deeper inspection, however, they are worse than that.

But first a caveat: taken broadly, the household sector remains in fine financial conditions, with net financial assets of W1,994.4tr (or US$1.78tr at 3Q exchange rate), and equivalent to 113% of GDP. The household sector has, in other words, net financial assets equivalent to more than an entire year's output.

Nevertheless, there are three problems:

First, during 2Q the household sector’s position deteriorated, falling by 0.7% qoq, or W14.68tr, thanks to a W18.2tr fall in holdings of equities. The 2Q results will, of course, have caught the full force of June's 8% fall in the KOSPI, from which there has as yet been no significant recovery. Still, the fall in net financial assets is a rare event: since 2009 it has happened only three times. In yoy terms, the growth has slowed to 4.1%, compared with a nominal GDP growth rate of 4.8%, and net financial assets/GDP ratio has been essentially stagnant since 2015.

The second problem owes nothing to market volatility: S Korea's households are now net debtors to the nation's credit institutions. During 2Q, household holdings of deposits are currency rose W16.78tr qoq, but loans rose W27.06tr. As a result, the net position of households with credit institutions deteriorated W10.29tr qoq to end the quarter with a net debt position of W5.1tr.

This is merely an extension of the continuing deterioration in household's banking position that's been continuing since 2015, but it is only the second time we've seen an actual net debt position. It is worth pointing out that it is now quite unusual in developed economies for the household sector to be net debtors to the banking system: net deposit positions are now the norm - a legacy of the financial crisis.

hhold net.jpg

Third, the willingness of the household sector to take on net debt has been useful in helping sustain S Korean domestic demand. In fact, the W10.29tr qoq deterioration in the position was equivalent to 69% of the qoq nominal GDP growth seen in 2Q. Over the last year, the deterioration came to W16.1tr, which was equivalent to 20% of the rise in nominal GDP growth. The implication is that if the Korean household sector becomes reluctant to continue letting its banking position slide, the impact will be felt on nominal GDP growth. And are they reluctant?

c conf.jpg

Asia's Economic Trajectories Diverge

It doesn’t take much thought to summarize current views on Asia: ‘Not interesting: US-China trade war; rising US interest rates.’ And the overall trajectory of Asia’s economic data doesn’t offer much challenge: Asia's overall shocks & surprises index has been mildly negative (with a short-lived reprieve in early June) since the end of January, and remains so today. However, the focus of weakness has changed sharply throughout the year.

asia ss.jpg

One of the advantages of tracking a large number of datapoints is that I can break this down these regional indexes into more specific countries or sub-regions. And that breakdown shows that there are quite distinct and divergent patterns of activity now emerging. Specifically, let’s look at the differences emerging between Greater China (China, Taiwan, HK), Japan, and the Rest of Asia.

asia ss split.jpg

Since mid-July, Japan's data has tended to break consensus or trend positively, and the Japan index is now persuasively strong. More significantly, over the last year, its trajectory has begun to move in concert with trends in the US, suggesting that the economic ties between these two are beginning once again to tighten, having decoupled around the middle of 2016. On this basis, Japan looks the principal Asian beneficiary of the vigour of the US’s business cycle.

By contrast, Greater China has sustained a period of disappointing results since the beginning of August, in a way which reverses the surprises achieved earlier in the year. It is not difficult to believe that this sustained disappointment is connected with the evolution of US-Chinese trade disputes.

Maybe the most interest result, however, is coming from the 'Rest of Asia, which since early September has strung together a series of surprises for the first time since the middle of January. As with Japan, it looks as if the Rest of Asia is beginning to feel the updraft from the US expansion.

This is a genuinely surprising development. It has happened despite the scepticism which has enveloped emerging markets globally in response an earlier strengthening of the dollar and an assumption of repeated US interest rate rises. And it has happened despite the mounting weakness of economic data in some major Asian economies (principally S Korea). But the recovery of the last few weeks has also coincided with the stabilization/weakening of the dollar, and the upturn in the Funk Index.

funk index.jpg

If it persists, the unexpected vitality of the 'rest of Asia' suggests that the positive impact on Asia of the US cyclical upswing is, at least for the time being, offsetting the negative impact of dollar strength and rising interest rates. It is an economic story which is being lost as commentators concentrate on US-China trade frictions and rising interest rates. If it continues, those stories will have to change.

Eurozone - Inflation Not 'Relatively Vigorous'

Messaging is taken as a crucial part of the central bank governor’s job, so we can’t be sure that headline writers and bond markets got it wrong this week when they latched on to ECB governor Mario Draghi’s description of Eurozone headline inflation as ‘relatively vigorous’. He did, after all, tell the European Parliament that ’underlying inflation is expected to increase further over the coming months as the tightening labour market is pushing up wage growth’.

In fact, the headlines belied Draghi’s rather drab forecasts: ‘Annual rates of HICP inflation are likely to hover around current levels in the coming months and are projected to reach 1.7% in each year between now and 2020. This stable profile conceals a slowing contribution from the non-core components of the general index, and a relatively vigorous pick-up in underlying inflation. Reflecting these dynamics, the ECB projections foresee inflation excluding food and energy reaching 1.8% in 2020.’

Given that in August, headline CPI was running at 2%, it is plain that ECB thinks headline inflation has already peaked, despite a modest uptick in core inflation (currently 1%). In general terms, that conforms with what one would expect if the deflections from 5yr seasonalized trends seen over the last six months are maintained. And given that those 6m deflections are already running at 0.8SDs for headline inflation and 0.9SDs for core inflation, these are already pretty punchy short-term forecasts.

cpi and core spread.jpg

In fact, it remains very difficult to make a case for a sustained uptick in the Eurozone’s inflation picture, with even the ECB’s longer-term steady-state at 1.7% looking hard to justify.

Rather, what evidence we have, including labour market evidence, gives no real hint that we are at an upward inflection point in Eurozone inflation. That evidence encompasses the state of inflationary expectations; the continued inability to pass through rises in factory prices to the consumer; and, crucially, labour market and wage trends.

First, consider expectations. Unlike in the US, where the inflationary expectations curve is now inverted, surveyed expectations in the Eurozone continue to expect a very modest uptick in inflation. On a quarterly basis, the ECB surveys professional forecasters, and compares that with consensus economics, and the European Commission’s monthly survey of consumer expectations: over a 2yr period, the expectation is for inflation of 1.6%-1.7%, with longer-term expectations of 1.8%-1.9%. These longer-term inflation expectations are almost entirely unchanged over the past four years.

But not only are longer-term expectations unchanged, they have fallen relative to 1yr inflation expectations. The Eurozone inflation expectations curve has not inverted, but is has quite clearly flattened over the last few years. Expectations, in other words, are an increasing damper on inflation, not a driver of it.

inflation curve.jpg

Second, there is no observable improvement of the ability of retailers to pass through increases in factory prices to consumers. Looking at the difference between monthly rises in PPI and monthly rises in CPI produces a CPI/PPI terms of trade. When it is rising, as it did between 2013 and 2016, it implies an improving ability of distributors of final goods and services to raise their margins. If there is underlying inflationary pressure from their suppliers, they are able to pass it along. But when the line falls, the reverse is true (ie, increased costs can’t be passed on to the consumer).

Since early 2017, the line has been stable, suggesting cost increases can be passed on, but final distributor margins are not rising. There is, in other words, consumer resistance to higher prices. As one would expect if underlying inflationary expectations are in retreat.

cpi ppi tot.jpg

The most curious feature of Mr Draghi’s presentation, however, was the apparent belief that inflation will be bolstered by wage pressures. This seems, on the face of it, utterly remarkable: the Eurozone’s unemployment rate is still above 8%: can it really be the case that the Eurozone’s labour market and supply-side rigidities are such that wage inflation kicks in at 8% unemployment? If so, the Eurozone is an extraordinary international outlier. The US unemployment rate of 3.9%, the UK rate of 4%, the Japanese rate of 2.5%, none of these has so far proved the foundation for inflationary wages rises.

If there is one economy within the Eurozone where this argument might develop some purchase, it is Germany, where the strictest definition of unemployment puts the rate at 3.4% (though the Bundesbank prefers a more generous figures of 5.2%).

But even in Germany, it is impossible to make the case that current wage rises are a potential source of inflation. In 2Q18, gross average monthly earnings growth was running at 2% yoy in Germany, which, deflated by CPI implies no real growth at all. Meanwhile, real GDP per worker was growing at 0.9% yoy. In other words, current wage rises are not even keeping up with productivity rises: if this continues for the medium term, the supply produced by Germany’s workers would therefore be rising faster than those worker’s ability to consume the product without either eroding savings or borrowing. Far from being inflationary, current wage settlements in Germany are currently disinflationary for the first time since 2012!

output per worker.jpg

Conclusion: If the Eurozone is about to see a modest sustained uplift in inflation, it will have to do so in the teeth of the underlying trends in wages vs productivity in its dominant economy.

Trade War Casualities or Business As Usual?

Let's imagine you are running a Chinese factory that exports to the US (maybe you are!), and all year President Trump has been tweeting his intentions to mess you up. How do you respond? I think your choices are: a) do nothing - he's bluffing; b) export like crazy because tomorrow you die; c) get your offshoring options nailed down.

What does this do to China's export statistics in the short term? a) nothing; b) spikes them around how; or c) depresses them but pushes up other Asian exports.

Now let's have a look at the evidence. First, China's exports are in only so-so health. The headline numbers are better than they have been for a few years: in August exports were up 9.5% yoy, and on a 12m basis they are rising 11.2%. But second, there's really no sign of any current sharp acceleration: in sequential terms, August's tally was 0.1SDs below where you'd expect it to be in August; in July the sequential movement it was down 0.5SDs. These results are not really consistent with the idea there's a lot of extra exporting in anticipation of Trump's tariffs.

china x.jpg

And third, putting China's export data in the context of regional export performance, there's no overall sign that its immediate neighbours (Japan, S Korea, Taiwan) are benefiting from much relocation of demand. In fact, China's share of Northeast Asian exports is rising, hitting 60.9% in August. But, as the chart shows, this recovery of market share has been underway since 2017, with little noticeable acceleration this year.

china x share.jpg

For all the angst, then, the evidence seems to suggest that if you're a Chinese exporter, for now you shrug your shoulders and keep doing what you do best - exporting.

China Debt: How Dangerous?

Not a month passes without someone somewhere assuring me that China’s debt bomb is about to blow. It’s been like that for years. China’s imminent financial explosion is a core part of the global worry-kit we all carry around with us. But as China’s debt/GDP ratio has risen and risen, the story has got increasingly plausible.

So let’s take another look at the data, and its context. Spoiler alert: yes it’s bad, but it’s not unprecedented, and probably not lethal provided we’re in an environment in which rates rise relatively modestly.

Taking the data from the BIS’s quarterly global credit survey for 4Q17, China’s private sector credit to GDP ratio ended last year at around 260% - it is probably slightly lower today.

debt gdp.jpg

How bad is that? It’s bad, but not entirely unprecedented:

  • For all emerging markets, the average stands at 176.5%;

  • For all reporting countries it stands at 244.4%;

  • For advanced economies it stands at 276%.

So it is obviously worse than emerging economies, somewhat worse than the global average, and slightly less than the average for advanced economies. .

Drilling down, it’s obvious where the problem is: private sector non-financial corporate debt ended 2017 at 160.3% of GDP.

corp debt gdp.jpg

That’s definitely out of line with what we see elsewhere in the world:

  • The average for emerging markets standing at 104.6%,

  • For all BIS reporting countries the average stands at 96.6%

  • For advanced economies, it stands at 91.6%

But whilst non-financial private corporate debt is sharply higher than elsewhere in the world, there are two things to notice. First, this ratio peaked in 2Q16, and had fallen 6.6pps by end-2017. And it will certainly be lower now. As a rule of thumb, it usually takes a couple of years of credit restrictions before the bad news of bankruptcies begins to be realized. If it’s going to happen dramatically, it should be happening around now.

Second, there is the perennial problem of knowing how to determine what really constitutes ‘private business’ rather than disguised state enterprises. Note that for practical financial purposes, this is not entirely a matter of checking the shareholder register (if any), since the Party will certainly have serious effective representation within effective management structures. (Though, of course, creditors should certainly not rely on ‘implicit guarantees’ offered by anything other than central government organizations.)

Nevertheless, part of the too-high corporate debt/GDP ratio is offset by China’s lower-than-expected government debt/ GDP ratio. In China, government debt is running at 47% of GDP, which compares with

  • Emerging markets average of 49%

  • All BIS countries average of 81%

  • Advanced economies ot 100.9%

Taking corporate and government debt together brings the total to 207% of GDP, which compares with 154% for emerging markets, and 193% for advanced economies. Bad, and worth watching, but not necessarily damnable.

With corporate debt/GDP now stable, where’s the current debt build-up coming from? The household sector of course. By end-2017 this was equivalent to 48.4% of GDP, and rising.

hhold debt gdp.jpg

How bad is that?

  • For emerging markets the average is 39.8%;

  • For all BIS countries the average is 62.1%,

  • For advanced countries it is 76.1%

The bit of private debt which is rising is h’hold debt, which is mainly mortgage debt. It continues to climb steeply, but the overall h’hold debt level does not look wildly out of the range where one should expect it to be.

Conclusion: China has a corporate debt problem, partly because China’s tax system doesn’t collect the revenues it needs to finance the growth it demands. It’s trying to deal with this problem, and is having some success in that, so far. The portion of debt which continues to grow in proportion to the economy is consumer debt. But for now, that’s not at atypical or worrying levels.

Interest Rates Rise: Spot the Credit Cycle?

The rise in US interest rates, echoed by tightening elsewhere in the world economy, is bad news for banking systems which:
i) show almost no sign of accelerating credit growth, and which
ii) are unlikely to be able to pass on the interest rate rises to borrowers at quite the same rate as they are obliged to raise their deposit rates.  

As a result, either banks will either have to accept lower interest rate margins, or they will protect those margins by raising the risk profile of their loanbooks. 

Anyone spotted a credit cycle?
Since the financial crisis, the economic recoveries seen in the US, the UK and now in the Eurozone have been long acyclical grinds, largely unsupported by any significant credit cycle. More, when we track the relatively small fluctuations in credit growth seen since then, most of the evidence suggests that the latest rise in interest rates is being made in economies where loan growth, and credit growth, is already slowing. 

If one takes all the bank lending of the US, Eurozone, China and Japan and expresses the total in US dollars, by February loan growth had slowed to 3.2% yoy, with the yoy rates falling almost continuously since the middle 2016, and with the 6m momentum deflecting slightly below historic seasonal trends. Moreover, comparisons are about to turn much more difficult: if current trends are maintained, yoy loan growth is likely to stop altogether by 3Q17, turning mildly negative thereafter.

Hints from China's 4Q GDP

China’s ‘real GDP’ headlines tell us little, with quarterly movements sticking very closely to both trend and official expectations. Nominal GDP, however, is accelerating sharply, which suggests: 

  1. rising inflationary pressures, denied by CPI but surfacing in PPI, but also;
  2. recovering return on capital, which underpins current profits growth and likely future GDP growth’; but also:
  3. the slowdown in growth of capital stock (down to nominal 8.4% yoy estimated in 2016) is implicated in China now losing market share of NE Asia exports. 
  4. the credit squeeze has produced a noticeable recovery in the marginal economic efficiency of finance, but there is a very long way to go before China is weaned off credit-fired growth; and
  5. the improvement in overall allocation of savings is much less convincing than the improvement in bank credit allocation. 

In short, the 4Q national accounts do show some progress being made in the structure of China’s growth. But the progress is painfully slow compared to the distance to be travelled, and is already developing unexpected diversions. 

US - Anything But Normal, Part II

Can the US could break out of the ‘new normal’ patterns of sluggish acyclical growth seen since 2010? Part I showed that improving trends in return on capital and real labour productivity generate a background in which we should expect rising capital investment and rising employment. At the least, this suggests an uptick in GDP growth in the short and medium term. In addition, it looked at two potential subcyclicals that could accelerate that growth into a genuine business cycle upswing - ie, a breakout from the ‘new normal’.  The dynamics for both the capital goods sector and for inventory management - both of which can act as business cycle accelerators - have been in a ‘negative adjustment’ phase over the last two years , but this has left them, ironically, increasingly sensitive to the current rise in global producer prices. 

But if all these provide fuel for a potential business cycle upswing, could these nevertheless be snuffed out by a deterioration in saving/consumption decisions in the household sector and, more generally, a tightening of credit and monetary conditions?  

This piece makes three points: 

  1. The current tightening in monetary conditions are not dramatic, and pose no material threat to GDP growth in 2017/18.
  2. Bond yields have already risen further than is easily explained either by current supply/demand conditions in bond markets, or by what consensus expects of Fed rate rises, and the prospects for inflation and GDP over the coming five quarters. In the absence of a monetary policy ‘mistake’ bond markets alone are not about to snuff out a significant acceleration in economic growth.
  3. Nevertheless, household savings rates fell gently but persistently throughout 2016 and are now at historically low levels. A reversion to the mean would cut approximately 30bps off GDP growth. In a ‘new normal’ scenario, in which GDP averages 2.1% with a standard deviation of 70bps, this is a material risk.